]> code.delx.au - gnu-emacs/blob - lisp/progmodes/cc-engine.el
Update copyright year to 2016
[gnu-emacs] / lisp / progmodes / cc-engine.el
1 ;;; cc-engine.el --- core syntax guessing engine for CC mode -*- coding: utf-8 -*-
2
3 ;; Copyright (C) 1985, 1987, 1992-2016 Free Software Foundation, Inc.
4
5 ;; Authors: 2001- Alan Mackenzie
6 ;; 1998- Martin Stjernholm
7 ;; 1992-1999 Barry A. Warsaw
8 ;; 1987 Dave Detlefs
9 ;; 1987 Stewart Clamen
10 ;; 1985 Richard M. Stallman
11 ;; Maintainer: bug-cc-mode@gnu.org
12 ;; Created: 22-Apr-1997 (split from cc-mode.el)
13 ;; Keywords: c languages
14 ;; Package: cc-mode
15
16 ;; This file is part of GNU Emacs.
17
18 ;; GNU Emacs is free software: you can redistribute it and/or modify
19 ;; it under the terms of the GNU General Public License as published by
20 ;; the Free Software Foundation, either version 3 of the License, or
21 ;; (at your option) any later version.
22
23 ;; GNU Emacs is distributed in the hope that it will be useful,
24 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
25 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26 ;; GNU General Public License for more details.
27
28 ;; You should have received a copy of the GNU General Public License
29 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
30
31 ;;; Commentary:
32
33 ;; The functions which have docstring documentation can be considered
34 ;; part of an API which other packages can use in CC Mode buffers.
35 ;; Otoh, undocumented functions and functions with the documentation
36 ;; in comments are considered purely internal and can change semantics
37 ;; or even disappear in the future.
38 ;;
39 ;; (This policy applies to CC Mode as a whole, not just this file. It
40 ;; probably also applies to many other Emacs packages, but here it's
41 ;; clearly spelled out.)
42
43 ;; Hidden buffer changes
44 ;;
45 ;; Various functions in CC Mode use text properties for caching and
46 ;; syntactic markup purposes, and those of them that might modify such
47 ;; properties but still don't modify the buffer in a visible way are
48 ;; said to do "hidden buffer changes". They should be used within
49 ;; `c-save-buffer-state' or a similar function that saves and restores
50 ;; buffer modifiedness, disables buffer change hooks, etc.
51 ;;
52 ;; Interactive functions are assumed to not do hidden buffer changes,
53 ;; except in the specific parts of them that do real changes.
54 ;;
55 ;; Lineup functions are assumed to do hidden buffer changes. They
56 ;; must not do real changes, though.
57 ;;
58 ;; All other functions that do hidden buffer changes have that noted
59 ;; in their doc string or comment.
60 ;;
61 ;; The intention with this system is to avoid wrapping every leaf
62 ;; function that do hidden buffer changes inside
63 ;; `c-save-buffer-state'. It should be used as near the top of the
64 ;; interactive functions as possible.
65 ;;
66 ;; Functions called during font locking are allowed to do hidden
67 ;; buffer changes since the font-lock package run them in a context
68 ;; similar to `c-save-buffer-state' (in fact, that function is heavily
69 ;; inspired by `save-buffer-state' in the font-lock package).
70
71 ;; Use of text properties
72 ;;
73 ;; CC Mode uses several text properties internally to mark up various
74 ;; positions, e.g. to improve speed and to eliminate glitches in
75 ;; interactive refontification.
76 ;;
77 ;; Note: This doc is for internal use only. Other packages should not
78 ;; assume that these text properties are used as described here.
79 ;;
80 ;; 'category
81 ;; Used for "indirection". With its help, some other property can
82 ;; be cheaply and easily switched on or off everywhere it occurs.
83 ;;
84 ;; 'syntax-table
85 ;; Used to modify the syntax of some characters. It is used to
86 ;; mark the "<" and ">" of angle bracket parens with paren syntax, and
87 ;; to "hide" obtrusive characters in preprocessor lines.
88 ;;
89 ;; This property is used on single characters and is therefore
90 ;; always treated as front and rear nonsticky (or start and end open
91 ;; in XEmacs vocabulary). It's therefore installed on
92 ;; `text-property-default-nonsticky' if that variable exists (Emacs
93 ;; >= 21).
94 ;;
95 ;; 'c-is-sws and 'c-in-sws
96 ;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
97 ;; speed them up. See the comment blurb before `c-put-is-sws'
98 ;; below for further details.
99 ;;
100 ;; 'c-type
101 ;; This property is used on single characters to mark positions with
102 ;; special syntactic relevance of various sorts. Its primary use is
103 ;; to avoid glitches when multiline constructs are refontified
104 ;; interactively (on font lock decoration level 3). It's cleared in
105 ;; a region before it's fontified and is then put on relevant chars
106 ;; in that region as they are encountered during the fontification.
107 ;; The value specifies the kind of position:
108 ;;
109 ;; 'c-decl-arg-start
110 ;; Put on the last char of the token preceding each declaration
111 ;; inside a declaration style arglist (typically in a function
112 ;; prototype).
113 ;;
114 ;; 'c-decl-end
115 ;; Put on the last char of the token preceding a declaration.
116 ;; This is used in cases where declaration boundaries can't be
117 ;; recognized simply by looking for a token like ";" or "}".
118 ;; `c-type-decl-end-used' must be set if this is used (see also
119 ;; `c-find-decl-spots').
120 ;;
121 ;; 'c-<>-arg-sep
122 ;; Put on the commas that separate arguments in angle bracket
123 ;; arglists like C++ template arglists.
124 ;;
125 ;; 'c-decl-id-start and 'c-decl-type-start
126 ;; Put on the last char of the token preceding each declarator
127 ;; in the declarator list of a declaration. They are also used
128 ;; between the identifiers cases like enum declarations.
129 ;; 'c-decl-type-start is used when the declarators are types,
130 ;; 'c-decl-id-start otherwise.
131 ;;
132 ;; 'c-awk-NL-prop
133 ;; Used in AWK mode to mark the various kinds of newlines. See
134 ;; cc-awk.el.
135
136 ;;; Code:
137
138 (eval-when-compile
139 (let ((load-path
140 (if (and (boundp 'byte-compile-dest-file)
141 (stringp byte-compile-dest-file))
142 (cons (file-name-directory byte-compile-dest-file) load-path)
143 load-path)))
144 (load "cc-bytecomp" nil t)))
145
146 (cc-require 'cc-defs)
147 (cc-require-when-compile 'cc-langs)
148 (cc-require 'cc-vars)
149
150 (eval-when-compile (require 'cl))
151
152 \f
153 ;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
154
155 (defmacro c-declare-lang-variables ()
156 `(progn
157 ,@(c--mapcan (lambda (init)
158 `(,(if (elt init 2)
159 `(defvar ,(car init) nil ,(elt init 2))
160 `(defvar ,(car init) nil))
161 (make-variable-buffer-local ',(car init))))
162 (cdr c-lang-variable-inits))))
163 (c-declare-lang-variables)
164
165 \f
166 ;;; Internal state variables.
167
168 ;; Internal state of hungry delete key feature
169 (defvar c-hungry-delete-key nil)
170 (make-variable-buffer-local 'c-hungry-delete-key)
171
172 ;; The electric flag (toggled by `c-toggle-electric-state').
173 ;; If t, electric actions (like automatic reindentation, and (if
174 ;; c-auto-newline is also set) auto newlining) will happen when an electric
175 ;; key like `{' is pressed (or an electric keyword like `else').
176 (defvar c-electric-flag t)
177 (make-variable-buffer-local 'c-electric-flag)
178
179 ;; Internal state of auto newline feature.
180 (defvar c-auto-newline nil)
181 (make-variable-buffer-local 'c-auto-newline)
182
183 ;; Included in the mode line to indicate the active submodes.
184 ;; (defvar c-submode-indicators nil)
185 ;; (make-variable-buffer-local 'c-submode-indicators)
186
187 (defun c-calculate-state (arg prevstate)
188 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
189 ;; arg is nil or zero, toggle the state. If arg is negative, turn
190 ;; the state off, and if arg is positive, turn the state on
191 (if (or (not arg)
192 (zerop (setq arg (prefix-numeric-value arg))))
193 (not prevstate)
194 (> arg 0)))
195
196 \f
197 ;; Basic handling of preprocessor directives.
198
199 ;; This is a dynamically bound cache used together with
200 ;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
201 ;; works as long as point doesn't cross a macro boundary.
202 (defvar c-macro-start 'unknown)
203
204 (defsubst c-query-and-set-macro-start ()
205 (if (symbolp c-macro-start)
206 (setq c-macro-start (save-excursion
207 (c-save-buffer-state ()
208 (and (c-beginning-of-macro)
209 (point)))))
210 c-macro-start))
211
212 (defsubst c-query-macro-start ()
213 (if (symbolp c-macro-start)
214 (save-excursion
215 (c-save-buffer-state ()
216 (and (c-beginning-of-macro)
217 (point))))
218 c-macro-start))
219
220 ;; One element macro cache to cope with continual movement within very large
221 ;; CPP macros.
222 (defvar c-macro-cache nil)
223 (make-variable-buffer-local 'c-macro-cache)
224 ;; Nil or cons of the bounds of the most recent CPP form probed by
225 ;; `c-beginning-of-macro', `c-end-of-macro' or `c-syntactic-end-of-macro'.
226 ;; The cdr will be nil if we know only the start of the CPP form.
227 (defvar c-macro-cache-start-pos nil)
228 (make-variable-buffer-local 'c-macro-cache-start-pos)
229 ;; The starting position from where we determined `c-macro-cache'.
230 (defvar c-macro-cache-syntactic nil)
231 (make-variable-buffer-local 'c-macro-cache-syntactic)
232 ;; non-nil iff `c-macro-cache' has both elements set AND the cdr is at a
233 ;; syntactic end of macro, not merely an apparent one.
234
235 (defun c-invalidate-macro-cache (beg end)
236 ;; Called from a before-change function. If the change region is before or
237 ;; in the macro characterized by `c-macro-cache' etc., nullify it
238 ;; appropriately. BEG and END are the standard before-change-functions
239 ;; parameters. END isn't used.
240 (cond
241 ((null c-macro-cache))
242 ((< beg (car c-macro-cache))
243 (setq c-macro-cache nil
244 c-macro-cache-start-pos nil
245 c-macro-cache-syntactic nil))
246 ((and (cdr c-macro-cache)
247 (< beg (cdr c-macro-cache)))
248 (setcdr c-macro-cache nil)
249 (setq c-macro-cache-start-pos beg
250 c-macro-cache-syntactic nil))))
251
252 (defun c-macro-is-genuine-p ()
253 ;; Check that the ostensible CPP construct at point is a real one. In
254 ;; particular, if point is on the first line of a narrowed buffer, make sure
255 ;; that the "#" isn't, say, the second character of a "##" operator. Return
256 ;; t when the macro is real, nil otherwise.
257 (let ((here (point)))
258 (beginning-of-line)
259 (prog1
260 (if (and (eq (point) (point-min))
261 (/= (point) 1))
262 (save-restriction
263 (widen)
264 (beginning-of-line)
265 (and (looking-at c-anchored-cpp-prefix)
266 (eq (match-beginning 1) here)))
267 t)
268 (goto-char here))))
269
270 (defun c-beginning-of-macro (&optional lim)
271 "Go to the beginning of a preprocessor directive.
272 Leave point at the beginning of the directive and return t if in one,
273 otherwise return nil and leave point unchanged.
274
275 Note that this function might do hidden buffer changes. See the
276 comment at the start of cc-engine.el for more info."
277 (let ((here (point)))
278 (when c-opt-cpp-prefix
279 (if (and (car c-macro-cache)
280 (>= (point) (car c-macro-cache))
281 (or (and (cdr c-macro-cache)
282 (<= (point) (cdr c-macro-cache)))
283 (<= (point) c-macro-cache-start-pos)))
284 (unless (< (car c-macro-cache) (or lim (point-min)))
285 (progn (goto-char (max (or lim (point-min)) (car c-macro-cache)))
286 (setq c-macro-cache-start-pos
287 (max c-macro-cache-start-pos here))
288 t))
289 (setq c-macro-cache nil
290 c-macro-cache-start-pos nil
291 c-macro-cache-syntactic nil)
292
293 (save-restriction
294 (if lim (narrow-to-region lim (point-max)))
295 (beginning-of-line)
296 (while (eq (char-before (1- (point))) ?\\)
297 (forward-line -1))
298 (back-to-indentation)
299 (if (and (<= (point) here)
300 (looking-at c-opt-cpp-start)
301 (c-macro-is-genuine-p))
302 (progn
303 (setq c-macro-cache (cons (point) nil)
304 c-macro-cache-start-pos here)
305 t)
306 (goto-char here)
307 nil))))))
308
309 (defun c-end-of-macro ()
310 "Go to the end of a preprocessor directive.
311 More accurately, move the point to the end of the closest following
312 line that doesn't end with a line continuation backslash - no check is
313 done that the point is inside a cpp directive to begin with.
314
315 Note that this function might do hidden buffer changes. See the
316 comment at the start of cc-engine.el for more info."
317 (if (and (cdr c-macro-cache)
318 (<= (point) (cdr c-macro-cache))
319 (>= (point) (car c-macro-cache)))
320 (goto-char (cdr c-macro-cache))
321 (unless (and (car c-macro-cache)
322 (<= (point) c-macro-cache-start-pos)
323 (>= (point) (car c-macro-cache)))
324 (setq c-macro-cache nil
325 c-macro-cache-start-pos nil
326 c-macro-cache-syntactic nil))
327 (while (progn
328 (end-of-line)
329 (when (and (eq (char-before) ?\\)
330 (not (eobp)))
331 (forward-char)
332 t)))
333 (when (car c-macro-cache)
334 (setcdr c-macro-cache (point)))))
335
336 (defun c-syntactic-end-of-macro ()
337 ;; Go to the end of a CPP directive, or a "safe" pos just before.
338 ;;
339 ;; This is normally the end of the next non-escaped line. A "safe"
340 ;; position is one not within a string or comment. (The EOL on a line
341 ;; comment is NOT "safe").
342 ;;
343 ;; This function must only be called from the beginning of a CPP construct.
344 ;;
345 ;; Note that this function might do hidden buffer changes. See the comment
346 ;; at the start of cc-engine.el for more info.
347 (let* ((here (point))
348 (there (progn (c-end-of-macro) (point)))
349 s)
350 (unless c-macro-cache-syntactic
351 (setq s (parse-partial-sexp here there))
352 (while (and (or (nth 3 s) ; in a string
353 (nth 4 s)) ; in a comment (maybe at end of line comment)
354 (> there here)) ; No infinite loops, please.
355 (setq there (1- (nth 8 s)))
356 (setq s (parse-partial-sexp here there)))
357 (setq c-macro-cache-syntactic (car c-macro-cache)))
358 (point)))
359
360 (defun c-forward-over-cpp-define-id ()
361 ;; Assuming point is at the "#" that introduces a preprocessor
362 ;; directive, it's moved forward to the end of the identifier which is
363 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
364 ;; is returned in this case, in all other cases nil is returned and
365 ;; point isn't moved.
366 ;;
367 ;; This function might do hidden buffer changes.
368 (when (and c-opt-cpp-macro-define-id
369 (looking-at c-opt-cpp-macro-define-id))
370 (goto-char (match-end 0))))
371
372 (defun c-forward-to-cpp-define-body ()
373 ;; Assuming point is at the "#" that introduces a preprocessor
374 ;; directive, it's moved forward to the start of the definition body
375 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
376 ;; specifies). Non-nil is returned in this case, in all other cases
377 ;; nil is returned and point isn't moved.
378 ;;
379 ;; This function might do hidden buffer changes.
380 (when (and c-opt-cpp-macro-define-start
381 (looking-at c-opt-cpp-macro-define-start)
382 (not (= (match-end 0) (c-point 'eol))))
383 (goto-char (match-end 0))))
384
385 \f
386 ;;; Basic utility functions.
387
388 (defun c-syntactic-content (from to paren-level)
389 ;; Return the given region as a string where all syntactic
390 ;; whitespace is removed or, where necessary, replaced with a single
391 ;; space. If PAREN-LEVEL is given then all parens in the region are
392 ;; collapsed to "()", "[]" etc.
393 ;;
394 ;; This function might do hidden buffer changes.
395
396 (save-excursion
397 (save-restriction
398 (narrow-to-region from to)
399 (goto-char from)
400 (let* ((parts (list nil)) (tail parts) pos in-paren)
401
402 (while (re-search-forward c-syntactic-ws-start to t)
403 (goto-char (setq pos (match-beginning 0)))
404 (c-forward-syntactic-ws)
405 (if (= (point) pos)
406 (forward-char)
407
408 (when paren-level
409 (save-excursion
410 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
411 pos (point))))
412
413 (if (and (> pos from)
414 (< (point) to)
415 (looking-at "\\w\\|\\s_")
416 (save-excursion
417 (goto-char (1- pos))
418 (looking-at "\\w\\|\\s_")))
419 (progn
420 (setcdr tail (list (buffer-substring-no-properties from pos)
421 " "))
422 (setq tail (cddr tail)))
423 (setcdr tail (list (buffer-substring-no-properties from pos)))
424 (setq tail (cdr tail)))
425
426 (when in-paren
427 (when (= (car (parse-partial-sexp pos to -1)) -1)
428 (setcdr tail (list (buffer-substring-no-properties
429 (1- (point)) (point))))
430 (setq tail (cdr tail))))
431
432 (setq from (point))))
433
434 (setcdr tail (list (buffer-substring-no-properties from to)))
435 (apply 'concat (cdr parts))))))
436
437 (defun c-shift-line-indentation (shift-amt)
438 ;; Shift the indentation of the current line with the specified
439 ;; amount (positive inwards). The buffer is modified only if
440 ;; SHIFT-AMT isn't equal to zero.
441 (let ((pos (- (point-max) (point)))
442 (c-macro-start c-macro-start)
443 tmp-char-inserted)
444 (if (zerop shift-amt)
445 nil
446 ;; If we're on an empty line inside a macro, we take the point
447 ;; to be at the current indentation and shift it to the
448 ;; appropriate column. This way we don't treat the extra
449 ;; whitespace out to the line continuation as indentation.
450 (when (and (c-query-and-set-macro-start)
451 (looking-at "[ \t]*\\\\$")
452 (save-excursion
453 (skip-chars-backward " \t")
454 (bolp)))
455 (insert ?x)
456 (backward-char)
457 (setq tmp-char-inserted t))
458 (unwind-protect
459 (let ((col (current-indentation)))
460 (delete-region (c-point 'bol) (c-point 'boi))
461 (beginning-of-line)
462 (indent-to (+ col shift-amt)))
463 (when tmp-char-inserted
464 (delete-char 1))))
465 ;; If initial point was within line's indentation and we're not on
466 ;; a line with a line continuation in a macro, position after the
467 ;; indentation. Else stay at same point in text.
468 (if (and (< (point) (c-point 'boi))
469 (not tmp-char-inserted))
470 (back-to-indentation)
471 (if (> (- (point-max) pos) (point))
472 (goto-char (- (point-max) pos))))))
473
474 (defsubst c-keyword-sym (keyword)
475 ;; Return non-nil if the string KEYWORD is a known keyword. More
476 ;; precisely, the value is the symbol for the keyword in
477 ;; `c-keywords-obarray'.
478 (intern-soft keyword c-keywords-obarray))
479
480 (defsubst c-keyword-member (keyword-sym lang-constant)
481 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
482 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
483 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
484 ;; nil then the result is nil.
485 (get keyword-sym lang-constant))
486
487 ;; String syntax chars, suitable for skip-syntax-(forward|backward).
488 (defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
489 "\"|"
490 "\""))
491
492 ;; Regexp matching string limit syntax.
493 (defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
494 "\\s\"\\|\\s|"
495 "\\s\""))
496
497 ;; Regexp matching WS followed by string limit syntax.
498 (defconst c-ws*-string-limit-regexp
499 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
500
501 ;; Holds formatted error strings for the few cases where parse errors
502 ;; are reported.
503 (defvar c-parsing-error nil)
504 (make-variable-buffer-local 'c-parsing-error)
505
506 (defun c-echo-parsing-error (&optional quiet)
507 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
508 (c-benign-error "%s" c-parsing-error))
509 c-parsing-error)
510
511 ;; Faces given to comments and string literals. This is used in some
512 ;; situations to speed up recognition; it isn't mandatory that font
513 ;; locking is in use. This variable is extended with the face in
514 ;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
515 (defvar c-literal-faces
516 (append '(font-lock-comment-face font-lock-string-face)
517 (when (facep 'font-lock-comment-delimiter-face)
518 ;; New in Emacs 22.
519 '(font-lock-comment-delimiter-face))))
520
521 (defsubst c-put-c-type-property (pos value)
522 ;; Put a c-type property with the given value at POS.
523 (c-put-char-property pos 'c-type value))
524
525 (defun c-clear-c-type-property (from to value)
526 ;; Remove all occurrences of the c-type property that has the given
527 ;; value in the region between FROM and TO. VALUE is assumed to not
528 ;; be nil.
529 ;;
530 ;; Note: This assumes that c-type is put on single chars only; it's
531 ;; very inefficient if matching properties cover large regions.
532 (save-excursion
533 (goto-char from)
534 (while (progn
535 (when (eq (get-text-property (point) 'c-type) value)
536 (c-clear-char-property (point) 'c-type))
537 (goto-char (c-next-single-property-change (point) 'c-type nil to))
538 (< (point) to)))))
539
540 \f
541 ;; Some debug tools to visualize various special positions. This
542 ;; debug code isn't as portable as the rest of CC Mode.
543
544 (cc-bytecomp-defun overlays-in)
545 (cc-bytecomp-defun overlay-get)
546 (cc-bytecomp-defun overlay-start)
547 (cc-bytecomp-defun overlay-end)
548 (cc-bytecomp-defun delete-overlay)
549 (cc-bytecomp-defun overlay-put)
550 (cc-bytecomp-defun make-overlay)
551
552 (defun c-debug-add-face (beg end face)
553 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
554 (while overlays
555 (setq overlay (car overlays)
556 overlays (cdr overlays))
557 (when (eq (overlay-get overlay 'face) face)
558 (setq beg (min beg (overlay-start overlay))
559 end (max end (overlay-end overlay)))
560 (delete-overlay overlay)))
561 (overlay-put (make-overlay beg end) 'face face)))
562
563 (defun c-debug-remove-face (beg end face)
564 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
565 (ol-beg beg) (ol-end end))
566 (while overlays
567 (setq overlay (car overlays)
568 overlays (cdr overlays))
569 (when (eq (overlay-get overlay 'face) face)
570 (setq ol-beg (min ol-beg (overlay-start overlay))
571 ol-end (max ol-end (overlay-end overlay)))
572 (delete-overlay overlay)))
573 (when (< ol-beg beg)
574 (overlay-put (make-overlay ol-beg beg) 'face face))
575 (when (> ol-end end)
576 (overlay-put (make-overlay end ol-end) 'face face))))
577
578 \f
579 ;; `c-beginning-of-statement-1' and accompanying stuff.
580
581 ;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
582 ;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
583 ;; better way should be implemented, but this will at least shut up
584 ;; the byte compiler.
585 (defvar c-maybe-labelp)
586
587 ;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
588
589 ;; Macros used internally in c-beginning-of-statement-1 for the
590 ;; automaton actions.
591 (defmacro c-bos-push-state ()
592 '(setq stack (cons (cons state saved-pos)
593 stack)))
594 (defmacro c-bos-pop-state (&optional do-if-done)
595 `(if (setq state (car (car stack))
596 saved-pos (cdr (car stack))
597 stack (cdr stack))
598 t
599 ,do-if-done
600 (throw 'loop nil)))
601 (defmacro c-bos-pop-state-and-retry ()
602 '(throw 'loop (setq state (car (car stack))
603 saved-pos (cdr (car stack))
604 ;; Throw nil if stack is empty, else throw non-nil.
605 stack (cdr stack))))
606 (defmacro c-bos-save-pos ()
607 '(setq saved-pos (vector pos tok ptok pptok)))
608 (defmacro c-bos-restore-pos ()
609 '(unless (eq (elt saved-pos 0) start)
610 (setq pos (elt saved-pos 0)
611 tok (elt saved-pos 1)
612 ptok (elt saved-pos 2)
613 pptok (elt saved-pos 3))
614 (goto-char pos)
615 (setq sym nil)))
616 (defmacro c-bos-save-error-info (missing got)
617 `(setq saved-pos (vector pos ,missing ,got)))
618 (defmacro c-bos-report-error ()
619 '(unless noerror
620 (setq c-parsing-error
621 (format-message
622 "No matching `%s' found for `%s' on line %d"
623 (elt saved-pos 1)
624 (elt saved-pos 2)
625 (1+ (count-lines (point-min)
626 (c-point 'bol (elt saved-pos 0))))))))
627
628 (defun c-beginning-of-statement-1 (&optional lim ignore-labels
629 noerror comma-delim)
630 "Move to the start of the current statement or declaration, or to
631 the previous one if already at the beginning of one. Only
632 statements/declarations on the same level are considered, i.e. don't
633 move into or out of sexps (not even normal expression parentheses).
634
635 If point is already at the earliest statement within braces or parens,
636 this function doesn't move back into any whitespace preceding it; it
637 returns `same' in this case.
638
639 Stop at statement continuation tokens like \"else\", \"catch\",
640 \"finally\" and the \"while\" in \"do ... while\" if the start point
641 is within the continuation. If starting at such a token, move to the
642 corresponding statement start. If at the beginning of a statement,
643 move to the closest containing statement if there is any. This might
644 also stop at a continuation clause.
645
646 Labels are treated as part of the following statements if
647 IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
648 statement start keyword.) Otherwise, each label is treated as a
649 separate statement.
650
651 Macros are ignored \(i.e. skipped over) unless point is within one, in
652 which case the content of the macro is treated as normal code. Aside
653 from any normal statement starts found in it, stop at the first token
654 of the content in the macro, i.e. the expression of an \"#if\" or the
655 start of the definition in a \"#define\". Also stop at start of
656 macros before leaving them.
657
658 Return:
659 `label' if stopped at a label or \"case...:\" or \"default:\";
660 `same' if stopped at the beginning of the current statement;
661 `up' if stepped to a containing statement;
662 `previous' if stepped to a preceding statement;
663 `beginning' if stepped from a statement continuation clause to
664 its start clause; or
665 `macro' if stepped to a macro start.
666 Note that `same' and not `label' is returned if stopped at the same
667 label without crossing the colon character.
668
669 LIM may be given to limit the search. If the search hits the limit,
670 point will be left at the closest following token, or at the start
671 position if that is less (`same' is returned in this case).
672
673 NOERROR turns off error logging to `c-parsing-error'.
674
675 Normally only `;' and virtual semicolons are considered to delimit
676 statements, but if COMMA-DELIM is non-nil then `,' is treated
677 as a delimiter too.
678
679 Note that this function might do hidden buffer changes. See the
680 comment at the start of cc-engine.el for more info."
681
682 ;; The bulk of this function is a pushdown automaton that looks at statement
683 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
684 ;; purpose is to keep track of nested statements, ensuring that such
685 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
686 ;; does with nested braces/brackets/parentheses).
687 ;;
688 ;; Note: The position of a boundary is the following token.
689 ;;
690 ;; Beginning with the current token (the one following point), move back one
691 ;; sexp at a time (where a sexp is, more or less, either a token or the
692 ;; entire contents of a brace/bracket/paren pair). Each time a statement
693 ;; boundary is crossed or a "while"-like token is found, update the state of
694 ;; the PDA. Stop at the beginning of a statement when the stack (holding
695 ;; nested statement info) is empty and the position has been moved.
696 ;;
697 ;; The following variables constitute the PDA:
698 ;;
699 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
700 ;; scanned back over, 'boundary if we've just gone back over a
701 ;; statement boundary, or nil otherwise.
702 ;; state: takes one of the values (nil else else-boundary while
703 ;; while-boundary catch catch-boundary).
704 ;; nil means "no "while"-like token yet scanned".
705 ;; 'else, for example, means "just gone back over an else".
706 ;; 'else-boundary means "just gone back over a statement boundary
707 ;; immediately after having gone back over an else".
708 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
709 ;; of error reporting information.
710 ;; stack: The stack onto which the PDA pushes its state. Each entry
711 ;; consists of a saved value of state and saved-pos. An entry is
712 ;; pushed when we move back over a "continuation" token (e.g. else)
713 ;; and popped when we encounter the corresponding opening token
714 ;; (e.g. if).
715 ;;
716 ;;
717 ;; The following diagram briefly outlines the PDA.
718 ;;
719 ;; Common state:
720 ;; "else": Push state, goto state `else'.
721 ;; "while": Push state, goto state `while'.
722 ;; "catch" or "finally": Push state, goto state `catch'.
723 ;; boundary: Pop state.
724 ;; other: Do nothing special.
725 ;;
726 ;; State `else':
727 ;; boundary: Goto state `else-boundary'.
728 ;; other: Error, pop state, retry token.
729 ;;
730 ;; State `else-boundary':
731 ;; "if": Pop state.
732 ;; boundary: Error, pop state.
733 ;; other: See common state.
734 ;;
735 ;; State `while':
736 ;; boundary: Save position, goto state `while-boundary'.
737 ;; other: Pop state, retry token.
738 ;;
739 ;; State `while-boundary':
740 ;; "do": Pop state.
741 ;; boundary: Restore position if it's not at start, pop state. [*see below]
742 ;; other: See common state.
743 ;;
744 ;; State `catch':
745 ;; boundary: Goto state `catch-boundary'.
746 ;; other: Error, pop state, retry token.
747 ;;
748 ;; State `catch-boundary':
749 ;; "try": Pop state.
750 ;; "catch": Goto state `catch'.
751 ;; boundary: Error, pop state.
752 ;; other: See common state.
753 ;;
754 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
755 ;; searching for a "do" which would have opened a do-while. If we didn't
756 ;; find it, we discard the analysis done since the "while", go back to this
757 ;; token in the buffer and restart the scanning there, this time WITHOUT
758 ;; pushing the 'while state onto the stack.
759 ;;
760 ;; In addition to the above there is some special handling of labels
761 ;; and macros.
762
763 (let ((case-fold-search nil)
764 (start (point))
765 macro-start
766 (delims (if comma-delim '(?\; ?,) '(?\;)))
767 (c-stmt-delim-chars (if comma-delim
768 c-stmt-delim-chars-with-comma
769 c-stmt-delim-chars))
770 c-in-literal-cache c-maybe-labelp after-case:-pos saved
771 ;; Current position.
772 pos
773 ;; Position of last stmt boundary character (e.g. ;).
774 boundary-pos
775 ;; The position of the last sexp or bound that follows the
776 ;; first found colon, i.e. the start of the nonlabel part of
777 ;; the statement. It's `start' if a colon is found just after
778 ;; the start.
779 after-labels-pos
780 ;; Like `after-labels-pos', but the first such position inside
781 ;; a label, i.e. the start of the last label before the start
782 ;; of the nonlabel part of the statement.
783 last-label-pos
784 ;; The last position where a label is possible provided the
785 ;; statement started there. It's nil as long as no invalid
786 ;; label content has been found (according to
787 ;; `c-nonlabel-token-key'). It's `start' if no valid label
788 ;; content was found in the label. Note that we might still
789 ;; regard it a label if it starts with `c-label-kwds'.
790 label-good-pos
791 ;; Putative positions of the components of a bitfield declaration,
792 ;; e.g. "int foo : NUM_FOO_BITS ;"
793 bitfield-type-pos bitfield-id-pos bitfield-size-pos
794 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
795 ;; See above.
796 sym
797 ;; Current state in the automaton. See above.
798 state
799 ;; Current saved positions. See above.
800 saved-pos
801 ;; Stack of conses (state . saved-pos).
802 stack
803 ;; Regexp which matches "for", "if", etc.
804 (cond-key (or c-opt-block-stmt-key
805 "\\<\\>")) ; Matches nothing.
806 ;; Return value.
807 (ret 'same)
808 ;; Positions of the last three sexps or bounds we've stopped at.
809 tok ptok pptok)
810
811 (save-restriction
812 (if lim (narrow-to-region lim (point-max)))
813
814 (if (save-excursion
815 (and (c-beginning-of-macro)
816 (/= (point) start)))
817 (setq macro-start (point)))
818
819 ;; Try to skip back over unary operator characters, to register
820 ;; that we've moved.
821 (while (progn
822 (setq pos (point))
823 (c-backward-syntactic-ws)
824 ;; Protect post-++/-- operators just before a virtual semicolon.
825 (and (not (c-at-vsemi-p))
826 (/= (skip-chars-backward "-+!*&~@`#") 0))))
827
828 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
829 ;; done. Later on we ignore the boundaries for statements that don't
830 ;; contain any sexp. The only thing that is affected is that the error
831 ;; checking is a little less strict, and we really don't bother.
832 (if (and (memq (char-before) delims)
833 (progn (forward-char -1)
834 (setq saved (point))
835 (c-backward-syntactic-ws)
836 (or (memq (char-before) delims)
837 (memq (char-before) '(?: nil))
838 (eq (char-syntax (char-before)) ?\()
839 (c-at-vsemi-p))))
840 (setq ret 'previous
841 pos saved)
842
843 ;; Begin at start and not pos to detect macros if we stand
844 ;; directly after the #.
845 (goto-char start)
846 (if (looking-at "\\<\\|\\W")
847 ;; Record this as the first token if not starting inside it.
848 (setq tok start))
849
850 ;; The following while loop goes back one sexp (balanced parens,
851 ;; etc. with contents, or symbol or suchlike) each iteration. This
852 ;; movement is accomplished with a call to c-backward-sexp approx 170
853 ;; lines below.
854 ;;
855 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
856 ;; 1. On reaching the start of a macro;
857 ;; 2. On having passed a stmt boundary with the PDA stack empty;
858 ;; 3. On reaching the start of an Objective C method def;
859 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
860 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
861 (while
862 (catch 'loop ;; Throw nil to break, non-nil to continue.
863 (cond
864 ;; Are we in a macro, just after the opening #?
865 ((save-excursion
866 (and macro-start ; Always NIL for AWK.
867 (progn (skip-chars-backward " \t")
868 (eq (char-before) ?#))
869 (progn (setq saved (1- (point)))
870 (beginning-of-line)
871 (not (eq (char-before (1- (point))) ?\\)))
872 (looking-at c-opt-cpp-start)
873 (progn (skip-chars-forward " \t")
874 (eq (point) saved))))
875 (goto-char saved)
876 (if (and (c-forward-to-cpp-define-body)
877 (progn (c-forward-syntactic-ws start)
878 (< (point) start)))
879 ;; Stop at the first token in the content of the macro.
880 (setq pos (point)
881 ignore-labels t) ; Avoid the label check on exit.
882 (setq pos saved
883 ret 'macro
884 ignore-labels t))
885 (throw 'loop nil)) ; 1. Start of macro.
886
887 ;; Do a round through the automaton if we've just passed a
888 ;; statement boundary or passed a "while"-like token.
889 ((or sym
890 (and (looking-at cond-key)
891 (setq sym (intern (match-string 1)))))
892
893 (when (and (< pos start) (null stack))
894 (throw 'loop nil)) ; 2. Statement boundary.
895
896 ;; The PDA state handling.
897 ;;
898 ;; Refer to the description of the PDA in the opening
899 ;; comments. In the following OR form, the first leaf
900 ;; attempts to handles one of the specific actions detailed
901 ;; (e.g., finding token "if" whilst in state `else-boundary').
902 ;; We drop through to the second leaf (which handles common
903 ;; state) if no specific handler is found in the first cond.
904 ;; If a parsing error is detected (e.g. an "else" with no
905 ;; preceding "if"), we throw to the enclosing catch.
906 ;;
907 ;; Note that the (eq state 'else) means
908 ;; "we've just passed an else", NOT "we're looking for an
909 ;; else".
910 (or (cond
911 ((eq state 'else)
912 (if (eq sym 'boundary)
913 (setq state 'else-boundary)
914 (c-bos-report-error)
915 (c-bos-pop-state-and-retry)))
916
917 ((eq state 'else-boundary)
918 (cond ((eq sym 'if)
919 (c-bos-pop-state (setq ret 'beginning)))
920 ((eq sym 'boundary)
921 (c-bos-report-error)
922 (c-bos-pop-state))))
923
924 ((eq state 'while)
925 (if (and (eq sym 'boundary)
926 ;; Since this can cause backtracking we do a
927 ;; little more careful analysis to avoid it:
928 ;; If there's a label in front of the while
929 ;; it can't be part of a do-while.
930 (not after-labels-pos))
931 (progn (c-bos-save-pos)
932 (setq state 'while-boundary))
933 (c-bos-pop-state-and-retry))) ; Can't be a do-while
934
935 ((eq state 'while-boundary)
936 (cond ((eq sym 'do)
937 (c-bos-pop-state (setq ret 'beginning)))
938 ((eq sym 'boundary) ; isn't a do-while
939 (c-bos-restore-pos) ; the position of the while
940 (c-bos-pop-state)))) ; no longer searching for do.
941
942 ((eq state 'catch)
943 (if (eq sym 'boundary)
944 (setq state 'catch-boundary)
945 (c-bos-report-error)
946 (c-bos-pop-state-and-retry)))
947
948 ((eq state 'catch-boundary)
949 (cond
950 ((eq sym 'try)
951 (c-bos-pop-state (setq ret 'beginning)))
952 ((eq sym 'catch)
953 (setq state 'catch))
954 ((eq sym 'boundary)
955 (c-bos-report-error)
956 (c-bos-pop-state)))))
957
958 ;; This is state common. We get here when the previous
959 ;; cond statement found no particular state handler.
960 (cond ((eq sym 'boundary)
961 ;; If we have a boundary at the start
962 ;; position we push a frame to go to the
963 ;; previous statement.
964 (if (>= pos start)
965 (c-bos-push-state)
966 (c-bos-pop-state)))
967 ((eq sym 'else)
968 (c-bos-push-state)
969 (c-bos-save-error-info 'if 'else)
970 (setq state 'else))
971 ((eq sym 'while)
972 ;; Is this a real while, or a do-while?
973 ;; The next `when' triggers unless we are SURE that
974 ;; the `while' is not the tail end of a `do-while'.
975 (when (or (not pptok)
976 (memq (char-after pptok) delims)
977 ;; The following kludge is to prevent
978 ;; infinite recursion when called from
979 ;; c-awk-after-if-for-while-condition-p,
980 ;; or the like.
981 (and (eq (point) start)
982 (c-vsemi-status-unknown-p))
983 (c-at-vsemi-p pptok))
984 ;; Since this can cause backtracking we do a
985 ;; little more careful analysis to avoid it: If
986 ;; the while isn't followed by a (possibly
987 ;; virtual) semicolon it can't be a do-while.
988 (c-bos-push-state)
989 (setq state 'while)))
990 ((memq sym '(catch finally))
991 (c-bos-push-state)
992 (c-bos-save-error-info 'try sym)
993 (setq state 'catch))))
994
995 (when c-maybe-labelp
996 ;; We're either past a statement boundary or at the
997 ;; start of a statement, so throw away any label data
998 ;; for the previous one.
999 (setq after-labels-pos nil
1000 last-label-pos nil
1001 c-maybe-labelp nil))))
1002
1003 ;; Step to the previous sexp, but not if we crossed a
1004 ;; boundary, since that doesn't consume an sexp.
1005 (if (eq sym 'boundary)
1006 (setq ret 'previous)
1007
1008 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
1009 ;; BACKWARDS THROUGH THE SOURCE.
1010
1011 (c-backward-syntactic-ws)
1012 (let ((before-sws-pos (point))
1013 ;; The end position of the area to search for statement
1014 ;; barriers in this round.
1015 (maybe-after-boundary-pos pos))
1016
1017 ;; Go back over exactly one logical sexp, taking proper
1018 ;; account of macros and escaped EOLs.
1019 (while
1020 (progn
1021 (unless (c-safe (c-backward-sexp) t)
1022 ;; Give up if we hit an unbalanced block. Since the
1023 ;; stack won't be empty the code below will report a
1024 ;; suitable error.
1025 (throw 'loop nil))
1026 (cond
1027 ;; Have we moved into a macro?
1028 ((and (not macro-start)
1029 (c-beginning-of-macro))
1030 ;; Have we crossed a statement boundary? If not,
1031 ;; keep going back until we find one or a "real" sexp.
1032 (and
1033 (save-excursion
1034 (c-end-of-macro)
1035 (not (c-crosses-statement-barrier-p
1036 (point) maybe-after-boundary-pos)))
1037 (setq maybe-after-boundary-pos (point))))
1038 ;; Have we just gone back over an escaped NL? This
1039 ;; doesn't count as a sexp.
1040 ((looking-at "\\\\$")))))
1041
1042 ;; Have we crossed a statement boundary?
1043 (setq boundary-pos
1044 (cond
1045 ;; Are we at a macro beginning?
1046 ((and (not macro-start)
1047 c-opt-cpp-prefix
1048 (looking-at c-opt-cpp-prefix))
1049 (save-excursion
1050 (c-end-of-macro)
1051 (c-crosses-statement-barrier-p
1052 (point) maybe-after-boundary-pos)))
1053 ;; Just gone back over a brace block?
1054 ((and
1055 (eq (char-after) ?{)
1056 (not (c-looking-at-inexpr-block lim nil t))
1057 (save-excursion
1058 (c-backward-token-2 1 t nil)
1059 (not (looking-at "=\\([^=]\\|$\\)"))))
1060 (save-excursion
1061 (c-forward-sexp) (point)))
1062 ;; Just gone back over some paren block?
1063 ((looking-at "\\s(")
1064 (save-excursion
1065 (goto-char (1+ (c-down-list-backward
1066 before-sws-pos)))
1067 (c-crosses-statement-barrier-p
1068 (point) maybe-after-boundary-pos)))
1069 ;; Just gone back over an ordinary symbol of some sort?
1070 (t (c-crosses-statement-barrier-p
1071 (point) maybe-after-boundary-pos))))
1072
1073 (when boundary-pos
1074 (setq pptok ptok
1075 ptok tok
1076 tok boundary-pos
1077 sym 'boundary)
1078 ;; Like a C "continue". Analyze the next sexp.
1079 (throw 'loop t))))
1080
1081 ;; ObjC method def?
1082 (when (and c-opt-method-key
1083 (setq saved (c-in-method-def-p)))
1084 (setq pos saved
1085 ignore-labels t) ; Avoid the label check on exit.
1086 (throw 'loop nil)) ; 3. ObjC method def.
1087
1088 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1089 (if c-has-bitfields
1090 (cond
1091 ;; The : <size> and <id> fields?
1092 ((and (numberp c-maybe-labelp)
1093 (not bitfield-size-pos)
1094 (save-excursion
1095 (goto-char (or tok start))
1096 (not (looking-at c-keywords-regexp)))
1097 (not (looking-at c-keywords-regexp))
1098 (not (c-punctuation-in (point) c-maybe-labelp)))
1099 (setq bitfield-size-pos (or tok start)
1100 bitfield-id-pos (point)))
1101 ;; The <type> field?
1102 ((and bitfield-id-pos
1103 (not bitfield-type-pos))
1104 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1105 (not (looking-at c-not-primitive-type-keywords-regexp))
1106 (not (c-punctuation-in (point) tok)))
1107 (setq bitfield-type-pos (point))
1108 (setq bitfield-size-pos nil
1109 bitfield-id-pos nil)))))
1110
1111 ;; Handle labels.
1112 (unless (eq ignore-labels t)
1113 (when (numberp c-maybe-labelp)
1114 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1115 ;; might be in a label now. Have we got a real label
1116 ;; (including a case label) or something like C++'s "public:"?
1117 ;; A case label might use an expression rather than a token.
1118 (setq after-case:-pos (or tok start))
1119 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1120 ;; Catch C++'s inheritance construct "class foo : bar".
1121 (save-excursion
1122 (and
1123 (c-safe (c-backward-sexp) t)
1124 (looking-at c-nonlabel-token-2-key))))
1125 (setq c-maybe-labelp nil)
1126 (if after-labels-pos ; Have we already encountered a label?
1127 (if (not last-label-pos)
1128 (setq last-label-pos (or tok start)))
1129 (setq after-labels-pos (or tok start)))
1130 (setq c-maybe-labelp t
1131 label-good-pos nil))) ; bogus "label"
1132
1133 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1134 ; been found.
1135 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
1136 ;; We're in a potential label and it's the first
1137 ;; time we've found something that isn't allowed in
1138 ;; one.
1139 (setq label-good-pos (or tok start))))
1140
1141 ;; We've moved back by a sexp, so update the token positions.
1142 (setq sym nil
1143 pptok ptok
1144 ptok tok
1145 tok (point)
1146 pos tok) ; always non-nil
1147 ) ; end of (catch loop ....)
1148 ) ; end of sexp-at-a-time (while ....)
1149
1150 ;; If the stack isn't empty there might be errors to report.
1151 (while stack
1152 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1153 (c-bos-report-error))
1154 (setq saved-pos (cdr (car stack))
1155 stack (cdr stack)))
1156
1157 (when (and (eq ret 'same)
1158 (not (memq sym '(boundary ignore nil))))
1159 ;; Need to investigate closer whether we've crossed
1160 ;; between a substatement and its containing statement.
1161 (if (setq saved
1162 (cond ((and (looking-at c-block-stmt-1-2-key)
1163 (eq (char-after ptok) ?\())
1164 pptok)
1165 ((looking-at c-block-stmt-1-key)
1166 ptok)
1167 (t pptok)))
1168 (cond ((> start saved) (setq pos saved))
1169 ((= start saved) (setq ret 'up)))))
1170
1171 (when (and (not ignore-labels)
1172 (eq c-maybe-labelp t)
1173 (not (eq ret 'beginning))
1174 after-labels-pos
1175 (not bitfield-type-pos) ; Bitfields take precedence over labels.
1176 (or (not label-good-pos)
1177 (<= label-good-pos pos)
1178 (progn
1179 (goto-char (if (and last-label-pos
1180 (< last-label-pos start))
1181 last-label-pos
1182 pos))
1183 (looking-at c-label-kwds-regexp))))
1184 ;; We're in a label. Maybe we should step to the statement
1185 ;; after it.
1186 (if (< after-labels-pos start)
1187 (setq pos after-labels-pos)
1188 (setq ret 'label)
1189 (if (and last-label-pos (< last-label-pos start))
1190 ;; Might have jumped over several labels. Go to the last one.
1191 (setq pos last-label-pos)))))
1192
1193 ;; Have we got "case <expression>:"?
1194 (goto-char pos)
1195 (when (and after-case:-pos
1196 (not (eq ret 'beginning))
1197 (looking-at c-case-kwds-regexp))
1198 (if (< after-case:-pos start)
1199 (setq pos after-case:-pos))
1200 (if (eq ret 'same)
1201 (setq ret 'label)))
1202
1203 ;; Skip over the unary operators that can start the statement.
1204 (while (progn
1205 (c-backward-syntactic-ws)
1206 ;; protect AWK post-inc/decrement operators, etc.
1207 (and (not (c-at-vsemi-p (point)))
1208 (/= (skip-chars-backward "-+!*&~@`#") 0)))
1209 (setq pos (point)))
1210 (goto-char pos)
1211 ret)))
1212
1213 (defun c-punctuation-in (from to)
1214 "Return non-nil if there is a non-comment non-macro punctuation character
1215 between FROM and TO. FROM must not be in a string or comment. The returned
1216 value is the position of the first such character."
1217 (save-excursion
1218 (goto-char from)
1219 (let ((pos (point)))
1220 (while (progn (skip-chars-forward c-symbol-chars to)
1221 (c-forward-syntactic-ws to)
1222 (> (point) pos))
1223 (setq pos (point))))
1224 (and (< (point) to) (point))))
1225
1226 (defun c-crosses-statement-barrier-p (from to)
1227 "Return non-nil if buffer positions FROM to TO cross one or more
1228 statement or declaration boundaries. The returned value is actually
1229 the position of the earliest boundary char. FROM must not be within
1230 a string or comment.
1231
1232 The variable `c-maybe-labelp' is set to the position of the first `:' that
1233 might start a label (i.e. not part of `::' and not preceded by `?'). If a
1234 single `?' is found, then `c-maybe-labelp' is cleared.
1235
1236 For AWK, a statement which is terminated by an EOL (not a ; or a }) is
1237 regarded as having a \"virtual semicolon\" immediately after the last token on
1238 the line. If this virtual semicolon is _at_ from, the function recognizes it.
1239
1240 Note that this function might do hidden buffer changes. See the
1241 comment at the start of cc-engine.el for more info."
1242 (let* ((skip-chars
1243 ;; If the current language has CPP macros, insert # into skip-chars.
1244 (if c-opt-cpp-symbol
1245 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1246 c-opt-cpp-symbol ; usually "#"
1247 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1248 c-stmt-delim-chars))
1249 (non-skip-list
1250 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1251 lit-range vsemi-pos)
1252 (save-restriction
1253 (widen)
1254 (save-excursion
1255 (catch 'done
1256 (goto-char from)
1257 (while (progn (skip-chars-forward
1258 skip-chars
1259 (min to (c-point 'bonl)))
1260 (< (point) to))
1261 (cond
1262 ;; Virtual semicolon?
1263 ((and (bolp)
1264 (save-excursion
1265 (progn
1266 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1267 (goto-char (car lit-range)))
1268 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1269 (setq vsemi-pos (point))
1270 (c-at-vsemi-p))))
1271 (throw 'done vsemi-pos))
1272 ;; In a string/comment?
1273 ((setq lit-range (c-literal-limits from))
1274 (goto-char (cdr lit-range)))
1275 ((eq (char-after) ?:)
1276 (forward-char)
1277 (if (and (eq (char-after) ?:)
1278 (< (point) to))
1279 ;; Ignore scope operators.
1280 (forward-char)
1281 (setq c-maybe-labelp (1- (point)))))
1282 ((eq (char-after) ??)
1283 ;; A question mark. Can't be a label, so stop
1284 ;; looking for more : and ?.
1285 (setq c-maybe-labelp nil
1286 skip-chars (substring c-stmt-delim-chars 0 -2)))
1287 ;; At a CPP construct or a "#" or "##" operator?
1288 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol))
1289 (if (save-excursion
1290 (skip-chars-backward " \t")
1291 (and (bolp)
1292 (or (bobp)
1293 (not (eq (char-before (1- (point))) ?\\)))))
1294 (c-end-of-macro)
1295 (skip-chars-forward c-opt-cpp-symbol)))
1296 ((memq (char-after) non-skip-list)
1297 (throw 'done (point)))))
1298 ;; In trailing space after an as yet undetected virtual semicolon?
1299 (c-backward-syntactic-ws from)
1300 (when (and (bolp) (not (bobp))) ; Can happen in AWK Mode with an
1301 ; unterminated string/regexp.
1302 (backward-char))
1303 (if (and (< (point) to)
1304 (c-at-vsemi-p))
1305 (point)
1306 nil))))))
1307
1308 (defun c-at-statement-start-p ()
1309 "Return non-nil if the point is at the first token in a statement
1310 or somewhere in the syntactic whitespace before it.
1311
1312 A \"statement\" here is not restricted to those inside code blocks.
1313 Any kind of declaration-like construct that occur outside function
1314 bodies is also considered a \"statement\".
1315
1316 Note that this function might do hidden buffer changes. See the
1317 comment at the start of cc-engine.el for more info."
1318
1319 (save-excursion
1320 (let ((end (point))
1321 c-maybe-labelp)
1322 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1323 (or (bobp)
1324 (eq (char-before) ?})
1325 (and (eq (char-before) ?{)
1326 (not (and c-special-brace-lists
1327 (progn (backward-char)
1328 (c-looking-at-special-brace-list)))))
1329 (c-crosses-statement-barrier-p (point) end)))))
1330
1331 (defun c-at-expression-start-p ()
1332 "Return non-nil if the point is at the first token in an expression or
1333 statement, or somewhere in the syntactic whitespace before it.
1334
1335 An \"expression\" here is a bit different from the normal language
1336 grammar sense: It's any sequence of expression tokens except commas,
1337 unless they are enclosed inside parentheses of some kind. Also, an
1338 expression never continues past an enclosing parenthesis, but it might
1339 contain parenthesis pairs of any sort except braces.
1340
1341 Since expressions never cross statement boundaries, this function also
1342 recognizes statement beginnings, just like `c-at-statement-start-p'.
1343
1344 Note that this function might do hidden buffer changes. See the
1345 comment at the start of cc-engine.el for more info."
1346
1347 (save-excursion
1348 (let ((end (point))
1349 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1350 c-maybe-labelp)
1351 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1352 (or (bobp)
1353 (memq (char-before) '(?{ ?}))
1354 (save-excursion (backward-char)
1355 (looking-at "\\s("))
1356 (c-crosses-statement-barrier-p (point) end)))))
1357
1358 \f
1359 ;; A set of functions that covers various idiosyncrasies in
1360 ;; implementations of `forward-comment'.
1361
1362 ;; Note: Some emacsen considers incorrectly that any line comment
1363 ;; ending with a backslash continues to the next line. I can't think
1364 ;; of any way to work around that in a reliable way without changing
1365 ;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1366 ;; changing the syntax for backslash doesn't work since we must treat
1367 ;; escapes in string literals correctly.)
1368
1369 (defun c-forward-single-comment ()
1370 "Move forward past whitespace and the closest following comment, if any.
1371 Return t if a comment was found, nil otherwise. In either case, the
1372 point is moved past the following whitespace. Line continuations,
1373 i.e. a backslashes followed by line breaks, are treated as whitespace.
1374 The line breaks that end line comments are considered to be the
1375 comment enders, so the point will be put on the beginning of the next
1376 line if it moved past a line comment.
1377
1378 This function does not do any hidden buffer changes."
1379
1380 (let ((start (point)))
1381 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1382 (goto-char (match-end 0)))
1383
1384 (when (forward-comment 1)
1385 (if (eobp)
1386 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1387 ;; forwards at eob.
1388 nil
1389
1390 ;; Emacs includes the ending newline in a b-style (c++)
1391 ;; comment, but XEmacs doesn't. We depend on the Emacs
1392 ;; behavior (which also is symmetric).
1393 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1394 (condition-case nil (forward-char 1)))
1395
1396 t))))
1397
1398 (defsubst c-forward-comments ()
1399 "Move forward past all following whitespace and comments.
1400 Line continuations, i.e. a backslashes followed by line breaks, are
1401 treated as whitespace.
1402
1403 Note that this function might do hidden buffer changes. See the
1404 comment at the start of cc-engine.el for more info."
1405
1406 (while (or
1407 ;; If forward-comment in at least XEmacs 21 is given a large
1408 ;; positive value, it'll loop all the way through if it hits
1409 ;; eob.
1410 (and (forward-comment 5)
1411 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1412 ;; forwards at eob.
1413 (not (eobp)))
1414
1415 (when (looking-at "\\\\[\n\r]")
1416 (forward-char 2)
1417 t))))
1418
1419 (defun c-backward-single-comment ()
1420 "Move backward past whitespace and the closest preceding comment, if any.
1421 Return t if a comment was found, nil otherwise. In either case, the
1422 point is moved past the preceding whitespace. Line continuations,
1423 i.e. a backslashes followed by line breaks, are treated as whitespace.
1424 The line breaks that end line comments are considered to be the
1425 comment enders, so the point cannot be at the end of the same line to
1426 move over a line comment.
1427
1428 This function does not do any hidden buffer changes."
1429
1430 (let ((start (point)))
1431 ;; When we got newline terminated comments, forward-comment in all
1432 ;; supported emacsen so far will stop at eol of each line not
1433 ;; ending with a comment when moving backwards. This corrects for
1434 ;; that, and at the same time handles line continuations.
1435 (while (progn
1436 (skip-chars-backward " \t\n\r\f\v")
1437 (and (looking-at "[\n\r]")
1438 (eq (char-before) ?\\)))
1439 (backward-char))
1440
1441 (if (bobp)
1442 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1443 ;; backwards at bob.
1444 nil
1445
1446 ;; Leave point after the closest following newline if we've
1447 ;; backed up over any above, since forward-comment won't move
1448 ;; backward over a line comment if point is at the end of the
1449 ;; same line.
1450 (re-search-forward "\\=\\s *[\n\r]" start t)
1451
1452 (if (if (forward-comment -1)
1453 (if (eolp)
1454 ;; If forward-comment above succeeded and we're at eol
1455 ;; then the newline we moved over above didn't end a
1456 ;; line comment, so we give it another go.
1457 (forward-comment -1)
1458 t))
1459
1460 ;; Emacs <= 20 and XEmacs move back over the closer of a
1461 ;; block comment that lacks an opener.
1462 (if (looking-at "\\*/")
1463 (progn (forward-char 2) nil)
1464 t)))))
1465
1466 (defsubst c-backward-comments ()
1467 "Move backward past all preceding whitespace and comments.
1468 Line continuations, i.e. a backslashes followed by line breaks, are
1469 treated as whitespace. The line breaks that end line comments are
1470 considered to be the comment enders, so the point cannot be at the end
1471 of the same line to move over a line comment. Unlike
1472 c-backward-syntactic-ws, this function doesn't move back over
1473 preprocessor directives.
1474
1475 Note that this function might do hidden buffer changes. See the
1476 comment at the start of cc-engine.el for more info."
1477
1478 (let ((start (point)))
1479 (while (and
1480 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
1481 ;; return t when moving backwards at bob.
1482 (not (bobp))
1483
1484 (if (let (moved-comment)
1485 (while
1486 (and (not (setq moved-comment (forward-comment -1)))
1487 ;; Cope specifically with ^M^J here -
1488 ;; forward-comment sometimes gets stuck after ^Ms,
1489 ;; sometimes after ^M^J.
1490 (or
1491 (when (eq (char-before) ?\r)
1492 (backward-char)
1493 t)
1494 (when (and (eq (char-before) ?\n)
1495 (eq (char-before (1- (point))) ?\r))
1496 (backward-char 2)
1497 t))))
1498 moved-comment)
1499 (if (looking-at "\\*/")
1500 ;; Emacs <= 20 and XEmacs move back over the
1501 ;; closer of a block comment that lacks an opener.
1502 (progn (forward-char 2) nil)
1503 t)
1504
1505 ;; XEmacs treats line continuations as whitespace but
1506 ;; only in the backward direction, which seems a bit
1507 ;; odd. Anyway, this is necessary for Emacs.
1508 (when (and (looking-at "[\n\r]")
1509 (eq (char-before) ?\\)
1510 (< (point) start))
1511 (backward-char)
1512 t))))))
1513
1514 \f
1515 ;; Tools for skipping over syntactic whitespace.
1516
1517 ;; The following functions use text properties to cache searches over
1518 ;; large regions of syntactic whitespace. It works as follows:
1519 ;;
1520 ;; o If a syntactic whitespace region contains anything but simple
1521 ;; whitespace (i.e. space, tab and line breaks), the text property
1522 ;; `c-in-sws' is put over it. At places where we have stopped
1523 ;; within that region there's also a `c-is-sws' text property.
1524 ;; That since there typically are nested whitespace inside that
1525 ;; must be handled separately, e.g. whitespace inside a comment or
1526 ;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1527 ;; to jump to another point with that property within the same
1528 ;; `c-in-sws' region. It can be likened to a ladder where
1529 ;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1530 ;;
1531 ;; o The `c-is-sws' property is put on the simple whitespace chars at
1532 ;; a "rung position" and also maybe on the first following char.
1533 ;; As many characters as can be conveniently found in this range
1534 ;; are marked, but no assumption can be made that the whole range
1535 ;; is marked (it could be clobbered by later changes, for
1536 ;; instance).
1537 ;;
1538 ;; Note that some part of the beginning of a sequence of simple
1539 ;; whitespace might be part of the end of a preceding line comment
1540 ;; or cpp directive and must not be considered part of the "rung".
1541 ;; Such whitespace is some amount of horizontal whitespace followed
1542 ;; by a newline. In the case of cpp directives it could also be
1543 ;; two newlines with horizontal whitespace between them.
1544 ;;
1545 ;; The reason to include the first following char is to cope with
1546 ;; "rung positions" that doesn't have any ordinary whitespace. If
1547 ;; `c-is-sws' is put on a token character it does not have
1548 ;; `c-in-sws' set simultaneously. That's the only case when that
1549 ;; can occur, and the reason for not extending the `c-in-sws'
1550 ;; region to cover it is that the `c-in-sws' region could then be
1551 ;; accidentally merged with a following one if the token is only
1552 ;; one character long.
1553 ;;
1554 ;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1555 ;; removed in the changed region. If the change was inside
1556 ;; syntactic whitespace that means that the "ladder" is broken, but
1557 ;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1558 ;; parts on either side and use an ordinary search only to "repair"
1559 ;; the gap.
1560 ;;
1561 ;; Special care needs to be taken if a region is removed: If there
1562 ;; are `c-in-sws' on both sides of it which do not connect inside
1563 ;; the region then they can't be joined. If e.g. a marked macro is
1564 ;; broken, syntactic whitespace inside the new text might be
1565 ;; marked. If those marks would become connected with the old
1566 ;; `c-in-sws' range around the macro then we could get a ladder
1567 ;; with one end outside the macro and the other at some whitespace
1568 ;; within it.
1569 ;;
1570 ;; The main motivation for this system is to increase the speed in
1571 ;; skipping over the large whitespace regions that can occur at the
1572 ;; top level in e.g. header files that contain a lot of comments and
1573 ;; cpp directives. For small comments inside code it's probably
1574 ;; slower than using `forward-comment' straightforwardly, but speed is
1575 ;; not a significant factor there anyway.
1576
1577 ; (defface c-debug-is-sws-face
1578 ; '((t (:background "GreenYellow")))
1579 ; "Debug face to mark the `c-is-sws' property.")
1580 ; (defface c-debug-in-sws-face
1581 ; '((t (:underline t)))
1582 ; "Debug face to mark the `c-in-sws' property.")
1583
1584 ; (defun c-debug-put-sws-faces ()
1585 ; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1586 ; ;; properties in the buffer.
1587 ; (interactive)
1588 ; (save-excursion
1589 ; (c-save-buffer-state (in-face)
1590 ; (goto-char (point-min))
1591 ; (setq in-face (if (get-text-property (point) 'c-is-sws)
1592 ; (point)))
1593 ; (while (progn
1594 ; (goto-char (next-single-property-change
1595 ; (point) 'c-is-sws nil (point-max)))
1596 ; (if in-face
1597 ; (progn
1598 ; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1599 ; (setq in-face nil))
1600 ; (setq in-face (point)))
1601 ; (not (eobp))))
1602 ; (goto-char (point-min))
1603 ; (setq in-face (if (get-text-property (point) 'c-in-sws)
1604 ; (point)))
1605 ; (while (progn
1606 ; (goto-char (next-single-property-change
1607 ; (point) 'c-in-sws nil (point-max)))
1608 ; (if in-face
1609 ; (progn
1610 ; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1611 ; (setq in-face nil))
1612 ; (setq in-face (point)))
1613 ; (not (eobp)))))))
1614
1615 (defmacro c-debug-sws-msg (&rest args)
1616 ;;`(message ,@args)
1617 )
1618
1619 (defmacro c-put-is-sws (beg end)
1620 ;; This macro does a hidden buffer change.
1621 `(let ((beg ,beg) (end ,end))
1622 (put-text-property beg end 'c-is-sws t)
1623 ,@(when (facep 'c-debug-is-sws-face)
1624 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1625
1626 (defmacro c-put-in-sws (beg end)
1627 ;; This macro does a hidden buffer change.
1628 `(let ((beg ,beg) (end ,end))
1629 (put-text-property beg end 'c-in-sws t)
1630 ,@(when (facep 'c-debug-is-sws-face)
1631 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1632
1633 (defmacro c-remove-is-sws (beg end)
1634 ;; This macro does a hidden buffer change.
1635 `(let ((beg ,beg) (end ,end))
1636 (remove-text-properties beg end '(c-is-sws nil))
1637 ,@(when (facep 'c-debug-is-sws-face)
1638 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1639
1640 (defmacro c-remove-in-sws (beg end)
1641 ;; This macro does a hidden buffer change.
1642 `(let ((beg ,beg) (end ,end))
1643 (remove-text-properties beg end '(c-in-sws nil))
1644 ,@(when (facep 'c-debug-is-sws-face)
1645 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1646
1647 (defmacro c-remove-is-and-in-sws (beg end)
1648 ;; This macro does a hidden buffer change.
1649 `(let ((beg ,beg) (end ,end))
1650 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1651 ,@(when (facep 'c-debug-is-sws-face)
1652 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1653 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1654
1655 (defsubst c-invalidate-sws-region-after (beg end)
1656 ;; Called from `after-change-functions'. Note that if
1657 ;; `c-forward-sws' or `c-backward-sws' are used outside
1658 ;; `c-save-buffer-state' or similar then this will remove the cache
1659 ;; properties right after they're added.
1660 ;;
1661 ;; This function does hidden buffer changes.
1662
1663 (save-excursion
1664 ;; Adjust the end to remove the properties in any following simple
1665 ;; ws up to and including the next line break, if there is any
1666 ;; after the changed region. This is necessary e.g. when a rung
1667 ;; marked empty line is converted to a line comment by inserting
1668 ;; "//" before the line break. In that case the line break would
1669 ;; keep the rung mark which could make a later `c-backward-sws'
1670 ;; move into the line comment instead of over it.
1671 (goto-char end)
1672 (skip-chars-forward " \t\f\v")
1673 (when (and (eolp) (not (eobp)))
1674 (setq end (1+ (point)))))
1675
1676 (when (and (= beg end)
1677 (get-text-property beg 'c-in-sws)
1678 (> beg (point-min))
1679 (get-text-property (1- beg) 'c-in-sws))
1680 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1681 ;; safe to keep a range that was continuous before the change. E.g:
1682 ;;
1683 ;; #define foo
1684 ;; \
1685 ;; bar
1686 ;;
1687 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1688 ;; after "foo" is removed then "bar" will become part of the cpp
1689 ;; directive instead of a syntactically relevant token. In that
1690 ;; case there's no longer syntactic ws from "#" to "b".
1691 (setq beg (1- beg)))
1692
1693 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1694 (c-remove-is-and-in-sws beg end))
1695
1696 (defun c-forward-sws ()
1697 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
1698 ;;
1699 ;; This function might do hidden buffer changes.
1700
1701 (let (;; `rung-pos' is set to a position as early as possible in the
1702 ;; unmarked part of the simple ws region.
1703 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1704 rung-is-marked next-rung-is-marked simple-ws-end
1705 ;; `safe-start' is set when it's safe to cache the start position.
1706 ;; It's not set if we've initially skipped over comments and line
1707 ;; continuations since we might have gone out through the end of a
1708 ;; macro then. This provision makes `c-forward-sws' not populate the
1709 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1710 ;; more common.
1711 safe-start)
1712
1713 ;; Skip simple ws and do a quick check on the following character to see
1714 ;; if it's anything that can't start syntactic ws, so we can bail out
1715 ;; early in the majority of cases when there just are a few ws chars.
1716 (skip-chars-forward " \t\n\r\f\v")
1717 (when (looking-at c-syntactic-ws-start)
1718
1719 (setq rung-end-pos (min (1+ (point)) (point-max)))
1720 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1721 'c-is-sws t))
1722 ;; Find the last rung position to avoid setting properties in all
1723 ;; the cases when the marked rung is complete.
1724 ;; (`next-single-property-change' is certain to move at least one
1725 ;; step forward.)
1726 (setq rung-pos (1- (c-next-single-property-change
1727 rung-is-marked 'c-is-sws nil rung-end-pos)))
1728 ;; Got no marked rung here. Since the simple ws might have started
1729 ;; inside a line comment or cpp directive we must set `rung-pos' as
1730 ;; high as possible.
1731 (setq rung-pos (point)))
1732
1733 (with-silent-modifications
1734 (while
1735 (progn
1736 (while
1737 (when (and rung-is-marked
1738 (get-text-property (point) 'c-in-sws))
1739
1740 ;; The following search is the main reason that `c-in-sws'
1741 ;; and `c-is-sws' aren't combined to one property.
1742 (goto-char (c-next-single-property-change
1743 (point) 'c-in-sws nil (point-max)))
1744 (unless (get-text-property (point) 'c-is-sws)
1745 ;; If the `c-in-sws' region extended past the last
1746 ;; `c-is-sws' char we have to go back a bit.
1747 (or (get-text-property (1- (point)) 'c-is-sws)
1748 (goto-char (previous-single-property-change
1749 (point) 'c-is-sws)))
1750 (backward-char))
1751
1752 (c-debug-sws-msg
1753 "c-forward-sws cached move %s -> %s (max %s)"
1754 rung-pos (point) (point-max))
1755
1756 (setq rung-pos (point))
1757 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1758 (not (eobp))))
1759
1760 ;; We'll loop here if there is simple ws after the last rung.
1761 ;; That means that there's been some change in it and it's
1762 ;; possible that we've stepped into another ladder, so extend
1763 ;; the previous one to join with it if there is one, and try to
1764 ;; use the cache again.
1765 (c-debug-sws-msg
1766 "c-forward-sws extending rung with [%s..%s] (max %s)"
1767 (1+ rung-pos) (1+ (point)) (point-max))
1768 (unless (get-text-property (point) 'c-is-sws)
1769 ;; Remove any `c-in-sws' property from the last char of
1770 ;; the rung before we mark it with `c-is-sws', so that we
1771 ;; won't connect with the remains of a broken "ladder".
1772 (c-remove-in-sws (point) (1+ (point))))
1773 (c-put-is-sws (1+ rung-pos)
1774 (1+ (point)))
1775 (c-put-in-sws rung-pos
1776 (setq rung-pos (point)
1777 last-put-in-sws-pos rung-pos)))
1778
1779 (setq simple-ws-end (point))
1780 (c-forward-comments)
1781
1782 (cond
1783 ((/= (point) simple-ws-end)
1784 ;; Skipped over comments. Don't cache at eob in case the buffer
1785 ;; is narrowed.
1786 (not (eobp)))
1787
1788 ((save-excursion
1789 (and c-opt-cpp-prefix
1790 (looking-at c-opt-cpp-start)
1791 (progn (skip-chars-backward " \t")
1792 (bolp))
1793 (or (bobp)
1794 (progn (backward-char)
1795 (not (eq (char-before) ?\\))))))
1796 ;; Skip a preprocessor directive.
1797 (end-of-line)
1798 (while (and (eq (char-before) ?\\)
1799 (= (forward-line 1) 0))
1800 (end-of-line))
1801 (forward-line 1)
1802 (setq safe-start t)
1803 ;; Don't cache at eob in case the buffer is narrowed.
1804 (not (eobp)))))
1805
1806 ;; We've searched over a piece of non-white syntactic ws. See if this
1807 ;; can be cached.
1808 (setq next-rung-pos (point))
1809 (skip-chars-forward " \t\n\r\f\v")
1810 (setq rung-end-pos (min (1+ (point)) (point-max)))
1811
1812 (if (or
1813 ;; Cache if we haven't skipped comments only, and if we started
1814 ;; either from a marked rung or from a completely uncached
1815 ;; position.
1816 (and safe-start
1817 (or rung-is-marked
1818 (not (get-text-property simple-ws-end 'c-in-sws))))
1819
1820 ;; See if there's a marked rung in the encountered simple ws. If
1821 ;; so then we can cache, unless `safe-start' is nil. Even then
1822 ;; we need to do this to check if the cache can be used for the
1823 ;; next step.
1824 (and (setq next-rung-is-marked
1825 (text-property-any next-rung-pos rung-end-pos
1826 'c-is-sws t))
1827 safe-start))
1828
1829 (progn
1830 (c-debug-sws-msg
1831 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1832 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1833 (point-max))
1834
1835 ;; Remove the properties for any nested ws that might be cached.
1836 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1837 ;; anyway.
1838 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1839 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1840 (c-put-is-sws rung-pos
1841 (1+ simple-ws-end))
1842 (setq rung-is-marked t))
1843 (c-put-in-sws rung-pos
1844 (setq rung-pos (point)
1845 last-put-in-sws-pos rung-pos))
1846 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1847 ;; Remove any `c-in-sws' property from the last char of
1848 ;; the rung before we mark it with `c-is-sws', so that we
1849 ;; won't connect with the remains of a broken "ladder".
1850 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1851 (c-put-is-sws next-rung-pos
1852 rung-end-pos))
1853
1854 (c-debug-sws-msg
1855 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1856 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1857 (point-max))
1858
1859 ;; Set `rung-pos' for the next rung. It's the same thing here as
1860 ;; initially, except that the rung position is set as early as
1861 ;; possible since we can't be in the ending ws of a line comment or
1862 ;; cpp directive now.
1863 (if (setq rung-is-marked next-rung-is-marked)
1864 (setq rung-pos (1- (c-next-single-property-change
1865 rung-is-marked 'c-is-sws nil rung-end-pos)))
1866 (setq rung-pos next-rung-pos))
1867 (setq safe-start t)))
1868
1869 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1870 ;; another one after the point (which might occur when editing inside a
1871 ;; comment or macro).
1872 (when (eq last-put-in-sws-pos (point))
1873 (cond ((< last-put-in-sws-pos (point-max))
1874 (c-debug-sws-msg
1875 "c-forward-sws clearing at %s for cache separation"
1876 last-put-in-sws-pos)
1877 (c-remove-in-sws last-put-in-sws-pos
1878 (1+ last-put-in-sws-pos)))
1879 (t
1880 ;; If at eob we have to clear the last character before the end
1881 ;; instead since the buffer might be narrowed and there might
1882 ;; be a `c-in-sws' after (point-max). In this case it's
1883 ;; necessary to clear both properties.
1884 (c-debug-sws-msg
1885 "c-forward-sws clearing thoroughly at %s for cache separation"
1886 (1- last-put-in-sws-pos))
1887 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1888 last-put-in-sws-pos))))
1889 ))))
1890
1891 (defun c-backward-sws ()
1892 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
1893 ;;
1894 ;; This function might do hidden buffer changes.
1895
1896 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1897 ;; part of the simple ws region.
1898 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1899 rung-is-marked simple-ws-beg cmt-skip-pos)
1900
1901 ;; Skip simple horizontal ws and do a quick check on the preceding
1902 ;; character to see if it's anything that can't end syntactic ws, so we can
1903 ;; bail out early in the majority of cases when there just are a few ws
1904 ;; chars. Newlines are complicated in the backward direction, so we can't
1905 ;; skip over them.
1906 (skip-chars-backward " \t\f")
1907 (when (and (not (bobp))
1908 (save-excursion
1909 (backward-char)
1910 (looking-at c-syntactic-ws-end)))
1911
1912 ;; Try to find a rung position in the simple ws preceding point, so that
1913 ;; we can get a cache hit even if the last bit of the simple ws has
1914 ;; changed recently.
1915 (setq simple-ws-beg (point))
1916 (skip-chars-backward " \t\n\r\f\v")
1917 (if (setq rung-is-marked (text-property-any
1918 (point) (min (1+ rung-pos) (point-max))
1919 'c-is-sws t))
1920 ;; `rung-pos' will be the earliest marked position, which means that
1921 ;; there might be later unmarked parts in the simple ws region.
1922 ;; It's not worth the effort to fix that; the last part of the
1923 ;; simple ws is also typically edited often, so it could be wasted.
1924 (goto-char (setq rung-pos rung-is-marked))
1925 (goto-char simple-ws-beg))
1926
1927 (with-silent-modifications
1928 (while
1929 (progn
1930 (while
1931 (when (and rung-is-marked
1932 (not (bobp))
1933 (get-text-property (1- (point)) 'c-in-sws))
1934
1935 ;; The following search is the main reason that `c-in-sws'
1936 ;; and `c-is-sws' aren't combined to one property.
1937 (goto-char (previous-single-property-change
1938 (point) 'c-in-sws nil (point-min)))
1939 (unless (get-text-property (point) 'c-is-sws)
1940 ;; If the `c-in-sws' region extended past the first
1941 ;; `c-is-sws' char we have to go forward a bit.
1942 (goto-char (c-next-single-property-change
1943 (point) 'c-is-sws)))
1944
1945 (c-debug-sws-msg
1946 "c-backward-sws cached move %s <- %s (min %s)"
1947 (point) rung-pos (point-min))
1948
1949 (setq rung-pos (point))
1950 (if (and (< (min (skip-chars-backward " \t\f\v")
1951 (progn
1952 (setq simple-ws-beg (point))
1953 (skip-chars-backward " \t\n\r\f\v")))
1954 0)
1955 (setq rung-is-marked
1956 (text-property-any (point) rung-pos
1957 'c-is-sws t)))
1958 t
1959 (goto-char simple-ws-beg)
1960 nil))
1961
1962 ;; We'll loop here if there is simple ws before the first rung.
1963 ;; That means that there's been some change in it and it's
1964 ;; possible that we've stepped into another ladder, so extend
1965 ;; the previous one to join with it if there is one, and try to
1966 ;; use the cache again.
1967 (c-debug-sws-msg
1968 "c-backward-sws extending rung with [%s..%s] (min %s)"
1969 rung-is-marked rung-pos (point-min))
1970 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1971 ;; Remove any `c-in-sws' property from the last char of
1972 ;; the rung before we mark it with `c-is-sws', so that we
1973 ;; won't connect with the remains of a broken "ladder".
1974 (c-remove-in-sws (1- rung-pos) rung-pos))
1975 (c-put-is-sws rung-is-marked
1976 rung-pos)
1977 (c-put-in-sws rung-is-marked
1978 (1- rung-pos))
1979 (setq rung-pos rung-is-marked
1980 last-put-in-sws-pos rung-pos))
1981
1982 (c-backward-comments)
1983 (setq cmt-skip-pos (point))
1984
1985 (cond
1986 ((and c-opt-cpp-prefix
1987 (/= cmt-skip-pos simple-ws-beg)
1988 (c-beginning-of-macro))
1989 ;; Inside a cpp directive. See if it should be skipped over.
1990 (let ((cpp-beg (point)))
1991
1992 ;; Move back over all line continuations in the region skipped
1993 ;; over by `c-backward-comments'. If we go past it then we
1994 ;; started inside the cpp directive.
1995 (goto-char simple-ws-beg)
1996 (beginning-of-line)
1997 (while (and (> (point) cmt-skip-pos)
1998 (progn (backward-char)
1999 (eq (char-before) ?\\)))
2000 (beginning-of-line))
2001
2002 (if (< (point) cmt-skip-pos)
2003 ;; Don't move past the cpp directive if we began inside
2004 ;; it. Note that the position at the end of the last line
2005 ;; of the macro is also considered to be within it.
2006 (progn (goto-char cmt-skip-pos)
2007 nil)
2008
2009 ;; It's worthwhile to spend a little bit of effort on finding
2010 ;; the end of the macro, to get a good `simple-ws-beg'
2011 ;; position for the cache. Note that `c-backward-comments'
2012 ;; could have stepped over some comments before going into
2013 ;; the macro, and then `simple-ws-beg' must be kept on the
2014 ;; same side of those comments.
2015 (goto-char simple-ws-beg)
2016 (skip-chars-backward " \t\n\r\f\v")
2017 (if (eq (char-before) ?\\)
2018 (forward-char))
2019 (forward-line 1)
2020 (if (< (point) simple-ws-beg)
2021 ;; Might happen if comments after the macro were skipped
2022 ;; over.
2023 (setq simple-ws-beg (point)))
2024
2025 (goto-char cpp-beg)
2026 t)))
2027
2028 ((/= (save-excursion
2029 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2030 (setq next-rung-pos (point)))
2031 simple-ws-beg)
2032 ;; Skipped over comments. Must put point at the end of
2033 ;; the simple ws at point since we might be after a line
2034 ;; comment or cpp directive that's been partially
2035 ;; narrowed out, and we can't risk marking the simple ws
2036 ;; at the end of it.
2037 (goto-char next-rung-pos)
2038 t)))
2039
2040 ;; We've searched over a piece of non-white syntactic ws. See if this
2041 ;; can be cached.
2042 (setq next-rung-pos (point))
2043 (skip-chars-backward " \t\f\v")
2044
2045 (if (or
2046 ;; Cache if we started either from a marked rung or from a
2047 ;; completely uncached position.
2048 rung-is-marked
2049 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2050
2051 ;; Cache if there's a marked rung in the encountered simple ws.
2052 (save-excursion
2053 (skip-chars-backward " \t\n\r\f\v")
2054 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2055 'c-is-sws t)))
2056
2057 (progn
2058 (c-debug-sws-msg
2059 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2060 (point) (1+ next-rung-pos)
2061 simple-ws-beg (min (1+ rung-pos) (point-max))
2062 (point-min))
2063
2064 ;; Remove the properties for any nested ws that might be cached.
2065 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2066 ;; anyway.
2067 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2068 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2069 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2070 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2071 ;; Remove any `c-in-sws' property from the last char of
2072 ;; the rung before we mark it with `c-is-sws', so that we
2073 ;; won't connect with the remains of a broken "ladder".
2074 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2075 (c-put-is-sws simple-ws-beg
2076 rung-end-pos)
2077 (setq rung-is-marked t)))
2078 (c-put-in-sws (setq simple-ws-beg (point)
2079 last-put-in-sws-pos simple-ws-beg)
2080 rung-pos)
2081 (c-put-is-sws (setq rung-pos simple-ws-beg)
2082 (1+ next-rung-pos)))
2083
2084 (c-debug-sws-msg
2085 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2086 (point) (1+ next-rung-pos)
2087 simple-ws-beg (min (1+ rung-pos) (point-max))
2088 (point-min))
2089 (setq rung-pos next-rung-pos
2090 simple-ws-beg (point))
2091 ))
2092
2093 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2094 ;; another one before the point (which might occur when editing inside a
2095 ;; comment or macro).
2096 (when (eq last-put-in-sws-pos (point))
2097 (cond ((< (point-min) last-put-in-sws-pos)
2098 (c-debug-sws-msg
2099 "c-backward-sws clearing at %s for cache separation"
2100 (1- last-put-in-sws-pos))
2101 (c-remove-in-sws (1- last-put-in-sws-pos)
2102 last-put-in-sws-pos))
2103 ((> (point-min) 1)
2104 ;; If at bob and the buffer is narrowed, we have to clear the
2105 ;; character we're standing on instead since there might be a
2106 ;; `c-in-sws' before (point-min). In this case it's necessary
2107 ;; to clear both properties.
2108 (c-debug-sws-msg
2109 "c-backward-sws clearing thoroughly at %s for cache separation"
2110 last-put-in-sws-pos)
2111 (c-remove-is-and-in-sws last-put-in-sws-pos
2112 (1+ last-put-in-sws-pos)))))
2113 ))))
2114
2115 \f
2116 ;; Other whitespace tools
2117 (defun c-partial-ws-p (beg end)
2118 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2119 ;; region? This is a "heuristic" function. .....
2120 ;;
2121 ;; The motivation for the second bit is to check whether removing this
2122 ;; region would coalesce two symbols.
2123 ;;
2124 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2125 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2126 (save-excursion
2127 (let ((end+1 (min (1+ end) (point-max))))
2128 (or (progn (goto-char (max (point-min) (1- beg)))
2129 (c-skip-ws-forward end)
2130 (eq (point) end))
2131 (progn (goto-char beg)
2132 (c-skip-ws-forward end+1)
2133 (eq (point) end+1))))))
2134 \f
2135 ;; A system for finding noteworthy parens before the point.
2136
2137 (defconst c-state-cache-too-far 5000)
2138 ;; A maximum comfortable scanning distance, e.g. between
2139 ;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
2140 ;; this distance is exceeded, we take "emergency measures", e.g. by clearing
2141 ;; the cache and starting again from point-min or a beginning of defun. This
2142 ;; value can be tuned for efficiency or set to a lower value for testing.
2143
2144 (defvar c-state-cache nil)
2145 (make-variable-buffer-local 'c-state-cache)
2146 ;; The state cache used by `c-parse-state' to cut down the amount of
2147 ;; searching. It's the result from some earlier `c-parse-state' call. See
2148 ;; `c-parse-state''s doc string for details of its structure.
2149 ;;
2150 ;; The use of the cached info is more effective if the next
2151 ;; `c-parse-state' call is on a line close by the one the cached state
2152 ;; was made at; the cache can actually slow down a little if the
2153 ;; cached state was made very far back in the buffer. The cache is
2154 ;; most effective if `c-parse-state' is used on each line while moving
2155 ;; forward.
2156
2157 (defvar c-state-cache-good-pos 1)
2158 (make-variable-buffer-local 'c-state-cache-good-pos)
2159 ;; This is a position where `c-state-cache' is known to be correct, or
2160 ;; nil (see below). It's a position inside one of the recorded unclosed
2161 ;; parens or the top level, but not further nested inside any literal or
2162 ;; subparen that is closed before the last recorded position.
2163 ;;
2164 ;; The exact position is chosen to try to be close to yet earlier than
2165 ;; the position where `c-state-cache' will be called next. Right now
2166 ;; the heuristic is to set it to the position after the last found
2167 ;; closing paren (of any type) before the line on which
2168 ;; `c-parse-state' was called. That is chosen primarily to work well
2169 ;; with refontification of the current line.
2170 ;;
2171 ;; 2009-07-28: When `c-state-point-min' and the last position where
2172 ;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2173 ;; both in the same literal, there is no such "good position", and
2174 ;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2175 ;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2176 ;;
2177 ;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2178 ;; the middle of the desert, as long as it is not within a brace pair
2179 ;; recorded in `c-state-cache' or a paren/bracket pair.
2180
2181 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2182 ;; We maintain a simple cache of positions which aren't in a literal, so as to
2183 ;; speed up testing for non-literality.
2184 (defconst c-state-nonlit-pos-interval 3000)
2185 ;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2186
2187 (defvar c-state-nonlit-pos-cache nil)
2188 (make-variable-buffer-local 'c-state-nonlit-pos-cache)
2189 ;; A list of buffer positions which are known not to be in a literal or a cpp
2190 ;; construct. This is ordered with higher positions at the front of the list.
2191 ;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
2192
2193 (defvar c-state-nonlit-pos-cache-limit 1)
2194 (make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2195 ;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2196 ;; reduced by buffer changes, and increased by invocations of
2197 ;; `c-state-literal-at'.
2198
2199 (defvar c-state-semi-nonlit-pos-cache nil)
2200 (make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2201 ;; A list of buffer positions which are known not to be in a literal. This is
2202 ;; ordered with higher positions at the front of the list. Only those which
2203 ;; are less than `c-state-semi-nonlit-pos-cache-limit' are valid.
2204
2205 (defvar c-state-semi-nonlit-pos-cache-limit 1)
2206 (make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2207 ;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This is
2208 ;; reduced by buffer changes, and increased by invocations of
2209 ;; `c-state-literal-at'. FIXME!!!
2210
2211 (defsubst c-state-pp-to-literal (from to &optional not-in-delimiter)
2212 ;; Do a parse-partial-sexp from FROM to TO, returning either
2213 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2214 ;; (STATE) otherwise,
2215 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2216 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2217 ;;
2218 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2219 ;; comment opener, this is recognized as being in a comment literal.
2220 ;;
2221 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2222 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2223 ;; STATE are valid.
2224 (save-excursion
2225 (let ((s (parse-partial-sexp from to))
2226 ty co-st)
2227 (cond
2228 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
2229 (setq ty (cond
2230 ((nth 3 s) 'string)
2231 ((nth 7 s) 'c++)
2232 (t 'c)))
2233 (parse-partial-sexp (point) (point-max)
2234 nil ; TARGETDEPTH
2235 nil ; STOPBEFORE
2236 s ; OLDSTATE
2237 'syntax-table) ; stop at end of literal
2238 `(,s ,ty (,(nth 8 s) . ,(point))))
2239
2240 ((and (not not-in-delimiter) ; inside a comment starter
2241 (not (bobp))
2242 (progn (backward-char)
2243 (and (not (looking-at "\\s!"))
2244 (looking-at c-comment-start-regexp))))
2245 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++)
2246 co-st (point))
2247 (forward-comment 1)
2248 `(,s ,ty (,co-st . ,(point))))
2249
2250 (t `(,s))))))
2251
2252 (defun c-state-safe-place (here)
2253 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2254 ;; string, comment, or macro.
2255 ;;
2256 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2257 ;; MAY NOT contain any positions within macros, since macros are frequently
2258 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2259 ;; We cannot rely on this mechanism whilst determining a cache pos since
2260 ;; this function is also called from outwith `c-parse-state'.
2261 (save-restriction
2262 (widen)
2263 (save-excursion
2264 (let ((c c-state-nonlit-pos-cache)
2265 pos npos high-pos lit macro-beg macro-end)
2266 ;; Trim the cache to take account of buffer changes.
2267 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2268 (setq c (cdr c)))
2269 (setq c-state-nonlit-pos-cache c)
2270
2271 (while (and c (> (car c) here))
2272 (setq high-pos (car c))
2273 (setq c (cdr c)))
2274 (setq pos (or (car c) (point-min)))
2275
2276 (unless high-pos
2277 (while
2278 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2279 (and
2280 (setq npos
2281 (when (<= (+ pos c-state-nonlit-pos-interval) here)
2282 (+ pos c-state-nonlit-pos-interval)))
2283
2284 ;; Test for being in a literal. If so, go to after it.
2285 (progn
2286 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2287 (or (null lit)
2288 (prog1 (<= (cdr lit) here)
2289 (setq npos (cdr lit)))))
2290
2291 ;; Test for being in a macro. If so, go to after it.
2292 (progn
2293 (goto-char npos)
2294 (setq macro-beg
2295 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2296 (when macro-beg
2297 (c-syntactic-end-of-macro)
2298 (or (eobp) (forward-char))
2299 (setq macro-end (point)))
2300 (or (null macro-beg)
2301 (prog1 (<= macro-end here)
2302 (setq npos macro-end)))))
2303
2304 (setq pos npos)
2305 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2306 ;; Add one extra element above HERE so as to to avoid the previous
2307 ;; expensive calculation when the next call is close to the current
2308 ;; one. This is especially useful when inside a large macro.
2309 (when npos
2310 (setq c-state-nonlit-pos-cache
2311 (cons npos c-state-nonlit-pos-cache))))
2312
2313 (if (> pos c-state-nonlit-pos-cache-limit)
2314 (setq c-state-nonlit-pos-cache-limit pos))
2315 pos))))
2316
2317 (defun c-state-semi-safe-place (here)
2318 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2319 ;; string or comment. It may be in a macro.
2320 (save-restriction
2321 (widen)
2322 (save-excursion
2323 (let ((c c-state-semi-nonlit-pos-cache)
2324 pos npos high-pos lit macro-beg macro-end)
2325 ;; Trim the cache to take account of buffer changes.
2326 (while (and c (> (car c) c-state-semi-nonlit-pos-cache-limit))
2327 (setq c (cdr c)))
2328 (setq c-state-semi-nonlit-pos-cache c)
2329
2330 (while (and c (> (car c) here))
2331 (setq high-pos (car c))
2332 (setq c (cdr c)))
2333 (setq pos (or (car c) (point-min)))
2334
2335 (unless high-pos
2336 (while
2337 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2338 (and
2339 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2340
2341 ;; Test for being in a literal. If so, go to after it.
2342 (progn
2343 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2344 (or (null lit)
2345 (prog1 (<= (cdr lit) here)
2346 (setq npos (cdr lit))))))
2347
2348 (setq pos npos)
2349 (setq c-state-semi-nonlit-pos-cache
2350 (cons pos c-state-semi-nonlit-pos-cache))))
2351
2352 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2353 (setq c-state-semi-nonlit-pos-cache-limit pos))
2354 pos))))
2355
2356 (defun c-state-literal-at (here)
2357 ;; If position HERE is inside a literal, return (START . END), the
2358 ;; boundaries of the literal (which may be outside the accessible bit of the
2359 ;; buffer). Otherwise, return nil.
2360 ;;
2361 ;; This function is almost the same as `c-literal-limits'. Previously, it
2362 ;; differed in that it was a lower level function, and that it rigorously
2363 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2364 ;; virtually identical to this function.
2365 (save-restriction
2366 (widen)
2367 (save-excursion
2368 (let ((pos (c-state-safe-place here)))
2369 (car (cddr (c-state-pp-to-literal pos here)))))))
2370
2371 (defsubst c-state-lit-beg (pos)
2372 ;; Return the start of the literal containing POS, or POS itself.
2373 (or (car (c-state-literal-at pos))
2374 pos))
2375
2376 (defsubst c-state-cache-non-literal-place (pos state)
2377 ;; Return a position outside of a string/comment/macro at or before POS.
2378 ;; STATE is the parse-partial-sexp state at POS.
2379 (let ((res (if (or (nth 3 state) ; in a string?
2380 (nth 4 state)) ; in a comment?
2381 (nth 8 state)
2382 pos)))
2383 (save-excursion
2384 (goto-char res)
2385 (if (c-beginning-of-macro)
2386 (point)
2387 res))))
2388
2389 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2390 ;; Stuff to do with point-min, and coping with any literal there.
2391 (defvar c-state-point-min 1)
2392 (make-variable-buffer-local 'c-state-point-min)
2393 ;; This is (point-min) when `c-state-cache' was last calculated. A change of
2394 ;; narrowing is likely to affect the parens that are visible before the point.
2395
2396 (defvar c-state-point-min-lit-type nil)
2397 (make-variable-buffer-local 'c-state-point-min-lit-type)
2398 (defvar c-state-point-min-lit-start nil)
2399 (make-variable-buffer-local 'c-state-point-min-lit-start)
2400 ;; These two variables define the literal, if any, containing point-min.
2401 ;; Their values are, respectively, 'string, c, or c++, and the start of the
2402 ;; literal. If there's no literal there, they're both nil.
2403
2404 (defvar c-state-min-scan-pos 1)
2405 (make-variable-buffer-local 'c-state-min-scan-pos)
2406 ;; This is the earliest buffer-pos from which scanning can be done. It is
2407 ;; either the end of the literal containing point-min, or point-min itself.
2408 ;; It becomes nil if the buffer is changed earlier than this point.
2409 (defun c-state-get-min-scan-pos ()
2410 ;; Return the lowest valid scanning pos. This will be the end of the
2411 ;; literal enclosing point-min, or point-min itself.
2412 (or c-state-min-scan-pos
2413 (save-restriction
2414 (save-excursion
2415 (widen)
2416 (goto-char c-state-point-min-lit-start)
2417 (if (eq c-state-point-min-lit-type 'string)
2418 (forward-sexp)
2419 (forward-comment 1))
2420 (setq c-state-min-scan-pos (point))))))
2421
2422 (defun c-state-mark-point-min-literal ()
2423 ;; Determine the properties of any literal containing POINT-MIN, setting the
2424 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2425 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2426 (let ((p-min (point-min))
2427 lit)
2428 (save-restriction
2429 (widen)
2430 (setq lit (c-state-literal-at p-min))
2431 (if lit
2432 (setq c-state-point-min-lit-type
2433 (save-excursion
2434 (goto-char (car lit))
2435 (cond
2436 ((looking-at c-block-comment-start-regexp) 'c)
2437 ((looking-at c-line-comment-starter) 'c++)
2438 (t 'string)))
2439 c-state-point-min-lit-start (car lit)
2440 c-state-min-scan-pos (cdr lit))
2441 (setq c-state-point-min-lit-type nil
2442 c-state-point-min-lit-start nil
2443 c-state-min-scan-pos p-min)))))
2444
2445
2446 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2447 ;; A variable which signals a brace dessert - helpful for reducing the number
2448 ;; of fruitless backward scans.
2449 (defvar c-state-brace-pair-desert nil)
2450 (make-variable-buffer-local 'c-state-brace-pair-desert)
2451 ;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
2452 ;; that defun has searched backwards for a brace pair and not found one. Its
2453 ;; value is either nil or a cons (PA . FROM), where PA is the position of the
2454 ;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2455 ;; nil when at top level) and FROM is where the backward search started. It
2456 ;; is reset to nil in `c-invalidate-state-cache'.
2457
2458
2459 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2460 ;; Lowish level functions/macros which work directly on `c-state-cache', or a
2461 ;; list of like structure.
2462 (defmacro c-state-cache-top-lparen (&optional cache)
2463 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2464 ;; (default `c-state-cache') (or nil).
2465 (let ((cash (or cache 'c-state-cache)))
2466 `(if (consp (car ,cash))
2467 (caar ,cash)
2468 (car ,cash))))
2469
2470 (defmacro c-state-cache-top-paren (&optional cache)
2471 ;; Return the address of the latest brace/bracket/paren (whether left or
2472 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2473 (let ((cash (or cache 'c-state-cache)))
2474 `(if (consp (car ,cash))
2475 (cdar ,cash)
2476 (car ,cash))))
2477
2478 (defmacro c-state-cache-after-top-paren (&optional cache)
2479 ;; Return the position just after the latest brace/bracket/paren (whether
2480 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2481 (let ((cash (or cache 'c-state-cache)))
2482 `(if (consp (car ,cash))
2483 (cdar ,cash)
2484 (and (car ,cash)
2485 (1+ (car ,cash))))))
2486
2487 (defun c-get-cache-scan-pos (here)
2488 ;; From the state-cache, determine the buffer position from which we might
2489 ;; scan forward to HERE to update this cache. This position will be just
2490 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2491 ;; return the earliest position in the accessible region which isn't within
2492 ;; a literal. If the visible portion of the buffer is entirely within a
2493 ;; literal, return NIL.
2494 (let ((c c-state-cache) elt)
2495 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2496 (while (and c
2497 (>= (c-state-cache-top-lparen c) here))
2498 (setq c (cdr c)))
2499
2500 (setq elt (car c))
2501 (cond
2502 ((consp elt)
2503 (if (> (cdr elt) here)
2504 (1+ (car elt))
2505 (cdr elt)))
2506 (elt (1+ elt))
2507 ((<= (c-state-get-min-scan-pos) here)
2508 (c-state-get-min-scan-pos))
2509 (t nil))))
2510
2511 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2512 ;; Variables which keep track of preprocessor constructs.
2513 (defvar c-state-old-cpp-beg-marker nil)
2514 (make-variable-buffer-local 'c-state-old-cpp-beg-marker)
2515 (defvar c-state-old-cpp-beg nil)
2516 (make-variable-buffer-local 'c-state-old-cpp-beg)
2517 (defvar c-state-old-cpp-end-marker nil)
2518 (make-variable-buffer-local 'c-state-old-cpp-end-marker)
2519 (defvar c-state-old-cpp-end nil)
2520 (make-variable-buffer-local 'c-state-old-cpp-end)
2521 ;; These are the limits of the macro containing point at the previous call of
2522 ;; `c-parse-state', or nil.
2523
2524 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2525 ;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
2526 (defun c-get-fallback-scan-pos (here)
2527 ;; Return a start position for building `c-state-cache' from
2528 ;; scratch. This will be at the top level, 2 defuns back.
2529 (save-excursion
2530 ;; Go back 2 bods, but ignore any bogus positions returned by
2531 ;; beginning-of-defun (i.e. open paren in column zero).
2532 (goto-char here)
2533 (let ((cnt 2))
2534 (while (not (or (bobp) (zerop cnt)))
2535 (c-beginning-of-defun-1) ; Pure elisp BOD.
2536 (if (eq (char-after) ?\{)
2537 (setq cnt (1- cnt)))))
2538 (point)))
2539
2540 (defun c-state-balance-parens-backwards (here- here+ top)
2541 ;; Return the position of the opening paren/brace/bracket before HERE- which
2542 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2543 ;; there's a macro, HERE- and HERE+ are the same. Like this:
2544 ;;
2545 ;; ............................................
2546 ;; | |
2547 ;; ( [ ( .........#macro.. ) ( ) ] )
2548 ;; ^ ^ ^ ^
2549 ;; | | | |
2550 ;; return HERE- HERE+ TOP
2551 ;;
2552 ;; If there aren't enough opening paren/brace/brackets, return the position
2553 ;; of the outermost one found, or HERE- if there are none. If there are no
2554 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
2555 ;; must not be inside literals. Only the accessible portion of the buffer
2556 ;; will be scanned.
2557
2558 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2559 ;; `here'. Go round the next loop each time we pass over such a ")". These
2560 ;; probably match "("s before `here-'.
2561 (let (pos pa ren+1 lonely-rens)
2562 (save-excursion
2563 (save-restriction
2564 (narrow-to-region (point-min) top) ; This can move point, sometimes.
2565 (setq pos here+)
2566 (c-safe
2567 (while
2568 (setq ren+1 (c-sc-scan-lists pos 1 1)) ; might signal
2569 (setq lonely-rens (cons ren+1 lonely-rens)
2570 pos ren+1)))))
2571
2572 ;; PART 2: Scan back before `here-' searching for the "("s
2573 ;; matching/mismatching the ")"s found above. We only need to direct the
2574 ;; caller to scan when we've encountered unmatched right parens.
2575 (setq pos here-)
2576 (when lonely-rens
2577 (c-safe
2578 (while
2579 (and lonely-rens ; actual values aren't used.
2580 (setq pa (c-sc-scan-lists pos -1 1)))
2581 (setq pos pa)
2582 (setq lonely-rens (cdr lonely-rens)))))
2583 pos))
2584
2585 (defun c-parse-state-get-strategy (here good-pos)
2586 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
2587 ;; to minimize the amount of scanning. HERE is the pertinent position in
2588 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2589 ;; its head trimmed) is known to be good, or nil if there is no such
2590 ;; position.
2591 ;;
2592 ;; The return value is a list, one of the following:
2593 ;;
2594 ;; o - ('forward START-POINT) - scan forward from START-POINT,
2595 ;; which is not less than the highest position in `c-state-cache' below HERE,
2596 ;; which is after GOOD-POS.
2597 ;; o - ('backward nil) - scan backwards (from HERE).
2598 ;; o - ('back-and-forward START-POINT) - like 'forward, but when HERE is earlier
2599 ;; than GOOD-POS.
2600 ;; o - ('BOD START-POINT) - scan forwards from START-POINT, which is at the
2601 ;; top level.
2602 ;; o - ('IN-LIT nil) - point is inside the literal containing point-min.
2603 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2604 BOD-pos ; position of 2nd BOD before HERE.
2605 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2606 start-point
2607 how-far) ; putative scanning distance.
2608 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2609 (cond
2610 ((< here (c-state-get-min-scan-pos))
2611 (setq strategy 'IN-LIT
2612 start-point nil
2613 cache-pos nil
2614 how-far 0))
2615 ((<= good-pos here)
2616 (setq strategy 'forward
2617 start-point (max good-pos cache-pos)
2618 how-far (- here start-point)))
2619 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2620 (setq strategy 'backward
2621 how-far (- good-pos here)))
2622 (t
2623 (setq strategy 'back-and-forward
2624 start-point cache-pos
2625 how-far (- here start-point))))
2626
2627 ;; Might we be better off starting from the top level, two defuns back,
2628 ;; instead? This heuristic no longer works well in C++, where
2629 ;; declarations inside namespace brace blocks are frequently placed at
2630 ;; column zero. (2015-11-10): Remove the condition on C++ Mode.
2631 (when (and (or (not (memq 'col-0-paren c-emacs-features))
2632 open-paren-in-column-0-is-defun-start)
2633 ;; (not (c-major-mode-is 'c++-mode))
2634 (> how-far c-state-cache-too-far))
2635 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2636 (if (< (- here BOD-pos) how-far)
2637 (setq strategy 'BOD
2638 start-point BOD-pos)))
2639
2640 (list strategy start-point)))
2641
2642
2643 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2644 ;; Routines which change `c-state-cache' and associated values.
2645 (defun c-renarrow-state-cache ()
2646 ;; The region (more precisely, point-min) has changed since we
2647 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2648 (if (< (point-min) c-state-point-min)
2649 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2650 ;; It would be possible to do a better job here and recalculate the top
2651 ;; only.
2652 (progn
2653 (c-state-mark-point-min-literal)
2654 (setq c-state-cache nil
2655 c-state-cache-good-pos c-state-min-scan-pos
2656 c-state-brace-pair-desert nil))
2657
2658 ;; point-min has MOVED FORWARD.
2659
2660 ;; Is the new point-min inside a (different) literal?
2661 (unless (and c-state-point-min-lit-start ; at prev. point-min
2662 (< (point-min) (c-state-get-min-scan-pos)))
2663 (c-state-mark-point-min-literal))
2664
2665 ;; Cut off a bit of the tail from `c-state-cache'.
2666 (let ((ptr (cons nil c-state-cache))
2667 pa)
2668 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2669 (>= pa (point-min)))
2670 (setq ptr (cdr ptr)))
2671
2672 (when (consp ptr)
2673 (if (eq (cdr ptr) c-state-cache)
2674 (setq c-state-cache nil
2675 c-state-cache-good-pos c-state-min-scan-pos)
2676 (setcdr ptr nil)
2677 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2678 )))
2679
2680 (setq c-state-point-min (point-min)))
2681
2682 (defun c-append-lower-brace-pair-to-state-cache (from here &optional upper-lim)
2683 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2684 ;; of nesting (not necessarily immediately preceding), push a cons onto
2685 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2686 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2687 ;; UPPER-LIM.
2688 ;;
2689 ;; Return non-nil when this has been done.
2690 ;;
2691 ;; The situation it copes with is this transformation:
2692 ;;
2693 ;; OLD: { (.) {...........}
2694 ;; ^ ^
2695 ;; FROM HERE
2696 ;;
2697 ;; NEW: { {....} (.) {.........
2698 ;; ^ ^ ^
2699 ;; LOWER BRACE PAIR HERE or HERE
2700 ;;
2701 ;; This routine should be fast. Since it can get called a LOT, we maintain
2702 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2703 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2704 (save-excursion
2705 (save-restriction
2706 (let* (new-cons
2707 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2708 (macro-start-or-from
2709 (progn (goto-char from)
2710 (c-beginning-of-macro)
2711 (point)))
2712 (bra ; Position of "{".
2713 ;; Don't start scanning in the middle of a CPP construct unless
2714 ;; it contains HERE - these constructs, in Emacs, are "commented
2715 ;; out" with category properties.
2716 (if (eq (c-get-char-property macro-start-or-from 'category)
2717 'c-cpp-delimiter)
2718 macro-start-or-from
2719 from))
2720 ce) ; Position of "}"
2721 (or upper-lim (setq upper-lim from))
2722
2723 ;; If we're essentially repeating a fruitless search, just give up.
2724 (unless (and c-state-brace-pair-desert
2725 (eq cache-pos (car c-state-brace-pair-desert))
2726 (or (null (car c-state-brace-pair-desert))
2727 (> from (car c-state-brace-pair-desert)))
2728 (<= from (cdr c-state-brace-pair-desert)))
2729 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
2730 (let ((desert-lim
2731 (and c-state-brace-pair-desert
2732 (eq cache-pos (car c-state-brace-pair-desert))
2733 (>= from (cdr c-state-brace-pair-desert))
2734 (cdr c-state-brace-pair-desert)))
2735 ;; CACHE-LIM. This limit will be necessary when an opening
2736 ;; paren at `cache-pos' has just had its matching close paren
2737 ;; inserted into the buffer. `cache-pos' continues to be a
2738 ;; search bound, even though the algorithm below would skip
2739 ;; over the new paren pair.
2740 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2741 (narrow-to-region
2742 (cond
2743 ((and desert-lim cache-lim)
2744 (max desert-lim cache-lim))
2745 (desert-lim)
2746 (cache-lim)
2747 ((point-min)))
2748 ;; The top limit is EOB to ensure that `bra' is inside the
2749 ;; accessible part of the buffer at the next scan operation.
2750 (1+ (buffer-size))))
2751
2752 ;; In the next pair of nested loops, the inner one moves back past a
2753 ;; pair of (mis-)matching parens or brackets; the outer one moves
2754 ;; back over a sequence of unmatched close brace/paren/bracket each
2755 ;; time round.
2756 (while
2757 (progn
2758 (c-safe
2759 (while
2760 (and (setq ce (c-sc-scan-lists bra -1 -1)) ; back past )/]/}; might signal
2761 (setq bra (c-sc-scan-lists ce -1 1)) ; back past (/[/{; might signal
2762 (or (> bra here) ;(> ce here)
2763 (and
2764 (< ce here)
2765 (or (not (eq (char-after bra) ?\{))
2766 (and (goto-char bra)
2767 (c-beginning-of-macro)
2768 (< (point) macro-start-or-from))))))))
2769 (and ce (< ce bra)))
2770 (setq bra ce)) ; If we just backed over an unbalanced closing
2771 ; brace, ignore it.
2772
2773 (if (and ce (< ce here) (< bra ce) (eq (char-after bra) ?\{))
2774 ;; We've found the desired brace-pair.
2775 (progn
2776 (setq new-cons (cons bra (1+ ce)))
2777 (cond
2778 ((consp (car c-state-cache))
2779 (setcar c-state-cache new-cons))
2780 ((and (numberp (car c-state-cache)) ; probably never happens
2781 (< ce (car c-state-cache)))
2782 (setcdr c-state-cache
2783 (cons new-cons (cdr c-state-cache))))
2784 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2785
2786 ;; We haven't found a brace pair. Record this in the cache.
2787 (setq c-state-brace-pair-desert
2788 (cons (if (and ce (< bra ce) (> ce here)) ; {..} straddling HERE?
2789 bra
2790 (point-min))
2791 (min here from)))))))))
2792
2793 (defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2794 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2795 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2796 ;; "push" "a" brace pair onto `c-state-cache'.
2797 ;;
2798 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2799 ;; otherwise push it normally.
2800 ;;
2801 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2802 ;; latter is inside a macro, not being a macro containing
2803 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2804 ;; base pair. This latter case is assumed to be rare.
2805 ;;
2806 ;; Note: POINT is not preserved in this routine.
2807 (if bra+1
2808 (if (or (> bra+1 macro-start-or-here)
2809 (progn (goto-char bra+1)
2810 (not (c-beginning-of-macro))))
2811 (setq c-state-cache
2812 (cons (cons (1- bra+1)
2813 (c-sc-scan-lists bra+1 1 1))
2814 (if (consp (car c-state-cache))
2815 (cdr c-state-cache)
2816 c-state-cache)))
2817 ;; N.B. This defsubst codes one method for the simple, normal case,
2818 ;; and a more sophisticated, slower way for the general case. Don't
2819 ;; eliminate this defsubst - it's a speed optimization.
2820 (c-append-lower-brace-pair-to-state-cache (1- bra+1) (point-max)))))
2821
2822 (defun c-append-to-state-cache (from here)
2823 ;; Scan the buffer from FROM to HERE, adding elements into `c-state-cache'
2824 ;; for braces etc. Return a candidate for `c-state-cache-good-pos'.
2825 ;;
2826 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2827 ;; any. Typically, it is immediately after it. It must not be inside a
2828 ;; literal.
2829 (let ((here-bol (c-point 'bol here))
2830 (macro-start-or-here
2831 (save-excursion (goto-char here)
2832 (if (c-beginning-of-macro)
2833 (point)
2834 here)))
2835 pa+1 ; pos just after an opening PAren (or brace).
2836 (ren+1 from) ; usually a pos just after an closing paREN etc.
2837 ; Is actually the pos. to scan for a (/{/[ from,
2838 ; which sometimes is after a silly )/}/].
2839 paren+1 ; Pos after some opening or closing paren.
2840 paren+1s ; A list of `paren+1's; used to determine a
2841 ; good-pos.
2842 bra+1 ; just after L bra-ce.
2843 bra+1s ; list of OLD values of bra+1.
2844 mstart) ; start of a macro.
2845
2846 (save-excursion
2847 (save-restriction
2848 (narrow-to-region (point-min) here)
2849 ;; Each time round the following loop, we enter a successively deeper
2850 ;; level of brace/paren nesting. (Except sometimes we "continue at
2851 ;; the existing level".) `pa+1' is a pos inside an opening
2852 ;; brace/paren/bracket, usually just after it.
2853 (while
2854 (progn
2855 ;; Each time round the next loop moves forward over an opening then
2856 ;; a closing brace/bracket/paren. This loop is white hot, so it
2857 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2858 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2859 ;; call of `scan-lists' signals an error, which happens when there
2860 ;; are no more b/b/p's to scan.
2861 (c-safe
2862 (while t
2863 (setq pa+1 (c-sc-scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2864 paren+1s (cons pa+1 paren+1s))
2865 (setq ren+1 (c-sc-scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2866 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2867 (setq bra+1 pa+1))
2868 (setcar paren+1s ren+1)))
2869
2870 (if (and pa+1 (> pa+1 ren+1))
2871 ;; We've just entered a deeper nesting level.
2872 (progn
2873 ;; Insert the brace pair (if present) and the single open
2874 ;; paren/brace/bracket into `c-state-cache' It cannot be
2875 ;; inside a macro, except one around point, because of what
2876 ;; `c-neutralize-syntax-in-CPP' has done.
2877 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2878 ;; Insert the opening brace/bracket/paren position.
2879 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2880 ;; Clear admin stuff for the next more nested part of the scan.
2881 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2882 t) ; Carry on the loop
2883
2884 ;; All open p/b/b's at this nesting level, if any, have probably
2885 ;; been closed by matching/mismatching ones. We're probably
2886 ;; finished - we just need to check for having found an
2887 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2888 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2889 (c-safe (setq ren+1 (c-sc-scan-lists ren+1 1 1)))))) ; acts as loop control.
2890
2891 ;; Record the final, innermost, brace-pair if there is one.
2892 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2893
2894 ;; Determine a good pos
2895 (while (and (setq paren+1 (car paren+1s))
2896 (> (if (> paren+1 macro-start-or-here)
2897 paren+1
2898 (goto-char paren+1)
2899 (setq mstart (and (c-beginning-of-macro)
2900 (point)))
2901 (or mstart paren+1))
2902 here-bol))
2903 (setq paren+1s (cdr paren+1s)))
2904 (cond
2905 ((and paren+1 mstart)
2906 (min paren+1 mstart))
2907 (paren+1)
2908 (t from))))))
2909
2910 (defun c-remove-stale-state-cache (start-point here pps-point)
2911 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2912 ;; not be in it when it is amended for position HERE. This may involve
2913 ;; replacing a CONS element for a brace pair containing HERE with its car.
2914 ;; Additionally, the "outermost" open-brace entry before HERE will be
2915 ;; converted to a cons if the matching close-brace is below HERE.
2916 ;;
2917 ;; START-POINT is a "maximal" "safe position" - there must be no open
2918 ;; parens/braces/brackets between START-POINT and HERE.
2919 ;;
2920 ;; As a second thing, calculate the result of parse-partial-sexp at
2921 ;; PPS-POINT, w.r.t. START-POINT. The motivation here is that
2922 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2923 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2924 ;; needs to be FAST).
2925 ;;
2926 ;; Return a list (GOOD-POS SCAN-BACK-POS CONS-SEPARATED PPS-STATE), where
2927 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2928 ;; to be good (we aim for this to be as high as possible);
2929 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2930 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2931 ;; position to scan backwards from. It is the position of the "{" of the
2932 ;; last element to be removed from `c-state-cache', when that elt is a
2933 ;; cons, otherwise nil.
2934 ;; o - CONS-SEPARATED is t when a cons element in `c-state-cache' has been
2935 ;; replaced by its car because HERE lies inside the brace pair represented
2936 ;; by the cons.
2937 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2938 (save-excursion
2939 (save-restriction
2940 (narrow-to-region 1 (point-max))
2941 (let* ((in-macro-start ; start of macro containing HERE or nil.
2942 (save-excursion
2943 (goto-char here)
2944 (and (c-beginning-of-macro)
2945 (point))))
2946 (start-point-actual-macro-start ; Start of macro containing
2947 ; start-point or nil
2948 (and (< start-point here)
2949 (save-excursion
2950 (goto-char start-point)
2951 (and (c-beginning-of-macro)
2952 (point)))))
2953 (start-point-actual-macro-end ; End of this macro, (maybe
2954 ; HERE), or nil.
2955 (and start-point-actual-macro-start
2956 (save-excursion
2957 (goto-char start-point-actual-macro-start)
2958 (c-end-of-macro)
2959 (point))))
2960 pps-state ; Will be 9 or 10 elements long.
2961 pos
2962 upper-lim ; ,beyond which `c-state-cache' entries are removed
2963 scan-back-pos
2964 cons-separated
2965 pair-beg pps-point-state target-depth)
2966
2967 ;; Remove entries beyond HERE. Also remove any entries inside
2968 ;; a macro, unless HERE is in the same macro.
2969 (setq upper-lim
2970 (if (or (null c-state-old-cpp-beg)
2971 (and (> here c-state-old-cpp-beg)
2972 (< here c-state-old-cpp-end)))
2973 here
2974 (min here c-state-old-cpp-beg)))
2975 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
2976 (setq scan-back-pos (car-safe (car c-state-cache)))
2977 (setq c-state-cache (cdr c-state-cache)))
2978
2979 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2980 ;; RBrace and indicate we'll need to search backwards for a previous
2981 ;; brace pair.
2982 (when (and c-state-cache
2983 (consp (car c-state-cache))
2984 (> (cdar c-state-cache) upper-lim))
2985 (setcar c-state-cache (caar c-state-cache))
2986 (setq scan-back-pos (car c-state-cache)
2987 cons-separated t))
2988
2989 ;; The next loop jumps forward out of a nested level of parens each
2990 ;; time round; the corresponding elements in `c-state-cache' are
2991 ;; removed. `pos' is just after the brace-pair or the open paren at
2992 ;; (car c-state-cache). There can be no open parens/braces/brackets
2993 ;; between `start-point'/`start-point-actual-macro-start' and HERE,
2994 ;; due to the interface spec to this function.
2995 (setq pos (if (and start-point-actual-macro-end
2996 (not (eq start-point-actual-macro-start
2997 in-macro-start)))
2998 (1+ start-point-actual-macro-end) ; get outside the macro as
2999 ; marked by a `category' text property.
3000 start-point))
3001 (goto-char pos)
3002 (while (and c-state-cache
3003 (or (numberp (car c-state-cache)) ; Have we a { at all?
3004 (cdr c-state-cache))
3005 (< (point) here))
3006 (cond
3007 ((null pps-state) ; first time through
3008 (setq target-depth -1))
3009 ((eq (car pps-state) target-depth) ; found closing ),},]
3010 (setq target-depth (1- (car pps-state))))
3011 ;; Do nothing when we've merely reached pps-point.
3012 )
3013
3014 ;; Scan!
3015 (setq pps-state
3016 (c-sc-parse-partial-sexp
3017 (point) (if (< (point) pps-point) pps-point here)
3018 target-depth
3019 nil pps-state))
3020
3021 (if (= (point) pps-point)
3022 (setq pps-point-state pps-state))
3023
3024 (when (eq (car pps-state) target-depth)
3025 (setq pos (point)) ; POS is now just after an R-paren/brace.
3026 (cond
3027 ((and (consp (car c-state-cache))
3028 (eq (point) (cdar c-state-cache)))
3029 ;; We've just moved out of the paren pair containing the brace-pair
3030 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
3031 ;; and is potentially where the open brace of a cons in
3032 ;; c-state-cache will be.
3033 (setq pair-beg (car-safe (cdr c-state-cache))
3034 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
3035 ((numberp (car c-state-cache))
3036 (setq pair-beg (car c-state-cache)
3037 c-state-cache (cdr c-state-cache))) ; remove this
3038 ; containing Lparen
3039 ((numberp (cadr c-state-cache))
3040 (setq pair-beg (cadr c-state-cache)
3041 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
3042 ; together with enclosed brace pair.
3043 ;; (t nil) ; Ignore an unmated Rparen.
3044 )))
3045
3046 (if (< (point) pps-point)
3047 (setq pps-state (c-sc-parse-partial-sexp
3048 (point) pps-point
3049 nil nil ; TARGETDEPTH, STOPBEFORE
3050 pps-state)))
3051
3052 ;; If the last paren pair we moved out of was actually a brace pair,
3053 ;; insert it into `c-state-cache'.
3054 (when (and pair-beg (eq (char-after pair-beg) ?{))
3055 (if (consp (car-safe c-state-cache))
3056 (setq c-state-cache (cdr c-state-cache)))
3057 (setq c-state-cache (cons (cons pair-beg pos)
3058 c-state-cache)))
3059
3060 (list pos scan-back-pos cons-separated pps-state)))))
3061
3062 (defun c-remove-stale-state-cache-backwards (here)
3063 ;; Strip stale elements of `c-state-cache' by moving backwards through the
3064 ;; buffer, and inform the caller of the scenario detected.
3065 ;;
3066 ;; HERE is the position we're setting `c-state-cache' for.
3067 ;; CACHE-POS (a locally bound variable) is just after the latest recorded
3068 ;; position in `c-state-cache' before HERE, or a position at or near
3069 ;; point-min which isn't in a literal.
3070 ;;
3071 ;; This function must only be called only when (> `c-state-cache-good-pos'
3072 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
3073 ;; optimized to eliminate (or minimize) scanning between these two
3074 ;; positions.
3075 ;;
3076 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
3077 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
3078 ;; could become so after missing elements are inserted into
3079 ;; `c-state-cache'. This is JUST AFTER an opening or closing
3080 ;; brace/paren/bracket which is already in `c-state-cache' or just before
3081 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
3082 ;; before `here''s line, or the start of the literal containing it.
3083 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
3084 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
3085 ;; to scan backwards from.
3086 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
3087 ;; POS and HERE which aren't recorded in `c-state-cache'.
3088 ;;
3089 ;; The comments in this defun use "paren" to mean parenthesis or square
3090 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
3091 ;;
3092 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
3093 ;; | | | | | |
3094 ;; CP E here D C good
3095 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
3096 (pos c-state-cache-good-pos)
3097 pa ren ; positions of "(" and ")"
3098 dropped-cons ; whether the last element dropped from `c-state-cache'
3099 ; was a cons (representing a brace-pair)
3100 good-pos ; see above.
3101 lit ; (START . END) of a literal containing some point.
3102 here-lit-start here-lit-end ; bounds of literal containing `here'
3103 ; or `here' itself.
3104 here- here+ ; start/end of macro around HERE, or HERE
3105 (here-bol (c-point 'bol here))
3106 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
3107
3108 ;; Remove completely irrelevant entries from `c-state-cache'.
3109 (while (and c-state-cache
3110 (>= (setq pa (c-state-cache-top-lparen)) here))
3111 (setq dropped-cons (consp (car c-state-cache)))
3112 (setq c-state-cache (cdr c-state-cache))
3113 (setq pos pa))
3114 ;; At this stage, (>= pos here);
3115 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3116
3117 (cond
3118 ((and (consp (car c-state-cache))
3119 (> (cdar c-state-cache) here))
3120 ;; CASE 1: The top of the cache is a brace pair which now encloses
3121 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3122 ;; knowledge of what's inside these braces, we have no alternative but
3123 ;; to direct the caller to scan the buffer from the opening brace.
3124 (setq pos (caar c-state-cache))
3125 (setcar c-state-cache pos)
3126 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3127 ; entry into a { entry, so the caller needs to
3128 ; search for a brace pair before the {.
3129
3130 ;; `here' might be inside a literal. Check for this.
3131 ((progn
3132 (setq lit (c-state-literal-at here)
3133 here-lit-start (or (car lit) here)
3134 here-lit-end (or (cdr lit) here))
3135 ;; Has `here' just "newly entered" a macro?
3136 (save-excursion
3137 (goto-char here-lit-start)
3138 (if (and (c-beginning-of-macro)
3139 (or (null c-state-old-cpp-beg)
3140 (not (= (point) c-state-old-cpp-beg))))
3141 (progn
3142 (setq here- (point))
3143 (c-end-of-macro)
3144 (setq here+ (point)))
3145 (setq here- here-lit-start
3146 here+ here-lit-end)))
3147
3148 ;; `here' might be nested inside any depth of parens (or brackets but
3149 ;; not braces). Scan backwards to find the outermost such opening
3150 ;; paren, if there is one. This will be the scan position to return.
3151 (save-restriction
3152 (narrow-to-region cache-pos (point-max))
3153 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3154 nil)) ; for the cond
3155
3156 ((< pos here-lit-start)
3157 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3158 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3159 ;; a brace pair preceding this, it will already be in `c-state-cache',
3160 ;; unless there was a brace pair after it, i.e. there'll only be one to
3161 ;; scan for if we've just deleted one.
3162 (list pos (and dropped-cons pos) t)) ; Return value.
3163
3164 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3165 ;; Further forward scanning isn't needed, but we still need to find a
3166 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3167 ((progn
3168 (save-restriction
3169 (narrow-to-region here-bol (point-max))
3170 (setq pos here-lit-start)
3171 (c-safe (while (setq pa (c-sc-scan-lists pos -1 1))
3172 (setq pos pa)))) ; might signal
3173 nil)) ; for the cond
3174
3175 ((save-restriction
3176 (narrow-to-region too-far-back (point-max))
3177 (setq ren (c-safe (c-sc-scan-lists pos -1 -1))))
3178 ;; CASE 3: After a }/)/] before `here''s BOL.
3179 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3180
3181 ((progn (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3182 (>= cache-pos good-pos))
3183 ;; CASE 3.5: Just after an existing entry in `c-state-cache' on `here''s
3184 ;; line or the previous line.
3185 (list cache-pos nil nil))
3186
3187 (t
3188 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3189 ;; literal containing it.
3190 (list good-pos (and dropped-cons good-pos) nil)))))
3191
3192
3193 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3194 ;; Externally visible routines.
3195
3196 (defun c-state-cache-init ()
3197 (setq c-state-cache nil
3198 c-state-cache-good-pos 1
3199 c-state-nonlit-pos-cache nil
3200 c-state-nonlit-pos-cache-limit 1
3201 c-state-semi-nonlit-pos-cache nil
3202 c-state-semi-nonlit-pos-cache-limit 1
3203 c-state-brace-pair-desert nil
3204 c-state-point-min 1
3205 c-state-point-min-lit-type nil
3206 c-state-point-min-lit-start nil
3207 c-state-min-scan-pos 1
3208 c-state-old-cpp-beg nil
3209 c-state-old-cpp-end nil)
3210 (c-state-mark-point-min-literal))
3211
3212 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3213 ;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3214 ;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3215 ;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3216 ;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3217 ;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3218 ;; (defun c-state-dump ()
3219 ;; ;; For debugging.
3220 ;; ;(message
3221 ;; (concat
3222 ;; (c-sc-qde c-state-cache)
3223 ;; (c-sc-de c-state-cache-good-pos)
3224 ;; (c-sc-qde c-state-nonlit-pos-cache)
3225 ;; (c-sc-de c-state-nonlit-pos-cache-limit)
3226 ;; (c-sc-qde c-state-brace-pair-desert)
3227 ;; (c-sc-de c-state-point-min)
3228 ;; (c-sc-de c-state-point-min-lit-type)
3229 ;; (c-sc-de c-state-point-min-lit-start)
3230 ;; (c-sc-de c-state-min-scan-pos)
3231 ;; (c-sc-de c-state-old-cpp-beg)
3232 ;; (c-sc-de c-state-old-cpp-end)))
3233 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3234
3235 (defun c-invalidate-state-cache-1 (here)
3236 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3237 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3238 ;; left in a consistent state.
3239 ;;
3240 ;; This is much like `c-whack-state-after', but it never changes a paren
3241 ;; pair element into an open paren element. Doing that would mean that the
3242 ;; new open paren wouldn't have the required preceding paren pair element.
3243 ;;
3244 ;; This function is called from c-after-change.
3245
3246 ;; The caches of non-literals:
3247 ;; Note that we use "<=" for the possibility of the second char of a two-char
3248 ;; comment opener being typed; this would invalidate any cache position at
3249 ;; HERE.
3250 (if (<= here c-state-nonlit-pos-cache-limit)
3251 (setq c-state-nonlit-pos-cache-limit (1- here)))
3252 (if (<= here c-state-semi-nonlit-pos-cache-limit)
3253 (setq c-state-semi-nonlit-pos-cache-limit (1- here)))
3254
3255 ;; `c-state-cache':
3256 ;; Case 1: if `here' is in a literal containing point-min, everything
3257 ;; becomes (or is already) nil.
3258 (if (or (null c-state-cache-good-pos)
3259 (< here (c-state-get-min-scan-pos)))
3260 (setq c-state-cache nil
3261 c-state-cache-good-pos nil
3262 c-state-min-scan-pos nil)
3263
3264 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3265 ;; below `here'. To maintain its consistency, we may need to insert a new
3266 ;; brace pair.
3267 (let ((here-bol (c-point 'bol here))
3268 too-high-pa ; recorded {/(/[ next above here, or nil.
3269 dropped-cons ; was the last removed element a brace pair?
3270 pa)
3271 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3272 (while (and c-state-cache
3273 (>= (setq pa (c-state-cache-top-paren)) here))
3274 (setq dropped-cons (consp (car c-state-cache))
3275 too-high-pa (c-state-cache-top-lparen)
3276 c-state-cache (cdr c-state-cache)))
3277
3278 ;; Do we need to add in an earlier brace pair, having lopped one off?
3279 (if (and dropped-cons
3280 (< too-high-pa (+ here c-state-cache-too-far)))
3281 (c-append-lower-brace-pair-to-state-cache too-high-pa here here-bol))
3282 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3283 (c-state-get-min-scan-pos)))))
3284
3285 ;; The brace-pair desert marker:
3286 (when (car c-state-brace-pair-desert)
3287 (if (< here (car c-state-brace-pair-desert))
3288 (setq c-state-brace-pair-desert nil)
3289 (if (< here (cdr c-state-brace-pair-desert))
3290 (setcdr c-state-brace-pair-desert here)))))
3291
3292 (defun c-parse-state-1 ()
3293 ;; Find and record all noteworthy parens between some good point earlier in
3294 ;; the file and point. That good point is at least the beginning of the
3295 ;; top-level construct we are in, or the beginning of the preceding
3296 ;; top-level construct if we aren't in one.
3297 ;;
3298 ;; The returned value is a list of the noteworthy parens with the last one
3299 ;; first. If an element in the list is an integer, it's the position of an
3300 ;; open paren (of any type) which has not been closed before the point. If
3301 ;; an element is a cons, it gives the position of a closed BRACE paren
3302 ;; pair[*]; the car is the start brace position and the cdr is the position
3303 ;; following the closing brace. Only the last closed brace paren pair
3304 ;; before each open paren and before the point is recorded, and thus the
3305 ;; state never contains two cons elements in succession. When a close brace
3306 ;; has no matching open brace (e.g., the matching brace is outside the
3307 ;; visible region), it is not represented in the returned value.
3308 ;;
3309 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3310 ;; This defun explicitly treats mismatching parens/braces/brackets as
3311 ;; matching. It is the open brace which makes it a "brace" pair.
3312 ;;
3313 ;; If POINT is within a macro, open parens and brace pairs within
3314 ;; THIS macro MIGHT be recorded. This depends on whether their
3315 ;; syntactic properties have been suppressed by
3316 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
3317 ;;
3318 ;; Currently no characters which are given paren syntax with the
3319 ;; syntax-table property are recorded, i.e. angle bracket arglist
3320 ;; parens are never present here. Note that this might change.
3321 ;;
3322 ;; BUG: This function doesn't cope entirely well with unbalanced
3323 ;; parens in macros. (2008-12-11: this has probably been resolved
3324 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3325 ;; following case the brace before the macro isn't balanced with the
3326 ;; one after it:
3327 ;;
3328 ;; {
3329 ;; #define X {
3330 ;; }
3331 ;;
3332 ;; Note to maintainers: this function DOES get called with point
3333 ;; within comments and strings, so don't assume it doesn't!
3334 ;;
3335 ;; This function might do hidden buffer changes.
3336 (let* ((here (point))
3337 (here-bopl (c-point 'bopl))
3338 strategy ; 'forward, 'backward etc..
3339 ;; Candidate positions to start scanning from:
3340 cache-pos ; highest position below HERE already existing in
3341 ; cache (or 1).
3342 good-pos
3343 start-point ; (when scanning forward) a place below HERE where there
3344 ; are no open parens/braces between it and HERE.
3345 bopl-state
3346 res
3347 cons-separated
3348 scan-backward-pos scan-forward-p) ; used for 'backward.
3349 ;; If POINT-MIN has changed, adjust the cache
3350 (unless (= (point-min) c-state-point-min)
3351 (c-renarrow-state-cache))
3352
3353 ;; Strategy?
3354 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3355 strategy (car res)
3356 start-point (cadr res))
3357
3358 (when (eq strategy 'BOD)
3359 (setq c-state-cache nil
3360 c-state-cache-good-pos start-point))
3361
3362 ;; SCAN!
3363 (cond
3364 ((memq strategy '(forward back-and-forward BOD))
3365 (setq res (c-remove-stale-state-cache start-point here here-bopl))
3366 (setq cache-pos (car res)
3367 scan-backward-pos (cadr res)
3368 cons-separated (car (cddr res))
3369 bopl-state (cadr (cddr res))) ; will be nil if (< here-bopl
3370 ; start-point)
3371 (if (and scan-backward-pos
3372 (or cons-separated (eq strategy 'forward))) ;scan-backward-pos
3373 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3374 (setq good-pos
3375 (c-append-to-state-cache cache-pos here))
3376 (setq c-state-cache-good-pos
3377 (if (and bopl-state
3378 (< good-pos (- here c-state-cache-too-far)))
3379 (c-state-cache-non-literal-place here-bopl bopl-state)
3380 good-pos)))
3381
3382 ((eq strategy 'backward)
3383 (setq res (c-remove-stale-state-cache-backwards here)
3384 good-pos (car res)
3385 scan-backward-pos (cadr res)
3386 scan-forward-p (car (cddr res)))
3387 (if scan-backward-pos
3388 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3389 (setq c-state-cache-good-pos
3390 (if scan-forward-p
3391 (c-append-to-state-cache good-pos here)
3392 good-pos)))
3393
3394 (t ; (eq strategy 'IN-LIT)
3395 (setq c-state-cache nil
3396 c-state-cache-good-pos nil))))
3397
3398 c-state-cache)
3399
3400 (defun c-invalidate-state-cache (here)
3401 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3402 ;;
3403 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3404 ;; of all parens in preprocessor constructs, except for any such construct
3405 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3406 ;; worrying further about macros and template delimiters.
3407 (if (eval-when-compile (memq 'category-properties c-emacs-features))
3408 ;; Emacs
3409 (c-with-<->-as-parens-suppressed
3410 (if (and c-state-old-cpp-beg
3411 (< c-state-old-cpp-beg here))
3412 (c-with-all-but-one-cpps-commented-out
3413 c-state-old-cpp-beg
3414 (min c-state-old-cpp-end here)
3415 (c-invalidate-state-cache-1 here))
3416 (c-with-cpps-commented-out
3417 (c-invalidate-state-cache-1 here))))
3418 ;; XEmacs
3419 (c-invalidate-state-cache-1 here)))
3420
3421 (defmacro c-state-maybe-marker (place marker)
3422 ;; If PLACE is non-nil, return a marker marking it, otherwise nil.
3423 ;; We (re)use MARKER.
3424 `(and ,place
3425 (or ,marker (setq ,marker (make-marker)))
3426 (set-marker ,marker ,place)))
3427
3428 (defun c-parse-state ()
3429 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3430 ;; description of the functionality and return value.
3431 ;;
3432 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3433 ;; of all parens in preprocessor constructs, except for any such construct
3434 ;; containing point. We can then call `c-parse-state-1' without worrying
3435 ;; further about macros and template delimiters.
3436 (let (here-cpp-beg here-cpp-end)
3437 (save-excursion
3438 (when (c-beginning-of-macro)
3439 (setq here-cpp-beg (point))
3440 (unless
3441 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3442 here-cpp-beg)
3443 (setq here-cpp-beg nil here-cpp-end nil))))
3444 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3445 ;; subsystem.
3446 (prog1
3447 (if (eval-when-compile (memq 'category-properties c-emacs-features))
3448 ;; Emacs
3449 (c-with-<->-as-parens-suppressed
3450 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3451 (c-with-all-but-one-cpps-commented-out
3452 here-cpp-beg here-cpp-end
3453 (c-parse-state-1))
3454 (c-with-cpps-commented-out
3455 (c-parse-state-1))))
3456 ;; XEmacs
3457 (c-parse-state-1))
3458 (setq c-state-old-cpp-beg
3459 (c-state-maybe-marker here-cpp-beg c-state-old-cpp-beg-marker)
3460 c-state-old-cpp-end
3461 (c-state-maybe-marker here-cpp-end c-state-old-cpp-end-marker)))))
3462
3463 ;; Debug tool to catch cache inconsistencies. This is called from
3464 ;; 000tests.el.
3465 (defvar c-debug-parse-state nil)
3466 (unless (fboundp 'c-real-parse-state)
3467 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3468 (cc-bytecomp-defun c-real-parse-state)
3469
3470 (defvar c-parse-state-point nil)
3471 (defvar c-parse-state-state nil)
3472 (make-variable-buffer-local 'c-parse-state-state)
3473 (defun c-record-parse-state-state ()
3474 (setq c-parse-state-point (point))
3475 (setq c-parse-state-state
3476 (mapcar
3477 (lambda (arg)
3478 (let ((val (symbol-value arg)))
3479 (cons arg
3480 (cond ((consp val) (copy-tree val))
3481 ((markerp val) (copy-marker val))
3482 (t val)))))
3483 '(c-state-cache
3484 c-state-cache-good-pos
3485 c-state-nonlit-pos-cache
3486 c-state-nonlit-pos-cache-limit
3487 c-state-semi-nonlit-pos-cache
3488 c-state-semi-nonlit-pos-cache-limit
3489 c-state-brace-pair-desert
3490 c-state-point-min
3491 c-state-point-min-lit-type
3492 c-state-point-min-lit-start
3493 c-state-min-scan-pos
3494 c-state-old-cpp-beg
3495 c-state-old-cpp-end
3496 c-parse-state-point))))
3497 (defun c-replay-parse-state-state ()
3498 (message
3499 (concat "(setq "
3500 (mapconcat
3501 (lambda (arg)
3502 (format "%s %s%s" (car arg)
3503 (if (atom (cdr arg)) "" "'")
3504 (if (markerp (cdr arg))
3505 (format "(copy-marker %s)" (marker-position (cdr arg)))
3506 (cdr arg))))
3507 c-parse-state-state " ")
3508 ")")))
3509
3510 (defun c-debug-parse-state-double-cons (state)
3511 (let (state-car conses-not-ok)
3512 (while state
3513 (setq state-car (car state)
3514 state (cdr state))
3515 (if (and (consp state-car)
3516 (consp (car state)))
3517 (setq conses-not-ok t)))
3518 conses-not-ok))
3519
3520 (defun c-debug-parse-state ()
3521 (let ((here (point)) (res1 (c-real-parse-state)) res2)
3522 (let ((c-state-cache nil)
3523 (c-state-cache-good-pos 1)
3524 (c-state-nonlit-pos-cache nil)
3525 (c-state-nonlit-pos-cache-limit 1)
3526 (c-state-brace-pair-desert nil)
3527 (c-state-point-min 1)
3528 (c-state-point-min-lit-type nil)
3529 (c-state-point-min-lit-start nil)
3530 (c-state-min-scan-pos 1)
3531 (c-state-old-cpp-beg nil)
3532 (c-state-old-cpp-end nil))
3533 (setq res2 (c-real-parse-state)))
3534 (unless (equal res1 res2)
3535 ;; The cache can actually go further back due to the ad-hoc way
3536 ;; the first paren is found, so try to whack off a bit of its
3537 ;; start before complaining.
3538 ;; (save-excursion
3539 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3540 ;; (c-beginning-of-defun-1)
3541 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3542 ;; (c-beginning-of-defun-1))
3543 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3544 ;; (message (concat "c-parse-state inconsistency at %s: "
3545 ;; "using cache: %s, from scratch: %s")
3546 ;; here res1 res2)))
3547 (message (concat "c-parse-state inconsistency at %s: "
3548 "using cache: %s, from scratch: %s")
3549 here res1 res2)
3550 (message "Old state:")
3551 (c-replay-parse-state-state))
3552
3553 (when (c-debug-parse-state-double-cons res1)
3554 (message "c-parse-state INVALIDITY at %s: %s"
3555 here res1)
3556 (message "Old state:")
3557 (c-replay-parse-state-state))
3558
3559 (c-record-parse-state-state)
3560 res2 ; res1 correct a cascading series of errors ASAP
3561 ))
3562
3563 (defun c-toggle-parse-state-debug (&optional arg)
3564 (interactive "P")
3565 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3566 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3567 'c-debug-parse-state
3568 'c-real-parse-state)))
3569 (c-keep-region-active)
3570 (message "c-debug-parse-state %sabled"
3571 (if c-debug-parse-state "en" "dis")))
3572 (when c-debug-parse-state
3573 (c-toggle-parse-state-debug 1))
3574
3575 \f
3576 (defun c-whack-state-before (bufpos paren-state)
3577 ;; Whack off any state information from PAREN-STATE which lies
3578 ;; before BUFPOS. Not destructive on PAREN-STATE.
3579 (let* ((newstate (list nil))
3580 (ptr newstate)
3581 car)
3582 (while paren-state
3583 (setq car (car paren-state)
3584 paren-state (cdr paren-state))
3585 (if (< (if (consp car) (car car) car) bufpos)
3586 (setq paren-state nil)
3587 (setcdr ptr (list car))
3588 (setq ptr (cdr ptr))))
3589 (cdr newstate)))
3590
3591 (defun c-whack-state-after (bufpos paren-state)
3592 ;; Whack off any state information from PAREN-STATE which lies at or
3593 ;; after BUFPOS. Not destructive on PAREN-STATE.
3594 (catch 'done
3595 (while paren-state
3596 (let ((car (car paren-state)))
3597 (if (consp car)
3598 ;; just check the car, because in a balanced brace
3599 ;; expression, it must be impossible for the corresponding
3600 ;; close brace to be before point, but the open brace to
3601 ;; be after.
3602 (if (<= bufpos (car car))
3603 nil ; whack it off
3604 (if (< bufpos (cdr car))
3605 ;; its possible that the open brace is before
3606 ;; bufpos, but the close brace is after. In that
3607 ;; case, convert this to a non-cons element. The
3608 ;; rest of the state is before bufpos, so we're
3609 ;; done.
3610 (throw 'done (cons (car car) (cdr paren-state)))
3611 ;; we know that both the open and close braces are
3612 ;; before bufpos, so we also know that everything else
3613 ;; on state is before bufpos.
3614 (throw 'done paren-state)))
3615 (if (<= bufpos car)
3616 nil ; whack it off
3617 ;; it's before bufpos, so everything else should too.
3618 (throw 'done paren-state)))
3619 (setq paren-state (cdr paren-state)))
3620 nil)))
3621
3622 (defun c-most-enclosing-brace (paren-state &optional bufpos)
3623 ;; Return the bufpos of the innermost enclosing open paren before
3624 ;; bufpos, or nil if none was found.
3625 (let (enclosingp)
3626 (or bufpos (setq bufpos 134217727))
3627 (while paren-state
3628 (setq enclosingp (car paren-state)
3629 paren-state (cdr paren-state))
3630 (if (or (consp enclosingp)
3631 (>= enclosingp bufpos))
3632 (setq enclosingp nil)
3633 (setq paren-state nil)))
3634 enclosingp))
3635
3636 (defun c-least-enclosing-brace (paren-state)
3637 ;; Return the bufpos of the outermost enclosing open paren, or nil
3638 ;; if none was found.
3639 (let (pos elem)
3640 (while paren-state
3641 (setq elem (car paren-state)
3642 paren-state (cdr paren-state))
3643 (if (integerp elem)
3644 (setq pos elem)))
3645 pos))
3646
3647 (defun c-safe-position (bufpos paren-state)
3648 ;; Return the closest "safe" position recorded on PAREN-STATE that
3649 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3650 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3651 ;; find the closest limit before a given limit that might be nil.
3652 ;;
3653 ;; A "safe" position is a position at or after a recorded open
3654 ;; paren, or after a recorded close paren. The returned position is
3655 ;; thus either the first position after a close brace, or the first
3656 ;; position after an enclosing paren, or at the enclosing paren in
3657 ;; case BUFPOS is immediately after it.
3658 (when bufpos
3659 (let (elem)
3660 (catch 'done
3661 (while paren-state
3662 (setq elem (car paren-state))
3663 (if (consp elem)
3664 (cond ((< (cdr elem) bufpos)
3665 (throw 'done (cdr elem)))
3666 ((< (car elem) bufpos)
3667 ;; See below.
3668 (throw 'done (min (1+ (car elem)) bufpos))))
3669 (if (< elem bufpos)
3670 ;; elem is the position at and not after the opening paren, so
3671 ;; we can go forward one more step unless it's equal to
3672 ;; bufpos. This is useful in some cases avoid an extra paren
3673 ;; level between the safe position and bufpos.
3674 (throw 'done (min (1+ elem) bufpos))))
3675 (setq paren-state (cdr paren-state)))))))
3676
3677 (defun c-beginning-of-syntax ()
3678 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3679 ;; goes to the closest previous point that is known to be outside
3680 ;; any string literal or comment. `c-state-cache' is used if it has
3681 ;; a position in the vicinity.
3682 (let* ((paren-state c-state-cache)
3683 elem
3684
3685 (pos (catch 'done
3686 ;; Note: Similar code in `c-safe-position'. The
3687 ;; difference is that we accept a safe position at
3688 ;; the point and don't bother to go forward past open
3689 ;; parens.
3690 (while paren-state
3691 (setq elem (car paren-state))
3692 (if (consp elem)
3693 (cond ((<= (cdr elem) (point))
3694 (throw 'done (cdr elem)))
3695 ((<= (car elem) (point))
3696 (throw 'done (car elem))))
3697 (if (<= elem (point))
3698 (throw 'done elem)))
3699 (setq paren-state (cdr paren-state)))
3700 (point-min))))
3701
3702 (if (> pos (- (point) 4000))
3703 (goto-char pos)
3704 ;; The position is far back. Try `c-beginning-of-defun-1'
3705 ;; (although we can't be entirely sure it will go to a position
3706 ;; outside a comment or string in current emacsen). FIXME:
3707 ;; Consult `syntax-ppss' here.
3708 (c-beginning-of-defun-1)
3709 (if (< (point) pos)
3710 (goto-char pos)))))
3711
3712 \f
3713 ;; Tools for scanning identifiers and other tokens.
3714
3715 (defun c-on-identifier ()
3716 "Return non-nil if the point is on or directly after an identifier.
3717 Keywords are recognized and not considered identifiers. If an
3718 identifier is detected, the returned value is its starting position.
3719 If an identifier ends at the point and another begins at it \(can only
3720 happen in Pike) then the point for the preceding one is returned.
3721
3722 Note that this function might do hidden buffer changes. See the
3723 comment at the start of cc-engine.el for more info."
3724
3725 ;; FIXME: Shouldn't this function handle "operator" in C++?
3726
3727 (save-excursion
3728 (skip-syntax-backward "w_")
3729
3730 (or
3731
3732 ;; Check for a normal (non-keyword) identifier.
3733 (and (looking-at c-symbol-start)
3734 (not (looking-at c-keywords-regexp))
3735 (point))
3736
3737 (when (c-major-mode-is 'pike-mode)
3738 ;; Handle the `<operator> syntax in Pike.
3739 (let ((pos (point)))
3740 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3741 (and (if (< (skip-chars-backward "`") 0)
3742 t
3743 (goto-char pos)
3744 (eq (char-after) ?\`))
3745 (looking-at c-symbol-key)
3746 (>= (match-end 0) pos)
3747 (point))))
3748
3749 ;; Handle the "operator +" syntax in C++.
3750 (when (and c-overloadable-operators-regexp
3751 (= (c-backward-token-2 0) 0))
3752
3753 (cond ((and (looking-at c-overloadable-operators-regexp)
3754 (or (not c-opt-op-identifier-prefix)
3755 (and (= (c-backward-token-2 1) 0)
3756 (looking-at c-opt-op-identifier-prefix))))
3757 (point))
3758
3759 ((save-excursion
3760 (and c-opt-op-identifier-prefix
3761 (looking-at c-opt-op-identifier-prefix)
3762 (= (c-forward-token-2 1) 0)
3763 (looking-at c-overloadable-operators-regexp)))
3764 (point))))
3765
3766 )))
3767
3768 (defsubst c-simple-skip-symbol-backward ()
3769 ;; If the point is at the end of a symbol then skip backward to the
3770 ;; beginning of it. Don't move otherwise. Return non-nil if point
3771 ;; moved.
3772 ;;
3773 ;; This function might do hidden buffer changes.
3774 (or (< (skip-syntax-backward "w_") 0)
3775 (and (c-major-mode-is 'pike-mode)
3776 ;; Handle the `<operator> syntax in Pike.
3777 (let ((pos (point)))
3778 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
3779 (< (skip-chars-backward "`") 0)
3780 (looking-at c-symbol-key)
3781 (>= (match-end 0) pos))
3782 t
3783 (goto-char pos)
3784 nil)))))
3785
3786 (defun c-beginning-of-current-token (&optional back-limit)
3787 ;; Move to the beginning of the current token. Do not move if not
3788 ;; in the middle of one. BACK-LIMIT may be used to bound the
3789 ;; backward search; if given it's assumed to be at the boundary
3790 ;; between two tokens. Return non-nil if the point is moved, nil
3791 ;; otherwise.
3792 ;;
3793 ;; This function might do hidden buffer changes.
3794 (let ((start (point)))
3795 (if (looking-at "\\w\\|\\s_")
3796 (skip-syntax-backward "w_" back-limit)
3797 (when (< (skip-syntax-backward ".()" back-limit) 0)
3798 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3799 (match-end 0))
3800 ;; `c-nonsymbol-token-regexp' should always match
3801 ;; since we've skipped backward over punctuation
3802 ;; or paren syntax, but consume one char in case
3803 ;; it doesn't so that we don't leave point before
3804 ;; some earlier incorrect token.
3805 (1+ (point)))))
3806 (if (<= pos start)
3807 (goto-char pos))))))
3808 (< (point) start)))
3809
3810 (defun c-end-of-current-token (&optional back-limit)
3811 ;; Move to the end of the current token. Do not move if not in the
3812 ;; middle of one. BACK-LIMIT may be used to bound the backward
3813 ;; search; if given it's assumed to be at the boundary between two
3814 ;; tokens. Return non-nil if the point is moved, nil otherwise.
3815 ;;
3816 ;; This function might do hidden buffer changes.
3817 (let ((start (point)))
3818 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3819 (skip-syntax-forward "w_"))
3820 ((< (skip-syntax-backward ".()" back-limit) 0)
3821 (while (progn
3822 (if (looking-at c-nonsymbol-token-regexp)
3823 (goto-char (match-end 0))
3824 ;; `c-nonsymbol-token-regexp' should always match since
3825 ;; we've skipped backward over punctuation or paren
3826 ;; syntax, but move forward in case it doesn't so that
3827 ;; we don't leave point earlier than we started with.
3828 (forward-char))
3829 (< (point) start)))))
3830 (> (point) start)))
3831
3832 (defconst c-jump-syntax-balanced
3833 (if (memq 'gen-string-delim c-emacs-features)
3834 "\\w\\|\\s_\\|\\s(\\|\\s)\\|\\s\"\\|\\s|"
3835 "\\w\\|\\s_\\|\\s(\\|\\s)\\|\\s\""))
3836
3837 (defconst c-jump-syntax-unbalanced
3838 (if (memq 'gen-string-delim c-emacs-features)
3839 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3840 "\\w\\|\\s_\\|\\s\""))
3841
3842 (defun c-forward-token-2 (&optional count balanced limit)
3843 "Move forward by tokens.
3844 A token is defined as all symbols and identifiers which aren't
3845 syntactic whitespace \(note that multicharacter tokens like \"==\" are
3846 treated properly). Point is always either left at the beginning of a
3847 token or not moved at all. COUNT specifies the number of tokens to
3848 move; a negative COUNT moves in the opposite direction. A COUNT of 0
3849 moves to the next token beginning only if not already at one. If
3850 BALANCED is true, move over balanced parens, otherwise move into them.
3851 Also, if BALANCED is true, never move out of an enclosing paren.
3852
3853 LIMIT sets the limit for the movement and defaults to the point limit.
3854 The case when LIMIT is set in the middle of a token, comment or macro
3855 is handled correctly, i.e. the point won't be left there.
3856
3857 Return the number of tokens left to move \(positive or negative). If
3858 BALANCED is true, a move over a balanced paren counts as one. Note
3859 that if COUNT is 0 and no appropriate token beginning is found, 1 will
3860 be returned. Thus, a return value of 0 guarantees that point is at
3861 the requested position and a return value less \(without signs) than
3862 COUNT guarantees that point is at the beginning of some token.
3863
3864 Note that this function might do hidden buffer changes. See the
3865 comment at the start of cc-engine.el for more info."
3866
3867 (or count (setq count 1))
3868 (if (< count 0)
3869 (- (c-backward-token-2 (- count) balanced limit))
3870
3871 (let ((jump-syntax (if balanced
3872 c-jump-syntax-balanced
3873 c-jump-syntax-unbalanced))
3874 (last (point))
3875 (prev (point)))
3876
3877 (if (zerop count)
3878 ;; If count is zero we should jump if in the middle of a token.
3879 (c-end-of-current-token))
3880
3881 (save-restriction
3882 (if limit (narrow-to-region (point-min) limit))
3883 (if (/= (point)
3884 (progn (c-forward-syntactic-ws) (point)))
3885 ;; Skip whitespace. Count this as a move if we did in
3886 ;; fact move.
3887 (setq count (max (1- count) 0)))
3888
3889 (if (eobp)
3890 ;; Moved out of bounds. Make sure the returned count isn't zero.
3891 (progn
3892 (if (zerop count) (setq count 1))
3893 (goto-char last))
3894
3895 ;; Use `condition-case' to avoid having the limit tests
3896 ;; inside the loop.
3897 (condition-case nil
3898 (while (and
3899 (> count 0)
3900 (progn
3901 (setq last (point))
3902 (cond ((looking-at jump-syntax)
3903 (goto-char (scan-sexps (point) 1))
3904 t)
3905 ((looking-at c-nonsymbol-token-regexp)
3906 (goto-char (match-end 0))
3907 t)
3908 ;; `c-nonsymbol-token-regexp' above should always
3909 ;; match if there are correct tokens. Try to
3910 ;; widen to see if the limit was set in the
3911 ;; middle of one, else fall back to treating
3912 ;; the offending thing as a one character token.
3913 ((and limit
3914 (save-restriction
3915 (widen)
3916 (looking-at c-nonsymbol-token-regexp)))
3917 nil)
3918 (t
3919 (forward-char)
3920 t))))
3921 (c-forward-syntactic-ws)
3922 (setq prev last
3923 count (1- count)))
3924 (error (goto-char last)))
3925
3926 (when (eobp)
3927 (goto-char prev)
3928 (setq count (1+ count)))))
3929
3930 count)))
3931
3932 (defun c-backward-token-2 (&optional count balanced limit)
3933 "Move backward by tokens.
3934 See `c-forward-token-2' for details."
3935
3936 (or count (setq count 1))
3937 (if (< count 0)
3938 (- (c-forward-token-2 (- count) balanced limit))
3939
3940 (or limit (setq limit (point-min)))
3941 (let ((jump-syntax (if balanced
3942 c-jump-syntax-balanced
3943 c-jump-syntax-unbalanced))
3944 (last (point)))
3945
3946 (if (zerop count)
3947 ;; The count is zero so try to skip to the beginning of the
3948 ;; current token.
3949 (if (> (point)
3950 (progn (c-beginning-of-current-token) (point)))
3951 (if (< (point) limit)
3952 ;; The limit is inside the same token, so return 1.
3953 (setq count 1))
3954
3955 ;; We're not in the middle of a token. If there's
3956 ;; whitespace after the point then we must move backward,
3957 ;; so set count to 1 in that case.
3958 (and (looking-at c-syntactic-ws-start)
3959 ;; If we're looking at a '#' that might start a cpp
3960 ;; directive then we have to do a more elaborate check.
3961 (or (/= (char-after) ?#)
3962 (not c-opt-cpp-prefix)
3963 (save-excursion
3964 (and (= (point)
3965 (progn (beginning-of-line)
3966 (looking-at "[ \t]*")
3967 (match-end 0)))
3968 (or (bobp)
3969 (progn (backward-char)
3970 (not (eq (char-before) ?\\)))))))
3971 (setq count 1))))
3972
3973 ;; Use `condition-case' to avoid having to check for buffer
3974 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3975 (condition-case nil
3976 (while (and
3977 (> count 0)
3978 (progn
3979 (c-backward-syntactic-ws)
3980 (backward-char)
3981 (if (looking-at jump-syntax)
3982 (goto-char (scan-sexps (1+ (point)) -1))
3983 ;; This can be very inefficient if there's a long
3984 ;; sequence of operator tokens without any separation.
3985 ;; That doesn't happen in practice, anyway.
3986 (c-beginning-of-current-token))
3987 (>= (point) limit)))
3988 (setq last (point)
3989 count (1- count)))
3990 (error (goto-char last)))
3991
3992 (if (< (point) limit)
3993 (goto-char last))
3994
3995 count)))
3996
3997 (defun c-forward-token-1 (&optional count balanced limit)
3998 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3999 tokens like \"==\" as single tokens, i.e. all sequences of symbol
4000 characters are jumped over character by character. This function is
4001 for compatibility only; it's only a wrapper over `c-forward-token-2'."
4002 (let ((c-nonsymbol-token-regexp "\\s."))
4003 (c-forward-token-2 count balanced limit)))
4004
4005 (defun c-backward-token-1 (&optional count balanced limit)
4006 "Like `c-backward-token-2' but doesn't treat multicharacter operator
4007 tokens like \"==\" as single tokens, i.e. all sequences of symbol
4008 characters are jumped over character by character. This function is
4009 for compatibility only; it's only a wrapper over `c-backward-token-2'."
4010 (let ((c-nonsymbol-token-regexp "\\s."))
4011 (c-backward-token-2 count balanced limit)))
4012
4013 \f
4014 ;; Tools for doing searches restricted to syntactically relevant text.
4015
4016 (defun c-syntactic-re-search-forward (regexp &optional bound noerror
4017 paren-level not-inside-token
4018 lookbehind-submatch)
4019 "Like `re-search-forward', but only report matches that are found
4020 in syntactically significant text. I.e. matches in comments, macros
4021 or string literals are ignored. The start point is assumed to be
4022 outside any comment, macro or string literal, or else the content of
4023 that region is taken as syntactically significant text.
4024
4025 If PAREN-LEVEL is non-nil, an additional restriction is added to
4026 ignore matches in nested paren sexps. The search will also not go
4027 outside the current list sexp, which has the effect that if the point
4028 should be moved to BOUND when no match is found \(i.e. NOERROR is
4029 neither nil nor t), then it will be at the closing paren if the end of
4030 the current list sexp is encountered first.
4031
4032 If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
4033 ignored. Things like multicharacter operators and special symbols
4034 \(e.g. \"`()\" in Pike) are handled but currently not floating point
4035 constants.
4036
4037 If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
4038 subexpression in REGEXP. The end of that submatch is used as the
4039 position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
4040 isn't used or if that subexpression didn't match then the start
4041 position of the whole match is used instead. The \"look behind\"
4042 subexpression is never tested before the starting position, so it
4043 might be a good idea to include \\=\\= as a match alternative in it.
4044
4045 Optimization note: Matches might be missed if the \"look behind\"
4046 subexpression can match the end of nonwhite syntactic whitespace,
4047 i.e. the end of comments or cpp directives. This since the function
4048 skips over such things before resuming the search. It's on the other
4049 hand not safe to assume that the \"look behind\" subexpression never
4050 matches syntactic whitespace.
4051
4052 Bug: Unbalanced parens inside cpp directives are currently not handled
4053 correctly \(i.e. they don't get ignored as they should) when
4054 PAREN-LEVEL is set.
4055
4056 Note that this function might do hidden buffer changes. See the
4057 comment at the start of cc-engine.el for more info."
4058
4059 (or bound (setq bound (point-max)))
4060 (if paren-level (setq paren-level -1))
4061
4062 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
4063
4064 (let ((start (point))
4065 tmp
4066 ;; Start position for the last search.
4067 search-pos
4068 ;; The `parse-partial-sexp' state between the start position
4069 ;; and the point.
4070 state
4071 ;; The current position after the last state update. The next
4072 ;; `parse-partial-sexp' continues from here.
4073 (state-pos (point))
4074 ;; The position at which to check the state and the state
4075 ;; there. This is separate from `state-pos' since we might
4076 ;; need to back up before doing the next search round.
4077 check-pos check-state
4078 ;; Last position known to end a token.
4079 (last-token-end-pos (point-min))
4080 ;; Set when a valid match is found.
4081 found)
4082
4083 (condition-case err
4084 (while
4085 (and
4086 (progn
4087 (setq search-pos (point))
4088 (re-search-forward regexp bound noerror))
4089
4090 (progn
4091 (setq state (parse-partial-sexp
4092 state-pos (match-beginning 0) paren-level nil state)
4093 state-pos (point))
4094 (if (setq check-pos (and lookbehind-submatch
4095 (or (not paren-level)
4096 (>= (car state) 0))
4097 (match-end lookbehind-submatch)))
4098 (setq check-state (parse-partial-sexp
4099 state-pos check-pos paren-level nil state))
4100 (setq check-pos state-pos
4101 check-state state))
4102
4103 ;; NOTE: If we got a look behind subexpression and get
4104 ;; an insignificant match in something that isn't
4105 ;; syntactic whitespace (i.e. strings or in nested
4106 ;; parentheses), then we can never skip more than a
4107 ;; single character from the match start position
4108 ;; (i.e. `state-pos' here) before continuing the
4109 ;; search. That since the look behind subexpression
4110 ;; might match the end of the insignificant region in
4111 ;; the next search.
4112
4113 (cond
4114 ((elt check-state 7)
4115 ;; Match inside a line comment. Skip to eol. Use
4116 ;; `re-search-forward' instead of `skip-chars-forward' to get
4117 ;; the right bound behavior.
4118 (re-search-forward "[\n\r]" bound noerror))
4119
4120 ((elt check-state 4)
4121 ;; Match inside a block comment. Skip to the '*/'.
4122 (search-forward "*/" bound noerror))
4123
4124 ((and (not (elt check-state 5))
4125 (eq (char-before check-pos) ?/)
4126 (not (c-get-char-property (1- check-pos) 'syntax-table))
4127 (memq (char-after check-pos) '(?/ ?*)))
4128 ;; Match in the middle of the opener of a block or line
4129 ;; comment.
4130 (if (= (char-after check-pos) ?/)
4131 (re-search-forward "[\n\r]" bound noerror)
4132 (search-forward "*/" bound noerror)))
4133
4134 ;; The last `parse-partial-sexp' above might have
4135 ;; stopped short of the real check position if the end
4136 ;; of the current sexp was encountered in paren-level
4137 ;; mode. The checks above are always false in that
4138 ;; case, and since they can do better skipping in
4139 ;; lookbehind-submatch mode, we do them before
4140 ;; checking the paren level.
4141
4142 ((and paren-level
4143 (/= (setq tmp (car check-state)) 0))
4144 ;; Check the paren level first since we're short of the
4145 ;; syntactic checking position if the end of the
4146 ;; current sexp was encountered by `parse-partial-sexp'.
4147 (if (> tmp 0)
4148
4149 ;; Inside a nested paren sexp.
4150 (if lookbehind-submatch
4151 ;; See the NOTE above.
4152 (progn (goto-char state-pos) t)
4153 ;; Skip out of the paren quickly.
4154 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4155 state-pos (point)))
4156
4157 ;; Have exited the current paren sexp.
4158 (if noerror
4159 (progn
4160 ;; The last `parse-partial-sexp' call above
4161 ;; has left us just after the closing paren
4162 ;; in this case, so we can modify the bound
4163 ;; to leave the point at the right position
4164 ;; upon return.
4165 (setq bound (1- (point)))
4166 nil)
4167 (signal 'search-failed (list regexp)))))
4168
4169 ((setq tmp (elt check-state 3))
4170 ;; Match inside a string.
4171 (if (or lookbehind-submatch
4172 (not (integerp tmp)))
4173 ;; See the NOTE above.
4174 (progn (goto-char state-pos) t)
4175 ;; Skip to the end of the string before continuing.
4176 (let ((ender (make-string 1 tmp)) (continue t))
4177 (while (if (search-forward ender bound noerror)
4178 (progn
4179 (setq state (parse-partial-sexp
4180 state-pos (point) nil nil state)
4181 state-pos (point))
4182 (elt state 3))
4183 (setq continue nil)))
4184 continue)))
4185
4186 ((save-excursion
4187 (save-match-data
4188 (c-beginning-of-macro start)))
4189 ;; Match inside a macro. Skip to the end of it.
4190 (c-end-of-macro)
4191 (cond ((<= (point) bound) t)
4192 (noerror nil)
4193 (t (signal 'search-failed (list regexp)))))
4194
4195 ((and not-inside-token
4196 (or (< check-pos last-token-end-pos)
4197 (< check-pos
4198 (save-excursion
4199 (goto-char check-pos)
4200 (save-match-data
4201 (c-end-of-current-token last-token-end-pos))
4202 (setq last-token-end-pos (point))))))
4203 ;; Inside a token.
4204 (if lookbehind-submatch
4205 ;; See the NOTE above.
4206 (goto-char state-pos)
4207 (goto-char (min last-token-end-pos bound))))
4208
4209 (t
4210 ;; A real match.
4211 (setq found t)
4212 nil)))
4213
4214 ;; Should loop to search again, but take care to avoid
4215 ;; looping on the same spot.
4216 (or (/= search-pos (point))
4217 (if (= (point) bound)
4218 (if noerror
4219 nil
4220 (signal 'search-failed (list regexp)))
4221 (forward-char)
4222 t))))
4223
4224 (error
4225 (goto-char start)
4226 (signal (car err) (cdr err))))
4227
4228 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
4229
4230 (if found
4231 (progn
4232 (goto-char (match-end 0))
4233 (match-end 0))
4234
4235 ;; Search failed. Set point as appropriate.
4236 (if (eq noerror t)
4237 (goto-char start)
4238 (goto-char bound))
4239 nil)))
4240
4241 (defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4242
4243 (defsubst c-ssb-lit-begin ()
4244 ;; Return the start of the literal point is in, or nil.
4245 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4246 ;; bound in the caller.
4247
4248 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4249 ;; if it's outside comments and strings.
4250 (save-excursion
4251 (let ((pos (point)) safe-pos state)
4252 ;; Pick a safe position as close to the point as possible.
4253 ;;
4254 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4255 ;; position.
4256
4257 (while (and safe-pos-list
4258 (> (car safe-pos-list) (point)))
4259 (setq safe-pos-list (cdr safe-pos-list)))
4260 (unless (setq safe-pos (car-safe safe-pos-list))
4261 (setq safe-pos (max (or (c-safe-position
4262 (point) (or c-state-cache
4263 (c-parse-state)))
4264 0)
4265 (point-min))
4266 safe-pos-list (list safe-pos)))
4267
4268 ;; Cache positions along the way to use if we have to back up more. We
4269 ;; cache every closing paren on the same level. If the paren cache is
4270 ;; relevant in this region then we're typically already on the same
4271 ;; level as the target position. Note that we might cache positions
4272 ;; after opening parens in case safe-pos is in a nested list. That's
4273 ;; both uncommon and harmless.
4274 (while (progn
4275 (setq state (parse-partial-sexp
4276 safe-pos pos 0))
4277 (< (point) pos))
4278 (setq safe-pos (point)
4279 safe-pos-list (cons safe-pos safe-pos-list)))
4280
4281 ;; If the state contains the start of the containing sexp we cache that
4282 ;; position too, so that parse-partial-sexp in the next run has a bigger
4283 ;; chance of starting at the same level as the target position and thus
4284 ;; will get more good safe positions into the list.
4285 (if (elt state 1)
4286 (setq safe-pos (1+ (elt state 1))
4287 safe-pos-list (cons safe-pos safe-pos-list)))
4288
4289 (if (or (elt state 3) (elt state 4))
4290 ;; Inside string or comment. Continue search at the
4291 ;; beginning of it.
4292 (elt state 8)))))
4293
4294 (defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4295 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4296 i.e. don't stop at positions inside syntactic whitespace or string
4297 literals. Preprocessor directives are also ignored, with the exception
4298 of the one that the point starts within, if any. If LIMIT is given,
4299 it's assumed to be at a syntactically relevant position.
4300
4301 If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4302 sexps, and the search will also not go outside the current paren sexp.
4303 However, if LIMIT or the buffer limit is reached inside a nested paren
4304 then the point will be left at the limit.
4305
4306 Non-nil is returned if the point moved, nil otherwise.
4307
4308 Note that this function might do hidden buffer changes. See the
4309 comment at the start of cc-engine.el for more info."
4310
4311 (let ((start (point))
4312 state-2
4313 ;; A list of syntactically relevant positions in descending
4314 ;; order. It's used to avoid scanning repeatedly over
4315 ;; potentially large regions with `parse-partial-sexp' to verify
4316 ;; each position. Used in `c-ssb-lit-begin'
4317 safe-pos-list
4318 ;; The result from `c-beginning-of-macro' at the start position or the
4319 ;; start position itself if it isn't within a macro. Evaluated on
4320 ;; demand.
4321 start-macro-beg
4322 ;; The earliest position after the current one with the same paren
4323 ;; level. Used only when `paren-level' is set.
4324 lit-beg
4325 (paren-level-pos (point)))
4326
4327 (while
4328 (progn
4329 ;; The next loop "tries" to find the end point each time round,
4330 ;; loops when it hasn't succeeded.
4331 (while
4332 (and
4333 (let ((pos (point)))
4334 (while (and
4335 (< (skip-chars-backward skip-chars limit) 0)
4336 ;; Don't stop inside a literal.
4337 (when (setq lit-beg (c-ssb-lit-begin))
4338 (goto-char lit-beg)
4339 t)))
4340 (< (point) pos))
4341
4342 (let ((pos (point)) state-2 pps-end-pos)
4343
4344 (cond
4345 ((and paren-level
4346 (save-excursion
4347 (setq state-2 (parse-partial-sexp
4348 pos paren-level-pos -1)
4349 pps-end-pos (point))
4350 (/= (car state-2) 0)))
4351 ;; Not at the right level.
4352
4353 (if (and (< (car state-2) 0)
4354 ;; We stop above if we go out of a paren.
4355 ;; Now check whether it precedes or is
4356 ;; nested in the starting sexp.
4357 (save-excursion
4358 (setq state-2
4359 (parse-partial-sexp
4360 pps-end-pos paren-level-pos
4361 nil nil state-2))
4362 (< (car state-2) 0)))
4363
4364 ;; We've stopped short of the starting position
4365 ;; so the hit was inside a nested list. Go up
4366 ;; until we are at the right level.
4367 (condition-case nil
4368 (progn
4369 (goto-char (scan-lists pos -1
4370 (- (car state-2))))
4371 (setq paren-level-pos (point))
4372 (if (and limit (>= limit paren-level-pos))
4373 (progn
4374 (goto-char limit)
4375 nil)
4376 t))
4377 (error
4378 (goto-char (or limit (point-min)))
4379 nil))
4380
4381 ;; The hit was outside the list at the start
4382 ;; position. Go to the start of the list and exit.
4383 (goto-char (1+ (elt state-2 1)))
4384 nil))
4385
4386 ((c-beginning-of-macro limit)
4387 ;; Inside a macro.
4388 (if (< (point)
4389 (or start-macro-beg
4390 (setq start-macro-beg
4391 (save-excursion
4392 (goto-char start)
4393 (c-beginning-of-macro limit)
4394 (point)))))
4395 t
4396
4397 ;; It's inside the same macro we started in so it's
4398 ;; a relevant match.
4399 (goto-char pos)
4400 nil))))))
4401
4402 (> (point)
4403 (progn
4404 ;; Skip syntactic ws afterwards so that we don't stop at the
4405 ;; end of a comment if `skip-chars' is something like "^/".
4406 (c-backward-syntactic-ws)
4407 (point)))))
4408
4409 ;; We might want to extend this with more useful return values in
4410 ;; the future.
4411 (/= (point) start)))
4412
4413 ;; The following is an alternative implementation of
4414 ;; `c-syntactic-skip-backward' that uses backward movement to keep
4415 ;; track of the syntactic context. It turned out to be generally
4416 ;; slower than the one above which uses forward checks from earlier
4417 ;; safe positions.
4418 ;;
4419 ;;(defconst c-ssb-stop-re
4420 ;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4421 ;; ;; stop at to avoid going into comments and literals.
4422 ;; (concat
4423 ;; ;; Match comment end syntax and string literal syntax. Also match
4424 ;; ;; '/' for block comment endings (not covered by comment end
4425 ;; ;; syntax).
4426 ;; "\\s>\\|/\\|\\s\""
4427 ;; (if (memq 'gen-string-delim c-emacs-features)
4428 ;; "\\|\\s|"
4429 ;; "")
4430 ;; (if (memq 'gen-comment-delim c-emacs-features)
4431 ;; "\\|\\s!"
4432 ;; "")))
4433 ;;
4434 ;;(defconst c-ssb-stop-paren-re
4435 ;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4436 ;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4437 ;;
4438 ;;(defconst c-ssb-sexp-end-re
4439 ;; ;; Regexp matching the ending syntax of a complex sexp.
4440 ;; (concat c-string-limit-regexp "\\|\\s)"))
4441 ;;
4442 ;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4443 ;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4444 ;;i.e. don't stop at positions inside syntactic whitespace or string
4445 ;;literals. Preprocessor directives are also ignored. However, if the
4446 ;;point is within a comment, string literal or preprocessor directory to
4447 ;;begin with, its contents is treated as syntactically relevant chars.
4448 ;;If LIMIT is given, it limits the backward search and the point will be
4449 ;;left there if no earlier position is found.
4450 ;;
4451 ;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4452 ;;sexps, and the search will also not go outside the current paren sexp.
4453 ;;However, if LIMIT or the buffer limit is reached inside a nested paren
4454 ;;then the point will be left at the limit.
4455 ;;
4456 ;;Non-nil is returned if the point moved, nil otherwise.
4457 ;;
4458 ;;Note that this function might do hidden buffer changes. See the
4459 ;;comment at the start of cc-engine.el for more info."
4460 ;;
4461 ;; (save-restriction
4462 ;; (when limit
4463 ;; (narrow-to-region limit (point-max)))
4464 ;;
4465 ;; (let ((start (point)))
4466 ;; (catch 'done
4467 ;; (while (let ((last-pos (point))
4468 ;; (stop-pos (progn
4469 ;; (skip-chars-backward skip-chars)
4470 ;; (point))))
4471 ;;
4472 ;; ;; Skip back over the same region as
4473 ;; ;; `skip-chars-backward' above, but keep to
4474 ;; ;; syntactically relevant positions.
4475 ;; (goto-char last-pos)
4476 ;; (while (and
4477 ;; ;; `re-search-backward' with a single char regexp
4478 ;; ;; should be fast.
4479 ;; (re-search-backward
4480 ;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4481 ;; stop-pos 'move)
4482 ;;
4483 ;; (progn
4484 ;; (cond
4485 ;; ((looking-at "\\s(")
4486 ;; ;; `paren-level' is set and we've found the
4487 ;; ;; start of the containing paren.
4488 ;; (forward-char)
4489 ;; (throw 'done t))
4490 ;;
4491 ;; ((looking-at c-ssb-sexp-end-re)
4492 ;; ;; We're at the end of a string literal or paren
4493 ;; ;; sexp (if `paren-level' is set).
4494 ;; (forward-char)
4495 ;; (condition-case nil
4496 ;; (c-backward-sexp)
4497 ;; (error
4498 ;; (goto-char limit)
4499 ;; (throw 'done t))))
4500 ;;
4501 ;; (t
4502 ;; (forward-char)
4503 ;; ;; At the end of some syntactic ws or possibly
4504 ;; ;; after a plain '/' operator.
4505 ;; (let ((pos (point)))
4506 ;; (c-backward-syntactic-ws)
4507 ;; (if (= pos (point))
4508 ;; ;; Was a plain '/' operator. Go past it.
4509 ;; (backward-char)))))
4510 ;;
4511 ;; (> (point) stop-pos))))
4512 ;;
4513 ;; ;; Now the point is either at `stop-pos' or at some
4514 ;; ;; position further back if `stop-pos' was at a
4515 ;; ;; syntactically irrelevant place.
4516 ;;
4517 ;; ;; Skip additional syntactic ws so that we don't stop
4518 ;; ;; at the end of a comment if `skip-chars' is
4519 ;; ;; something like "^/".
4520 ;; (c-backward-syntactic-ws)
4521 ;;
4522 ;; (< (point) stop-pos))))
4523 ;;
4524 ;; ;; We might want to extend this with more useful return values
4525 ;; ;; in the future.
4526 ;; (/= (point) start))))
4527
4528 \f
4529 ;; Tools for handling comments and string literals.
4530
4531 (defun c-in-literal (&optional lim detect-cpp)
4532 "Return the type of literal point is in, if any.
4533 The return value is `c' if in a C-style comment, `c++' if in a C++
4534 style comment, `string' if in a string literal, `pound' if DETECT-CPP
4535 is non-nil and in a preprocessor line, or nil if somewhere else.
4536 Optional LIM is used as the backward limit of the search. If omitted,
4537 or nil, `c-beginning-of-defun' is used.
4538
4539 The last point calculated is cached if the cache is enabled, i.e. if
4540 `c-in-literal-cache' is bound to a two element vector.
4541
4542 Note that this function might do hidden buffer changes. See the
4543 comment at the start of cc-engine.el for more info."
4544 (save-restriction
4545 (widen)
4546 (let* ((safe-place (c-state-semi-safe-place (point)))
4547 (lit (c-state-pp-to-literal safe-place (point))))
4548 (or (cadr lit)
4549 (and detect-cpp
4550 (save-excursion (c-beginning-of-macro))
4551 'pound)))))
4552
4553 (defun c-literal-limits (&optional lim near not-in-delimiter)
4554 "Return a cons of the beginning and end positions of the comment or
4555 string surrounding point (including both delimiters), or nil if point
4556 isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4557 to start parsing from. If NEAR is non-nil, then the limits of any
4558 literal next to point is returned. \"Next to\" means there's only
4559 spaces and tabs between point and the literal. The search for such a
4560 literal is done first in forward direction. If NOT-IN-DELIMITER is
4561 non-nil, the case when point is inside a starting delimiter won't be
4562 recognized. This only has effect for comments which have starting
4563 delimiters with more than one character.
4564
4565 Note that this function might do hidden buffer changes. See the
4566 comment at the start of cc-engine.el for more info."
4567
4568 (save-excursion
4569 (let* ((pos (point))
4570 (lim (or lim (c-state-semi-safe-place pos)))
4571 (pp-to-lit (save-restriction
4572 (widen)
4573 (c-state-pp-to-literal lim pos not-in-delimiter)))
4574 (state (car pp-to-lit))
4575 (lit-limits (car (cddr pp-to-lit))))
4576
4577 (cond
4578 (lit-limits)
4579
4580 (near
4581 (goto-char pos)
4582 ;; Search forward for a literal.
4583 (skip-chars-forward " \t")
4584 (cond
4585 ((looking-at c-string-limit-regexp) ; String.
4586 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4587 (point-max))))
4588
4589 ((looking-at c-comment-start-regexp) ; Line or block comment.
4590 (cons (point) (progn (c-forward-single-comment) (point))))
4591
4592 (t
4593 ;; Search backward.
4594 (skip-chars-backward " \t")
4595
4596 (let ((end (point)) beg)
4597 (cond
4598 ((save-excursion
4599 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4600 (setq beg (c-safe (c-backward-sexp 1) (point))))
4601
4602 ((and (c-safe (forward-char -2) t)
4603 (looking-at "*/"))
4604 ;; Block comment. Due to the nature of line
4605 ;; comments, they will always be covered by the
4606 ;; normal case above.
4607 (goto-char end)
4608 (c-backward-single-comment)
4609 ;; If LIM is bogus, beg will be bogus.
4610 (setq beg (point))))
4611
4612 (if beg (cons beg end))))))
4613 ))))
4614
4615 ;; In case external callers use this; it did have a docstring.
4616 (defalias 'c-literal-limits-fast 'c-literal-limits)
4617
4618 (defun c-collect-line-comments (range)
4619 "If the argument is a cons of two buffer positions (such as returned by
4620 `c-literal-limits'), and that range contains a C++ style line comment,
4621 then an extended range is returned that contains all adjacent line
4622 comments (i.e. all comments that starts in the same column with no
4623 empty lines or non-whitespace characters between them). Otherwise the
4624 argument is returned.
4625
4626 Note that this function might do hidden buffer changes. See the
4627 comment at the start of cc-engine.el for more info."
4628
4629 (save-excursion
4630 (condition-case nil
4631 (if (and (consp range) (progn
4632 (goto-char (car range))
4633 (looking-at c-line-comment-starter)))
4634 (let ((col (current-column))
4635 (beg (point))
4636 (bopl (c-point 'bopl))
4637 (end (cdr range)))
4638 ;; Got to take care in the backward direction to handle
4639 ;; comments which are preceded by code.
4640 (while (and (c-backward-single-comment)
4641 (>= (point) bopl)
4642 (looking-at c-line-comment-starter)
4643 (= col (current-column)))
4644 (setq beg (point)
4645 bopl (c-point 'bopl)))
4646 (goto-char end)
4647 (while (and (progn (skip-chars-forward " \t")
4648 (looking-at c-line-comment-starter))
4649 (= col (current-column))
4650 (prog1 (zerop (forward-line 1))
4651 (setq end (point)))))
4652 (cons beg end))
4653 range)
4654 (error range))))
4655
4656 (defun c-literal-type (range)
4657 "Convenience function that given the result of `c-literal-limits',
4658 returns nil or the type of literal that the range surrounds, one
4659 of the symbols `c', `c++' or `string'. It's much faster than using
4660 `c-in-literal' and is intended to be used when you need both the
4661 type of a literal and its limits.
4662
4663 Note that this function might do hidden buffer changes. See the
4664 comment at the start of cc-engine.el for more info."
4665
4666 (if (consp range)
4667 (save-excursion
4668 (goto-char (car range))
4669 (cond ((looking-at c-string-limit-regexp) 'string)
4670 ((or (looking-at "//") ; c++ line comment
4671 (and (looking-at "\\s<") ; comment starter
4672 (looking-at "#"))) ; awk comment.
4673 'c++)
4674 (t 'c))) ; Assuming the range is valid.
4675 range))
4676
4677 (defsubst c-determine-limit-get-base (start try-size)
4678 ;; Get a "safe place" approximately TRY-SIZE characters before START.
4679 ;; This doesn't preserve point.
4680 (let* ((pos (max (- start try-size) (point-min)))
4681 (base (c-state-semi-safe-place pos))
4682 (s (parse-partial-sexp base pos)))
4683 (if (or (nth 4 s) (nth 3 s)) ; comment or string
4684 (nth 8 s)
4685 (point))))
4686
4687 (defun c-determine-limit (how-far-back &optional start try-size)
4688 ;; Return a buffer position HOW-FAR-BACK non-literal characters from START
4689 ;; (default point). This is done by going back further in the buffer then
4690 ;; searching forward for literals. The position found won't be in a
4691 ;; literal. We start searching for the sought position TRY-SIZE (default
4692 ;; twice HOW-FAR-BACK) bytes back from START. This function must be fast.
4693 ;; :-)
4694 (save-excursion
4695 (let* ((start (or start (point)))
4696 (try-size (or try-size (* 2 how-far-back)))
4697 (base (c-determine-limit-get-base start try-size))
4698 (pos base)
4699
4700 (s (parse-partial-sexp pos pos)) ; null state.
4701 stack elt size
4702 (count 0))
4703 (while (< pos start)
4704 ;; Move forward one literal each time round this loop.
4705 ;; Move forward to the start of a comment or string.
4706 (setq s (parse-partial-sexp
4707 pos
4708 start
4709 nil ; target-depth
4710 nil ; stop-before
4711 s ; state
4712 'syntax-table)) ; stop-comment
4713
4714 ;; Gather details of the non-literal-bit - starting pos and size.
4715 (setq size (- (if (or (nth 4 s) (nth 3 s))
4716 (nth 8 s)
4717 (point))
4718 pos))
4719 (if (> size 0)
4720 (setq stack (cons (cons pos size) stack)))
4721
4722 ;; Move forward to the end of the comment/string.
4723 (if (or (nth 4 s) (nth 3 s))
4724 (setq s (parse-partial-sexp
4725 (point)
4726 start
4727 nil ; target-depth
4728 nil ; stop-before
4729 s ; state
4730 'syntax-table))) ; stop-comment
4731 (setq pos (point)))
4732
4733 ;; Now try and find enough non-literal characters recorded on the stack.
4734 ;; Go back one recorded literal each time round this loop.
4735 (while (and (< count how-far-back)
4736 stack)
4737 (setq elt (car stack)
4738 stack (cdr stack))
4739 (setq count (+ count (cdr elt))))
4740
4741 ;; Have we found enough yet?
4742 (cond
4743 ((>= count how-far-back)
4744 (+ (car elt) (- count how-far-back)))
4745 ((eq base (point-min))
4746 (point-min))
4747 (t
4748 (c-determine-limit (- how-far-back count) base try-size))))))
4749
4750 (defun c-determine-+ve-limit (how-far &optional start-pos)
4751 ;; Return a buffer position about HOW-FAR non-literal characters forward
4752 ;; from START-POS (default point), which must not be inside a literal.
4753 (save-excursion
4754 (let ((pos (or start-pos (point)))
4755 (count how-far)
4756 (s (parse-partial-sexp (point) (point)))) ; null state
4757 (while (and (not (eobp))
4758 (> count 0))
4759 ;; Scan over counted characters.
4760 (setq s (parse-partial-sexp
4761 pos
4762 (min (+ pos count) (point-max))
4763 nil ; target-depth
4764 nil ; stop-before
4765 s ; state
4766 'syntax-table)) ; stop-comment
4767 (setq count (- count (- (point) pos) 1)
4768 pos (point))
4769 ;; Scan over literal characters.
4770 (if (nth 8 s)
4771 (setq s (parse-partial-sexp
4772 pos
4773 (point-max)
4774 nil ; target-depth
4775 nil ; stop-before
4776 s ; state
4777 'syntax-table) ; stop-comment
4778 pos (point))))
4779 (point))))
4780
4781 \f
4782 ;; `c-find-decl-spots' and accompanying stuff.
4783
4784 ;; Variables used in `c-find-decl-spots' to cache the search done for
4785 ;; the first declaration in the last call. When that function starts,
4786 ;; it needs to back up over syntactic whitespace to look at the last
4787 ;; token before the region being searched. That can sometimes cause
4788 ;; moves back and forth over a quite large region of comments and
4789 ;; macros, which would be repeated for each changed character when
4790 ;; we're called during fontification, since font-lock refontifies the
4791 ;; current line for each change. Thus it's worthwhile to cache the
4792 ;; first match.
4793 ;;
4794 ;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4795 ;; the syntactic whitespace less or equal to some start position.
4796 ;; There's no cached value if it's nil.
4797 ;;
4798 ;; `c-find-decl-match-pos' is the match position if
4799 ;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4800 ;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4801 (defvar c-find-decl-syntactic-pos nil)
4802 (make-variable-buffer-local 'c-find-decl-syntactic-pos)
4803 (defvar c-find-decl-match-pos nil)
4804 (make-variable-buffer-local 'c-find-decl-match-pos)
4805
4806 (defsubst c-invalidate-find-decl-cache (change-min-pos)
4807 (and c-find-decl-syntactic-pos
4808 (< change-min-pos c-find-decl-syntactic-pos)
4809 (setq c-find-decl-syntactic-pos nil)))
4810
4811 ; (defface c-debug-decl-spot-face
4812 ; '((t (:background "Turquoise")))
4813 ; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4814 ; (defface c-debug-decl-sws-face
4815 ; '((t (:background "Khaki")))
4816 ; "Debug face to mark the syntactic whitespace between the declaration
4817 ; spots and the preceding token end.")
4818
4819 (defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4820 (when (facep 'c-debug-decl-spot-face)
4821 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
4822 (c-debug-add-face (max match-pos (point-min)) decl-pos
4823 'c-debug-decl-sws-face)
4824 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4825 'c-debug-decl-spot-face))))
4826 (defmacro c-debug-remove-decl-spot-faces (beg end)
4827 (when (facep 'c-debug-decl-spot-face)
4828 `(c-save-buffer-state ()
4829 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4830 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4831
4832 (defmacro c-find-decl-prefix-search ()
4833 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4834 ;; but it contains lots of free variables that refer to things
4835 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4836 ;; if there is a match, otherwise at `cfd-limit'.
4837 ;;
4838 ;; The macro moves point forward to the next putative start of a declaration
4839 ;; or cfd-limit. This decl start is the next token after a "declaration
4840 ;; prefix". The declaration prefix is the earlier of `cfd-prop-match' and
4841 ;; `cfd-re-match'. `cfd-match-pos' is set to the decl prefix.
4842 ;;
4843 ;; This macro might do hidden buffer changes.
4844
4845 '(progn
4846 ;; Find the next property match position if we haven't got one already.
4847 (unless cfd-prop-match
4848 (save-excursion
4849 (while (progn
4850 (goto-char (c-next-single-property-change
4851 (point) 'c-type nil cfd-limit))
4852 (and (< (point) cfd-limit)
4853 (not (eq (c-get-char-property (1- (point)) 'c-type)
4854 'c-decl-end)))))
4855 (setq cfd-prop-match (point))))
4856
4857 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4858 ;; got one already.
4859 (unless cfd-re-match
4860
4861 (if (> cfd-re-match-end (point))
4862 (goto-char cfd-re-match-end))
4863
4864 ;; Each time round, the next `while' moves forward over a pseudo match
4865 ;; of `c-decl-prefix-or-start-re' which is either inside a literal, or
4866 ;; is a ":" not preceded by "public", etc.. `cfd-re-match' and
4867 ;; `cfd-re-match-end' get set.
4868 (while
4869 (progn
4870 (setq cfd-re-match-end (re-search-forward c-decl-prefix-or-start-re
4871 cfd-limit 'move))
4872 (cond
4873 ((null cfd-re-match-end)
4874 ;; No match. Finish up and exit the loop.
4875 (setq cfd-re-match cfd-limit)
4876 nil)
4877 ((c-got-face-at
4878 (if (setq cfd-re-match (match-end 1))
4879 ;; Matched the end of a token preceding a decl spot.
4880 (progn
4881 (goto-char cfd-re-match)
4882 (1- cfd-re-match))
4883 ;; Matched a token that start a decl spot.
4884 (goto-char (match-beginning 0))
4885 (point))
4886 c-literal-faces)
4887 ;; Pseudo match inside a comment or string literal. Skip out
4888 ;; of comments and string literals.
4889 (while (progn
4890 (goto-char (c-next-single-property-change
4891 (point) 'face nil cfd-limit))
4892 (and (< (point) cfd-limit)
4893 (c-got-face-at (point) c-literal-faces))))
4894 t) ; Continue the loop over pseudo matches.
4895 ((and (match-string 1)
4896 (string= (match-string 1) ":")
4897 (save-excursion
4898 (or (/= (c-backward-token-2 2) 0) ; no search limit. :-(
4899 (not (looking-at c-decl-start-colon-kwd-re)))))
4900 ;; Found a ":" which isn't part of "public:", etc.
4901 t)
4902 (t nil)))) ;; Found a real match. Exit the pseudo-match loop.
4903
4904 ;; If our match was at the decl start, we have to back up over the
4905 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4906 ;; any decl spots in the syntactic ws.
4907 (unless cfd-re-match
4908 (c-backward-syntactic-ws)
4909 (setq cfd-re-match (point))))
4910
4911 ;; Choose whichever match is closer to the start.
4912 (if (< cfd-re-match cfd-prop-match)
4913 (setq cfd-match-pos cfd-re-match
4914 cfd-re-match nil)
4915 (setq cfd-match-pos cfd-prop-match
4916 cfd-prop-match nil))
4917
4918 (goto-char cfd-match-pos)
4919
4920 (when (< cfd-match-pos cfd-limit)
4921 ;; Skip forward past comments only so we don't skip macros.
4922 (c-forward-comments)
4923 ;; Set the position to continue at. We can avoid going over
4924 ;; the comments skipped above a second time, but it's possible
4925 ;; that the comment skipping has taken us past `cfd-prop-match'
4926 ;; since the property might be used inside comments.
4927 (setq cfd-continue-pos (if cfd-prop-match
4928 (min cfd-prop-match (point))
4929 (point))))))
4930
4931 (defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
4932 ;; Call CFD-FUN for each possible spot for a declaration, cast or
4933 ;; label from the point to CFD-LIMIT.
4934 ;;
4935 ;; CFD-FUN is called with point at the start of the spot. It's passed two
4936 ;; arguments: The first is the end position of the token preceding the spot,
4937 ;; or 0 for the implicit match at bob. The second is a flag that is t when
4938 ;; the match is inside a macro. Point should be moved forward by at least
4939 ;; one token.
4940 ;;
4941 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
4942 ;; it should return non-nil to ensure that the next search will find them.
4943 ;;
4944 ;; Such a spot is:
4945 ;; o The first token after bob.
4946 ;; o The first token after the end of submatch 1 in
4947 ;; `c-decl-prefix-or-start-re' when that submatch matches. This
4948 ;; submatch is typically a (L or R) brace or paren, a ;, or a ,.
4949 ;; o The start of each `c-decl-prefix-or-start-re' match when
4950 ;; submatch 1 doesn't match. This is, for example, the keyword
4951 ;; "class" in Pike.
4952 ;; o The start of a previously recognized declaration; "recognized"
4953 ;; means that the last char of the previous token has a `c-type'
4954 ;; text property with the value `c-decl-end'; this only holds
4955 ;; when `c-type-decl-end-used' is set.
4956 ;;
4957 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4958 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4959 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
4960 ;;
4961 ;; If the match is inside a macro then the buffer is narrowed to the
4962 ;; end of it, so that CFD-FUN can investigate the following tokens
4963 ;; without matching something that begins inside a macro and ends
4964 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4965 ;; CFD-FACE-CHECKLIST checks exist.
4966 ;;
4967 ;; The spots are visited approximately in order from top to bottom.
4968 ;; It's however the positions where `c-decl-prefix-or-start-re'
4969 ;; matches and where `c-decl-end' properties are found that are in
4970 ;; order. Since the spots often are at the following token, they
4971 ;; might be visited out of order insofar as more spots are reported
4972 ;; later on within the syntactic whitespace between the match
4973 ;; positions and their spots.
4974 ;;
4975 ;; It's assumed that comments and strings are fontified in the
4976 ;; searched range.
4977 ;;
4978 ;; This is mainly used in fontification, and so has an elaborate
4979 ;; cache to handle repeated calls from the same start position; see
4980 ;; the variables above.
4981 ;;
4982 ;; All variables in this function begin with `cfd-' to avoid name
4983 ;; collision with the (dynamically bound) variables used in CFD-FUN.
4984 ;;
4985 ;; This function might do hidden buffer changes.
4986
4987 (let ((cfd-start-pos (point)) ; never changed
4988 (cfd-buffer-end (point-max))
4989 ;; The end of the token preceding the decl spot last found
4990 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4991 ;; no match.
4992 cfd-re-match
4993 ;; The end position of the last `c-decl-prefix-or-start-re'
4994 ;; match. If this is greater than `cfd-continue-pos', the
4995 ;; next regexp search is started here instead.
4996 (cfd-re-match-end (point-min))
4997 ;; The end of the last `c-decl-end' found by
4998 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4999 ;; match. If searching for the property isn't needed then we
5000 ;; disable it by setting it to `cfd-limit' directly.
5001 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
5002 ;; The end of the token preceding the decl spot last found by
5003 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
5004 ;; bob. `cfd-limit' if there's no match. In other words,
5005 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
5006 (cfd-match-pos cfd-limit)
5007 ;; The position to continue searching at.
5008 cfd-continue-pos
5009 ;; The position of the last "real" token we've stopped at.
5010 ;; This can be greater than `cfd-continue-pos' when we get
5011 ;; hits inside macros or at `c-decl-end' positions inside
5012 ;; comments.
5013 (cfd-token-pos 0)
5014 ;; The end position of the last entered macro.
5015 (cfd-macro-end 0))
5016
5017 ;; Initialize by finding a syntactically relevant start position
5018 ;; before the point, and do the first `c-decl-prefix-or-start-re'
5019 ;; search unless we're at bob.
5020
5021 (let (start-in-literal start-in-macro syntactic-pos)
5022 ;; Must back up a bit since we look for the end of the previous
5023 ;; statement or declaration, which is earlier than the first
5024 ;; returned match.
5025
5026 ;; This `cond' moves back over any literals or macros. It has special
5027 ;; handling for when the region being searched is entirely within a
5028 ;; macro. It sets `cfd-continue-pos' (unless we've reached
5029 ;; `cfd-limit').
5030 (cond
5031 ;; First we need to move to a syntactically relevant position.
5032 ;; Begin by backing out of comment or string literals.
5033 ;;
5034 ;; This arm of the cond actually triggers if we're in a literal,
5035 ;; and cfd-limit is at most at BONL.
5036 ((and
5037 ;; This arm of the `and' moves backwards out of a literal when
5038 ;; the face at point is a literal face. In this case, its value
5039 ;; is always non-nil.
5040 (when (c-got-face-at (point) c-literal-faces)
5041 ;; Try to use the faces to back up to the start of the
5042 ;; literal. FIXME: What if the point is on a declaration
5043 ;; inside a comment?
5044 (while (and (not (bobp))
5045 (c-got-face-at (1- (point)) c-literal-faces))
5046 (goto-char (previous-single-property-change
5047 (point) 'face nil (point-min))))
5048
5049 ;; XEmacs doesn't fontify the quotes surrounding string
5050 ;; literals.
5051 (and (featurep 'xemacs)
5052 (eq (get-text-property (point) 'face)
5053 'font-lock-string-face)
5054 (not (bobp))
5055 (progn (backward-char)
5056 (not (looking-at c-string-limit-regexp)))
5057 (forward-char))
5058
5059 ;; Don't trust the literal to contain only literal faces
5060 ;; (the font lock package might not have fontified the
5061 ;; start of it at all, for instance) so check that we have
5062 ;; arrived at something that looks like a start or else
5063 ;; resort to `c-literal-limits'.
5064 (unless (looking-at c-literal-start-regexp)
5065 (let ((range (c-literal-limits)))
5066 (if range (goto-char (car range)))))
5067
5068 (setq start-in-literal (point))) ; end of `and' arm.
5069
5070 ;; The start is in a literal. If the limit is in the same
5071 ;; one we don't have to find a syntactic position etc. We
5072 ;; only check that if the limit is at or before bonl to save
5073 ;; time; it covers the by far most common case when font-lock
5074 ;; refontifies the current line only.
5075 (<= cfd-limit (c-point 'bonl cfd-start-pos))
5076 (save-excursion
5077 (goto-char cfd-start-pos)
5078 (while (progn
5079 (goto-char (c-next-single-property-change
5080 (point) 'face nil cfd-limit))
5081 (and (< (point) cfd-limit)
5082 (c-got-face-at (point) c-literal-faces))))
5083 (= (point) cfd-limit))) ; end of `cond' arm condition
5084
5085 ;; Completely inside a literal. Set up variables to trig the
5086 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
5087 ;; find a suitable start position.
5088 (setq cfd-continue-pos start-in-literal)) ; end of `cond' arm
5089
5090 ;; Check if the region might be completely inside a macro, to
5091 ;; optimize that like the completely-inside-literal above.
5092 ((save-excursion
5093 (and (= (forward-line 1) 0)
5094 (bolp) ; forward-line has funny behavior at eob.
5095 (>= (point) cfd-limit)
5096 (progn (backward-char)
5097 (eq (char-before) ?\\))))
5098 ;; (Maybe) completely inside a macro. Only need to trig the
5099 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
5100 ;; set things up.
5101 (setq cfd-continue-pos (1- cfd-start-pos)
5102 start-in-macro t))
5103
5104 ;; The default arm of the `cond' moves back over any macro we're in
5105 ;; and over any syntactic WS. It sets `c-find-decl-syntactic-pos'.
5106 (t
5107 ;; Back out of any macro so we don't miss any declaration
5108 ;; that could follow after it.
5109 (when (c-beginning-of-macro)
5110 (setq start-in-macro t))
5111
5112 ;; Now we're at a proper syntactically relevant position so we
5113 ;; can use the cache. But first clear it if it applied
5114 ;; further down.
5115 (c-invalidate-find-decl-cache cfd-start-pos)
5116
5117 (setq syntactic-pos (point))
5118 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
5119 ;; Don't have to do this if the cache is relevant here,
5120 ;; typically if the same line is refontified again. If
5121 ;; we're just some syntactic whitespace further down we can
5122 ;; still use the cache to limit the skipping.
5123 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
5124
5125 ;; If we hit `c-find-decl-syntactic-pos' and
5126 ;; `c-find-decl-match-pos' is set then we install the cached
5127 ;; values. If we hit `c-find-decl-syntactic-pos' and
5128 ;; `c-find-decl-match-pos' is nil then we know there's no decl
5129 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
5130 ;; and so we can continue the search from this point. If we
5131 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
5132 ;; the right spot to begin searching anyway.
5133 (if (and (eq (point) c-find-decl-syntactic-pos)
5134 c-find-decl-match-pos)
5135 (setq cfd-match-pos c-find-decl-match-pos
5136 cfd-continue-pos syntactic-pos)
5137
5138 (setq c-find-decl-syntactic-pos syntactic-pos)
5139
5140 (when (if (bobp)
5141 ;; Always consider bob a match to get the first
5142 ;; declaration in the file. Do this separately instead of
5143 ;; letting `c-decl-prefix-or-start-re' match bob, so that
5144 ;; regexp always can consume at least one character to
5145 ;; ensure that we won't get stuck in an infinite loop.
5146 (setq cfd-re-match 0)
5147 (backward-char)
5148 (c-beginning-of-current-token)
5149 (< (point) cfd-limit))
5150 ;; Do an initial search now. In the bob case above it's
5151 ;; only done to search for a `c-decl-end' spot.
5152 (c-find-decl-prefix-search)) ; sets cfd-continue-pos
5153
5154 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5155 cfd-match-pos))))) ; end of `cond'
5156
5157 ;; Advance `cfd-continue-pos' if it's before the start position.
5158 ;; The closest continue position that might have effect at or
5159 ;; after the start depends on what we started in. This also
5160 ;; finds a suitable start position in the special cases when the
5161 ;; region is completely within a literal or macro.
5162 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5163
5164 (cond
5165 (start-in-macro
5166 ;; If we're in a macro then it's the closest preceding token
5167 ;; in the macro. Check this before `start-in-literal',
5168 ;; since if we're inside a literal in a macro, the preceding
5169 ;; token is earlier than any `c-decl-end' spot inside the
5170 ;; literal (comment).
5171 (goto-char (or start-in-literal cfd-start-pos))
5172 ;; The only syntactic ws in macros are comments.
5173 (c-backward-comments)
5174 (backward-char)
5175 (c-beginning-of-current-token))
5176
5177 (start-in-literal
5178 ;; If we're in a comment it can only be the closest
5179 ;; preceding `c-decl-end' position within that comment, if
5180 ;; any. Go back to the beginning of such a property so that
5181 ;; `c-find-decl-prefix-search' will find the end of it.
5182 ;; (Can't stop at the end and install it directly on
5183 ;; `cfd-prop-match' since that variable might be cleared
5184 ;; after `cfd-fun' below.)
5185 ;;
5186 ;; Note that if the literal is a string then the property
5187 ;; search will simply skip to the beginning of it right
5188 ;; away.
5189 (if (not c-type-decl-end-used)
5190 (goto-char start-in-literal)
5191 (goto-char cfd-start-pos)
5192 (while (progn
5193 (goto-char (previous-single-property-change
5194 (point) 'c-type nil start-in-literal))
5195 (and (> (point) start-in-literal)
5196 (not (eq (c-get-char-property (point) 'c-type)
5197 'c-decl-end))))))
5198
5199 (when (= (point) start-in-literal)
5200 ;; Didn't find any property inside the comment, so we can
5201 ;; skip it entirely. (This won't skip past a string, but
5202 ;; that'll be handled quickly by the next
5203 ;; `c-find-decl-prefix-search' anyway.)
5204 (c-forward-single-comment)
5205 (if (> (point) cfd-limit)
5206 (goto-char cfd-limit))))
5207
5208 (t
5209 ;; If we started in normal code, the only match that might
5210 ;; apply before the start is what we already got in
5211 ;; `cfd-match-pos' so we can continue at the start position.
5212 ;; (Note that we don't get here if the first match is below
5213 ;; it.)
5214 (goto-char cfd-start-pos))) ; end of `cond'
5215
5216 ;; Delete found matches if they are before our new continue
5217 ;; position, so that `c-find-decl-prefix-search' won't back up
5218 ;; to them later on.
5219 (setq cfd-continue-pos (point))
5220 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5221 (setq cfd-re-match nil))
5222 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5223 (setq cfd-prop-match nil))) ; end of `when'
5224
5225 (if syntactic-pos
5226 ;; This is the normal case and we got a proper syntactic
5227 ;; position. If there's a match then it's always outside
5228 ;; macros and comments, so advance to the next token and set
5229 ;; `cfd-token-pos'. The loop below will later go back using
5230 ;; `cfd-continue-pos' to fix declarations inside the
5231 ;; syntactic ws.
5232 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5233 (goto-char syntactic-pos)
5234 (c-forward-syntactic-ws)
5235 (and cfd-continue-pos
5236 (< cfd-continue-pos (point))
5237 (setq cfd-token-pos (point))))
5238
5239 ;; Have one of the special cases when the region is completely
5240 ;; within a literal or macro. `cfd-continue-pos' is set to a
5241 ;; good start position for the search, so do it.
5242 (c-find-decl-prefix-search)))
5243
5244 ;; Now loop, one decl spot per iteration. We already have the first
5245 ;; match in `cfd-match-pos'.
5246 (while (progn
5247 ;; Go forward over "false matches", one per iteration.
5248 (while (and
5249 (< cfd-match-pos cfd-limit)
5250
5251 (or
5252 ;; Kludge to filter out matches on the "<" that
5253 ;; aren't open parens, for the sake of languages
5254 ;; that got `c-recognize-<>-arglists' set.
5255 (and (eq (char-before cfd-match-pos) ?<)
5256 (not (c-get-char-property (1- cfd-match-pos)
5257 'syntax-table)))
5258
5259 ;; If `cfd-continue-pos' is less or equal to
5260 ;; `cfd-token-pos', we've got a hit inside a macro
5261 ;; that's in the syntactic whitespace before the last
5262 ;; "real" declaration we've checked. If they're equal
5263 ;; we've arrived at the declaration a second time, so
5264 ;; there's nothing to do.
5265 (= cfd-continue-pos cfd-token-pos)
5266
5267 (progn
5268 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5269 ;; we're still searching for declarations embedded in
5270 ;; the syntactic whitespace. In that case we need
5271 ;; only to skip comments and not macros, since they
5272 ;; can't be nested, and that's already been done in
5273 ;; `c-find-decl-prefix-search'.
5274 (when (> cfd-continue-pos cfd-token-pos)
5275 (c-forward-syntactic-ws)
5276 (setq cfd-token-pos (point)))
5277
5278 ;; Continue if the following token fails the
5279 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5280 (when (or (>= (point) cfd-limit)
5281 (not (looking-at cfd-decl-re))
5282 (and cfd-face-checklist
5283 (not (c-got-face-at
5284 (point) cfd-face-checklist))))
5285 (goto-char cfd-continue-pos)
5286 t)))
5287
5288 (< (point) cfd-limit)) ; end of "false matches" condition
5289 (c-find-decl-prefix-search)) ; end of "false matches" loop
5290
5291 (< (point) cfd-limit)) ; end of condition for "decl-spot" while
5292
5293 (when (and
5294 (>= (point) cfd-start-pos)
5295
5296 (progn
5297 ;; Narrow to the end of the macro if we got a hit inside
5298 ;; one, to avoid recognizing things that start inside the
5299 ;; macro and end outside it.
5300 (when (> cfd-match-pos cfd-macro-end)
5301 ;; Not in the same macro as in the previous round.
5302 (save-excursion
5303 (goto-char cfd-match-pos)
5304 (setq cfd-macro-end
5305 (if (save-excursion (and (c-beginning-of-macro)
5306 (< (point) cfd-match-pos)))
5307 (progn (c-end-of-macro)
5308 (point))
5309 0))))
5310
5311 (if (zerop cfd-macro-end)
5312 t
5313 (if (> cfd-macro-end (point))
5314 (progn (narrow-to-region (point-min) cfd-macro-end)
5315 t)
5316 ;; The matched token was the last thing in the macro,
5317 ;; so the whole match is bogus.
5318 (setq cfd-macro-end 0)
5319 nil)))) ; end of when condition
5320
5321 (c-debug-put-decl-spot-faces cfd-match-pos (point))
5322 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5323 (setq cfd-prop-match nil))
5324
5325 (when (/= cfd-macro-end 0)
5326 ;; Restore limits if we did macro narrowing above.
5327 (narrow-to-region (point-min) cfd-buffer-end)))
5328
5329 (goto-char cfd-continue-pos)
5330 (if (= cfd-continue-pos cfd-limit)
5331 (setq cfd-match-pos cfd-limit)
5332 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5333 ; cfd-match-pos, etc.
5334
5335 \f
5336 ;; A cache for found types.
5337
5338 ;; Buffer local variable that contains an obarray with the types we've
5339 ;; found. If a declaration is recognized somewhere we record the
5340 ;; fully qualified identifier in it to recognize it as a type
5341 ;; elsewhere in the file too. This is not accurate since we do not
5342 ;; bother with the scoping rules of the languages, but in practice the
5343 ;; same name is seldom used as both a type and something else in a
5344 ;; file, and we only use this as a last resort in ambiguous cases (see
5345 ;; `c-forward-decl-or-cast-1').
5346 ;;
5347 ;; Not every type need be in this cache. However, things which have
5348 ;; ceased to be types must be removed from it.
5349 ;;
5350 ;; Template types in C++ are added here too but with the template
5351 ;; arglist replaced with "<>" in references or "<" for the one in the
5352 ;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5353 ;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5354 ;; template specs can be fairly sized programs in themselves) and
5355 ;; improves the hit ratio (it's a type regardless of the template
5356 ;; args; it's just not the same type, but we're only interested in
5357 ;; recognizing types, not telling distinct types apart). Note that
5358 ;; template types in references are added here too; from the example
5359 ;; above there will also be an entry "Foo<".
5360 (defvar c-found-types nil)
5361 (make-variable-buffer-local 'c-found-types)
5362
5363 (defsubst c-clear-found-types ()
5364 ;; Clears `c-found-types'.
5365 (setq c-found-types (make-vector 53 0)))
5366
5367 (defun c-add-type (from to)
5368 ;; Add the given region as a type in `c-found-types'. If the region
5369 ;; doesn't match an existing type but there is a type which is equal
5370 ;; to the given one except that the last character is missing, then
5371 ;; the shorter type is removed. That's done to avoid adding all
5372 ;; prefixes of a type as it's being entered and font locked. This
5373 ;; doesn't cover cases like when characters are removed from a type
5374 ;; or added in the middle. We'd need the position of point when the
5375 ;; font locking is invoked to solve this well.
5376 ;;
5377 ;; This function might do hidden buffer changes.
5378 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5379 (unless (intern-soft type c-found-types)
5380 (unintern (substring type 0 -1) c-found-types)
5381 (intern type c-found-types))))
5382
5383 (defun c-unfind-type (name)
5384 ;; Remove the "NAME" from c-found-types, if present.
5385 (unintern name c-found-types))
5386
5387 (defsubst c-check-type (from to)
5388 ;; Return non-nil if the given region contains a type in
5389 ;; `c-found-types'.
5390 ;;
5391 ;; This function might do hidden buffer changes.
5392 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5393 c-found-types))
5394
5395 (defun c-list-found-types ()
5396 ;; Return all the types in `c-found-types' as a sorted list of
5397 ;; strings.
5398 (let (type-list)
5399 (mapatoms (lambda (type)
5400 (setq type-list (cons (symbol-name type)
5401 type-list)))
5402 c-found-types)
5403 (sort type-list 'string-lessp)))
5404
5405 ;; Shut up the byte compiler.
5406 (defvar c-maybe-stale-found-type)
5407
5408 (defun c-trim-found-types (beg end old-len)
5409 ;; An after change function which, in conjunction with the info in
5410 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5411 ;; from `c-found-types', should this type have become stale. For
5412 ;; example, this happens to "foo" when "foo \n bar();" becomes
5413 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5414 ;; the fontification.
5415 ;;
5416 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5417 ;; type?
5418 (when (> end beg)
5419 (save-excursion
5420 (when (< end (point-max))
5421 (goto-char end)
5422 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5423 (progn (goto-char end)
5424 (c-end-of-current-token)))
5425 (c-unfind-type (buffer-substring-no-properties
5426 end (point)))))
5427 (when (> beg (point-min))
5428 (goto-char beg)
5429 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5430 (progn (goto-char beg)
5431 (c-beginning-of-current-token)))
5432 (c-unfind-type (buffer-substring-no-properties
5433 (point) beg))))))
5434
5435 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5436 (cond
5437 ;; Changing the amount of (already existing) whitespace - don't do anything.
5438 ((and (c-partial-ws-p beg end)
5439 (or (= beg end) ; removal of WS
5440 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5441
5442 ;; The syntactic relationship which defined a "found type" has been
5443 ;; destroyed.
5444 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5445 (c-unfind-type (cadr c-maybe-stale-found-type)))
5446 ;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5447 )))
5448
5449 \f
5450 ;; Setting and removing syntax properties on < and > in languages (C++
5451 ;; and Java) where they can be template/generic delimiters as well as
5452 ;; their normal meaning of "less/greater than".
5453
5454 ;; Normally, < and > have syntax 'punctuation'. When they are found to
5455 ;; be delimiters, they are marked as such with the category properties
5456 ;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5457
5458 ;; STRATEGY:
5459 ;;
5460 ;; It is impossible to determine with certainty whether a <..> pair in
5461 ;; C++ is two comparison operators or is template delimiters, unless
5462 ;; one duplicates a lot of a C++ compiler. For example, the following
5463 ;; code fragment:
5464 ;;
5465 ;; foo (a < b, c > d) ;
5466 ;;
5467 ;; could be a function call with two integer parameters (each a
5468 ;; relational expression), or it could be a constructor for class foo
5469 ;; taking one parameter d of templated type "a < b, c >". They are
5470 ;; somewhat easier to distinguish in Java.
5471 ;;
5472 ;; The strategy now (2010-01) adopted is to mark and unmark < and
5473 ;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5474 ;; individually when their context so indicated. This gave rise to
5475 ;; intractable problems when one of a matching pair was deleted, or
5476 ;; pulled into a literal.]
5477 ;;
5478 ;; At each buffer change, the syntax-table properties are removed in a
5479 ;; before-change function and reapplied, when needed, in an
5480 ;; after-change function. It is far more important that the
5481 ;; properties get removed when they they are spurious than that they
5482 ;; be present when wanted.
5483 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5484 (defun c-clear-<-pair-props (&optional pos)
5485 ;; POS (default point) is at a < character. If it is marked with
5486 ;; open paren syntax-table text property, remove the property,
5487 ;; together with the close paren property on the matching > (if
5488 ;; any).
5489 (save-excursion
5490 (if pos
5491 (goto-char pos)
5492 (setq pos (point)))
5493 (when (equal (c-get-char-property (point) 'syntax-table)
5494 c-<-as-paren-syntax)
5495 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5496 (c-go-list-forward))
5497 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5498 c->-as-paren-syntax) ; should always be true.
5499 (c-unmark-<->-as-paren (1- (point))))
5500 (c-unmark-<->-as-paren pos))))
5501
5502 (defun c-clear->-pair-props (&optional pos)
5503 ;; POS (default point) is at a > character. If it is marked with
5504 ;; close paren syntax-table property, remove the property, together
5505 ;; with the open paren property on the matching < (if any).
5506 (save-excursion
5507 (if pos
5508 (goto-char pos)
5509 (setq pos (point)))
5510 (when (equal (c-get-char-property (point) 'syntax-table)
5511 c->-as-paren-syntax)
5512 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5513 (c-go-up-list-backward))
5514 (when (equal (c-get-char-property (point) 'syntax-table)
5515 c-<-as-paren-syntax) ; should always be true.
5516 (c-unmark-<->-as-paren (point)))
5517 (c-unmark-<->-as-paren pos))))
5518
5519 (defun c-clear-<>-pair-props (&optional pos)
5520 ;; POS (default point) is at a < or > character. If it has an
5521 ;; open/close paren syntax-table property, remove this property both
5522 ;; from the current character and its partner (which will also be
5523 ;; thusly marked).
5524 (cond
5525 ((eq (char-after) ?\<)
5526 (c-clear-<-pair-props pos))
5527 ((eq (char-after) ?\>)
5528 (c-clear->-pair-props pos))
5529 (t (c-benign-error
5530 "c-clear-<>-pair-props called from wrong position"))))
5531
5532 (defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5533 ;; POS (default point) is at a < character. If it is both marked
5534 ;; with open/close paren syntax-table property, and has a matching >
5535 ;; (also marked) which is after LIM, remove the property both from
5536 ;; the current > and its partner. Return t when this happens, nil
5537 ;; when it doesn't.
5538 (save-excursion
5539 (if pos
5540 (goto-char pos)
5541 (setq pos (point)))
5542 (when (equal (c-get-char-property (point) 'syntax-table)
5543 c-<-as-paren-syntax)
5544 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5545 (c-go-list-forward))
5546 (when (and (>= (point) lim)
5547 (equal (c-get-char-property (1- (point)) 'syntax-table)
5548 c->-as-paren-syntax)) ; should always be true.
5549 (c-unmark-<->-as-paren (1- (point)))
5550 (c-unmark-<->-as-paren pos))
5551 t)))
5552
5553 (defun c-clear->-pair-props-if-match-before (lim &optional pos)
5554 ;; POS (default point) is at a > character. If it is both marked
5555 ;; with open/close paren syntax-table property, and has a matching <
5556 ;; (also marked) which is before LIM, remove the property both from
5557 ;; the current < and its partner. Return t when this happens, nil
5558 ;; when it doesn't.
5559 (save-excursion
5560 (if pos
5561 (goto-char pos)
5562 (setq pos (point)))
5563 (when (equal (c-get-char-property (point) 'syntax-table)
5564 c->-as-paren-syntax)
5565 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5566 (c-go-up-list-backward))
5567 (when (and (<= (point) lim)
5568 (equal (c-get-char-property (point) 'syntax-table)
5569 c-<-as-paren-syntax)) ; should always be true.
5570 (c-unmark-<->-as-paren (point))
5571 (c-unmark-<->-as-paren pos))
5572 t)))
5573
5574 ;; Set by c-common-init in cc-mode.el.
5575 (defvar c-new-BEG)
5576 (defvar c-new-END)
5577
5578 (defun c-before-change-check-<>-operators (beg end)
5579 ;; Unmark certain pairs of "< .... >" which are currently marked as
5580 ;; template/generic delimiters. (This marking is via syntax-table
5581 ;; text properties).
5582 ;;
5583 ;; These pairs are those which are in the current "statement" (i.e.,
5584 ;; the region between the {, }, or ; before BEG and the one after
5585 ;; END), and which enclose any part of the interval (BEG END).
5586 ;;
5587 ;; Note that in C++ (?and Java), template/generic parens cannot
5588 ;; enclose a brace or semicolon, so we use these as bounds on the
5589 ;; region we must work on.
5590 ;;
5591 ;; This function is called from before-change-functions (via
5592 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5593 ;; and point is undefined, both at entry and exit.
5594 ;;
5595 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5596 ;; 2010-01-29.
5597 (save-excursion
5598 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
5599 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5600 new-beg new-end need-new-beg need-new-end)
5601 ;; Locate the barrier before the changed region
5602 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5603 (c-syntactic-skip-backward "^;{}" (c-determine-limit 512))
5604 (setq new-beg (point))
5605
5606 ;; Remove the syntax-table/category properties from each pertinent <...>
5607 ;; pair. Firsly, the ones with the < before beg and > after beg.
5608 (while
5609 (c-search-forward-char-property 'syntax-table c-<-as-paren-syntax beg)
5610 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5611 (setq need-new-beg t)))
5612
5613 ;; Locate the barrier after END.
5614 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5615 (c-syntactic-re-search-forward "[;{}]" (c-determine-+ve-limit 512) 'end)
5616 (setq new-end (point))
5617
5618 ;; Remove syntax-table properties from the remaining pertinent <...>
5619 ;; pairs, those with a > after end and < before end.
5620 (while (c-search-backward-char-property 'syntax-table c->-as-paren-syntax end)
5621 (if (c-clear->-pair-props-if-match-before end)
5622 (setq need-new-end t)))
5623
5624 ;; Extend the fontification region, if needed.
5625 (when need-new-beg
5626 (goto-char new-beg)
5627 (c-forward-syntactic-ws)
5628 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5629
5630 (when need-new-end
5631 (and (> new-end c-new-END) (setq c-new-END new-end))))))
5632
5633 (defun c-after-change-check-<>-operators (beg end)
5634 ;; This is called from `after-change-functions' when
5635 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5636 ;; chars with paren syntax become part of another operator like "<<"
5637 ;; or ">=".
5638 ;;
5639 ;; This function might do hidden buffer changes.
5640
5641 (save-excursion
5642 (goto-char beg)
5643 (when (or (looking-at "[<>]")
5644 (< (skip-chars-backward "<>") 0))
5645
5646 (goto-char beg)
5647 (c-beginning-of-current-token)
5648 (when (and (< (point) beg)
5649 (looking-at c-<>-multichar-token-regexp)
5650 (< beg (setq beg (match-end 0))))
5651 (while (progn (skip-chars-forward "^<>" beg)
5652 (< (point) beg))
5653 (c-clear-<>-pair-props)
5654 (forward-char))))
5655
5656 (when (< beg end)
5657 (goto-char end)
5658 (when (or (looking-at "[<>]")
5659 (< (skip-chars-backward "<>") 0))
5660
5661 (goto-char end)
5662 (c-beginning-of-current-token)
5663 (when (and (< (point) end)
5664 (looking-at c-<>-multichar-token-regexp)
5665 (< end (setq end (match-end 0))))
5666 (while (progn (skip-chars-forward "^<>" end)
5667 (< (point) end))
5668 (c-clear-<>-pair-props)
5669 (forward-char)))))))
5670
5671
5672 \f
5673 ;; Handling of small scale constructs like types and names.
5674
5675 ;; Dynamically bound variable that instructs `c-forward-type' to also
5676 ;; treat possible types (i.e. those that it normally returns 'maybe or
5677 ;; 'found for) as actual types (and always return 'found for them).
5678 ;; This means that it records them in `c-record-type-identifiers' if
5679 ;; that is set, and that it adds them to `c-found-types'.
5680 (defvar c-promote-possible-types nil)
5681
5682 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5683 ;; mark up successfully parsed arglists with paren syntax properties on
5684 ;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5685 ;; `c-type' property of each argument separating comma.
5686 ;;
5687 ;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5688 ;; all arglists for side effects (i.e. recording types), otherwise it
5689 ;; exploits any existing paren syntax properties to quickly jump to the
5690 ;; end of already parsed arglists.
5691 ;;
5692 ;; Marking up the arglists is not the default since doing that correctly
5693 ;; depends on a proper value for `c-restricted-<>-arglists'.
5694 (defvar c-parse-and-markup-<>-arglists nil)
5695
5696 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5697 ;; not accept arglists that contain binary operators.
5698 ;;
5699 ;; This is primarily used to handle C++ template arglists. C++
5700 ;; disambiguates them by checking whether the preceding name is a
5701 ;; template or not. We can't do that, so we assume it is a template
5702 ;; if it can be parsed as one. That usually works well since
5703 ;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5704 ;; in almost all cases would be pointless.
5705 ;;
5706 ;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5707 ;; should let the comma separate the function arguments instead. And
5708 ;; in a context where the value of the expression is taken, e.g. in
5709 ;; "if (a < b || c > d)", it's probably not a template.
5710 (defvar c-restricted-<>-arglists nil)
5711
5712 ;; Dynamically bound variables that instructs
5713 ;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5714 ;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5715 ;; `c-forward-label' to record the ranges of all the type and
5716 ;; reference identifiers they encounter. They will build lists on
5717 ;; these variables where each element is a cons of the buffer
5718 ;; positions surrounding each identifier. This recording is only
5719 ;; activated when `c-record-type-identifiers' is non-nil.
5720 ;;
5721 ;; All known types that can't be identifiers are recorded, and also
5722 ;; other possible types if `c-promote-possible-types' is set.
5723 ;; Recording is however disabled inside angle bracket arglists that
5724 ;; are encountered inside names and other angle bracket arglists.
5725 ;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
5726 ;; instead.
5727 ;;
5728 ;; Only the names in C++ template style references (e.g. "tmpl" in
5729 ;; "tmpl<a,b>::foo") are recorded as references, other references
5730 ;; aren't handled here.
5731 ;;
5732 ;; `c-forward-label' records the label identifier(s) on
5733 ;; `c-record-ref-identifiers'.
5734 (defvar c-record-type-identifiers nil)
5735 (defvar c-record-ref-identifiers nil)
5736
5737 ;; This variable will receive a cons cell of the range of the last
5738 ;; single identifier symbol stepped over by `c-forward-name' if it's
5739 ;; successful. This is the range that should be put on one of the
5740 ;; record lists above by the caller. It's assigned nil if there's no
5741 ;; such symbol in the name.
5742 (defvar c-last-identifier-range nil)
5743
5744 (defmacro c-record-type-id (range)
5745 (if (eq (car-safe range) 'cons)
5746 ;; Always true.
5747 `(setq c-record-type-identifiers
5748 (cons ,range c-record-type-identifiers))
5749 `(let ((range ,range))
5750 (if range
5751 (setq c-record-type-identifiers
5752 (cons range c-record-type-identifiers))))))
5753
5754 (defmacro c-record-ref-id (range)
5755 (if (eq (car-safe range) 'cons)
5756 ;; Always true.
5757 `(setq c-record-ref-identifiers
5758 (cons ,range c-record-ref-identifiers))
5759 `(let ((range ,range))
5760 (if range
5761 (setq c-record-ref-identifiers
5762 (cons range c-record-ref-identifiers))))))
5763
5764 ;; Dynamically bound variable that instructs `c-forward-type' to
5765 ;; record the ranges of types that only are found. Behaves otherwise
5766 ;; like `c-record-type-identifiers'.
5767 (defvar c-record-found-types nil)
5768
5769 (defmacro c-forward-keyword-prefixed-id (type)
5770 ;; Used internally in `c-forward-keyword-clause' to move forward
5771 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5772 ;; possibly is prefixed by keywords and their associated clauses.
5773 ;; Try with a type/name first to not trip up on those that begin
5774 ;; with a keyword. Return t if a known or found type is moved
5775 ;; over. The point is clobbered if nil is returned. If range
5776 ;; recording is enabled, the identifier is recorded on as a type
5777 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
5778 ;;
5779 ;; This macro might do hidden buffer changes.
5780 `(let (res)
5781 (while (if (setq res ,(if (eq type 'type)
5782 `(c-forward-type)
5783 `(c-forward-name)))
5784 nil
5785 (and (looking-at c-keywords-regexp)
5786 (c-forward-keyword-clause 1))))
5787 (when (memq res '(t known found prefix))
5788 ,(when (eq type 'ref)
5789 `(when c-record-type-identifiers
5790 (c-record-ref-id c-last-identifier-range)))
5791 t)))
5792
5793 (defmacro c-forward-id-comma-list (type update-safe-pos)
5794 ;; Used internally in `c-forward-keyword-clause' to move forward
5795 ;; over a comma separated list of types or names using
5796 ;; `c-forward-keyword-prefixed-id'.
5797 ;;
5798 ;; This macro might do hidden buffer changes.
5799 `(while (and (progn
5800 ,(when update-safe-pos
5801 `(setq safe-pos (point)))
5802 (eq (char-after) ?,))
5803 (progn
5804 (forward-char)
5805 (c-forward-syntactic-ws)
5806 (c-forward-keyword-prefixed-id ,type)))))
5807
5808 (defun c-forward-keyword-clause (match)
5809 ;; Submatch MATCH in the current match data is assumed to surround a
5810 ;; token. If it's a keyword, move over it and any immediately
5811 ;; following clauses associated with it, stopping at the start of
5812 ;; the next token. t is returned in that case, otherwise the point
5813 ;; stays and nil is returned. The kind of clauses that are
5814 ;; recognized are those specified by `c-type-list-kwds',
5815 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5816 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5817 ;; and `c-<>-arglist-kwds'.
5818 ;;
5819 ;; This function records identifier ranges on
5820 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5821 ;; `c-record-type-identifiers' is non-nil.
5822 ;;
5823 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5824 ;; apply directly after the keyword, the type list is moved over
5825 ;; only when there is no unaccounted token before it (i.e. a token
5826 ;; that isn't moved over due to some other keyword list). The
5827 ;; identifier ranges in the list are still recorded if that should
5828 ;; be done, though.
5829 ;;
5830 ;; This function might do hidden buffer changes.
5831
5832 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5833 ;; The call to `c-forward-<>-arglist' below is made after
5834 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5835 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5836 ;; should therefore be nil.
5837 (c-parse-and-markup-<>-arglists t)
5838 c-restricted-<>-arglists)
5839
5840 (when kwd-sym
5841 (goto-char (match-end match))
5842 (c-forward-syntactic-ws)
5843 (setq safe-pos (point))
5844
5845 (cond
5846 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5847 (c-forward-keyword-prefixed-id type))
5848 ;; There's a type directly after a keyword in `c-type-list-kwds'.
5849 (c-forward-id-comma-list type t))
5850
5851 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5852 (c-forward-keyword-prefixed-id ref))
5853 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
5854 (c-forward-id-comma-list ref t))
5855
5856 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5857 (eq (char-after) ?\())
5858 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5859
5860 (forward-char)
5861 (when (and (setq pos (c-up-list-forward))
5862 (eq (char-before pos) ?\)))
5863 (when (and c-record-type-identifiers
5864 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5865 ;; Use `c-forward-type' on every identifier we can find
5866 ;; inside the paren, to record the types.
5867 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5868 (goto-char (match-beginning 0))
5869 (unless (c-forward-type)
5870 (looking-at c-symbol-key) ; Always matches.
5871 (goto-char (match-end 0)))))
5872
5873 (goto-char pos)
5874 (c-forward-syntactic-ws)
5875 (setq safe-pos (point))))
5876
5877 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5878 (eq (char-after) ?<)
5879 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
5880 (c-forward-syntactic-ws)
5881 (setq safe-pos (point)))
5882
5883 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
5884 (not (looking-at c-symbol-start))
5885 (c-safe (c-forward-sexp) t))
5886 (c-forward-syntactic-ws)
5887 (setq safe-pos (point))))
5888
5889 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5890 (if (eq (char-after) ?:)
5891 ;; If we are at the colon already, we move over the type
5892 ;; list after it.
5893 (progn
5894 (forward-char)
5895 (c-forward-syntactic-ws)
5896 (when (c-forward-keyword-prefixed-id type)
5897 (c-forward-id-comma-list type t)))
5898 ;; Not at the colon, so stop here. But the identifier
5899 ;; ranges in the type list later on should still be
5900 ;; recorded.
5901 (and c-record-type-identifiers
5902 (progn
5903 ;; If a keyword matched both one of the types above and
5904 ;; this one, we match `c-colon-type-list-re' after the
5905 ;; clause matched above.
5906 (goto-char safe-pos)
5907 (looking-at c-colon-type-list-re))
5908 (progn
5909 (goto-char (match-end 0))
5910 (c-forward-syntactic-ws)
5911 (c-forward-keyword-prefixed-id type))
5912 ;; There's a type after the `c-colon-type-list-re' match
5913 ;; after a keyword in `c-colon-type-list-kwds'.
5914 (c-forward-id-comma-list type nil))))
5915
5916 (goto-char safe-pos)
5917 t)))
5918
5919 ;; cc-mode requires cc-fonts.
5920 (declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5921
5922 (defun c-forward-<>-arglist (all-types)
5923 ;; The point is assumed to be at a "<". Try to treat it as the open
5924 ;; paren of an angle bracket arglist and move forward to the
5925 ;; corresponding ">". If successful, the point is left after the
5926 ;; ">" and t is returned, otherwise the point isn't moved and nil is
5927 ;; returned. If ALL-TYPES is t then all encountered arguments in
5928 ;; the arglist that might be types are treated as found types.
5929 ;;
5930 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5931 ;; function handles text properties on the angle brackets and argument
5932 ;; separating commas.
5933 ;;
5934 ;; `c-restricted-<>-arglists' controls how lenient the template
5935 ;; arglist recognition should be.
5936 ;;
5937 ;; This function records identifier ranges on
5938 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5939 ;; `c-record-type-identifiers' is non-nil.
5940 ;;
5941 ;; This function might do hidden buffer changes.
5942
5943 (let ((start (point))
5944 ;; If `c-record-type-identifiers' is set then activate
5945 ;; recording of any found types that constitute an argument in
5946 ;; the arglist.
5947 (c-record-found-types (if c-record-type-identifiers t)))
5948 (if (catch 'angle-bracket-arglist-escape
5949 (setq c-record-found-types
5950 (c-forward-<>-arglist-recur all-types)))
5951 (progn
5952 (when (consp c-record-found-types)
5953 (setq c-record-type-identifiers
5954 ;; `nconc' doesn't mind that the tail of
5955 ;; `c-record-found-types' is t.
5956 (nconc c-record-found-types c-record-type-identifiers)))
5957 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
5958 t)
5959
5960 (goto-char start)
5961 nil)))
5962
5963 (defun c-forward-<>-arglist-recur (all-types)
5964 ;; Recursive part of `c-forward-<>-arglist'.
5965 ;;
5966 ;; This function might do hidden buffer changes.
5967 (let ((start (point)) res pos
5968 ;; Cover this so that any recorded found type ranges are
5969 ;; automatically lost if it turns out to not be an angle
5970 ;; bracket arglist. It's propagated through the return value
5971 ;; on successful completion.
5972 (c-record-found-types c-record-found-types)
5973 ;; List that collects the positions after the argument
5974 ;; separating ',' in the arglist.
5975 arg-start-pos)
5976 ;; If the '<' has paren open syntax then we've marked it as an angle
5977 ;; bracket arglist before, so skip to the end.
5978 (if (and (not c-parse-and-markup-<>-arglists)
5979 (c-get-char-property (point) 'syntax-table))
5980
5981 (progn
5982 (forward-char)
5983 (if (and (c-go-up-list-forward)
5984 (eq (char-before) ?>))
5985 t
5986 ;; Got unmatched paren angle brackets. We don't clear the paren
5987 ;; syntax properties and retry, on the basis that it's very
5988 ;; unlikely that paren angle brackets become operators by code
5989 ;; manipulation. It's far more likely that it doesn't match due
5990 ;; to narrowing or some temporary change.
5991 (goto-char start)
5992 nil))
5993
5994 (forward-char) ; Forward over the opening '<'.
5995
5996 (unless (looking-at c-<-op-cont-regexp)
5997 ;; go forward one non-alphanumeric character (group) per iteration of
5998 ;; this loop.
5999 (while (and
6000 (progn
6001 (c-forward-syntactic-ws)
6002 (when (or (and c-record-type-identifiers all-types)
6003 (c-major-mode-is 'java-mode))
6004 ;; All encountered identifiers are types, so set the
6005 ;; promote flag and parse the type.
6006 (progn
6007 (c-forward-syntactic-ws)
6008 (if (looking-at "\\?")
6009 (forward-char)
6010 (when (looking-at c-identifier-start)
6011 (let ((c-promote-possible-types t)
6012 (c-record-found-types t))
6013 (c-forward-type))))
6014
6015 (c-forward-syntactic-ws)
6016
6017 (when (or (looking-at "extends")
6018 (looking-at "super"))
6019 (forward-word)
6020 (c-forward-syntactic-ws)
6021 (let ((c-promote-possible-types t)
6022 (c-record-found-types t))
6023 (c-forward-type)
6024 (c-forward-syntactic-ws)))))
6025
6026 (setq pos (point)) ; e.g. first token inside the '<'
6027
6028 ;; Note: These regexps exploit the match order in \| so
6029 ;; that "<>" is matched by "<" rather than "[^>:-]>".
6030 (c-syntactic-re-search-forward
6031 ;; Stop on ',', '|', '&', '+' and '-' to catch
6032 ;; common binary operators that could be between
6033 ;; two comparison expressions "a<b" and "c>d".
6034 "[<;{},|+&-]\\|[>)]"
6035 nil t t))
6036
6037 (cond
6038 ((eq (char-before) ?>)
6039 ;; Either an operator starting with '>' or the end of
6040 ;; the angle bracket arglist.
6041
6042 (if (looking-at c->-op-without->-cont-regexp)
6043 (progn
6044 (goto-char (match-end 0))
6045 t) ; Continue the loop.
6046
6047 ;; The angle bracket arglist is finished.
6048 (when c-parse-and-markup-<>-arglists
6049 (while arg-start-pos
6050 (c-put-c-type-property (1- (car arg-start-pos))
6051 'c-<>-arg-sep)
6052 (setq arg-start-pos (cdr arg-start-pos)))
6053 (c-mark-<-as-paren start)
6054 (c-mark->-as-paren (1- (point))))
6055 (setq res t)
6056 nil)) ; Exit the loop.
6057
6058 ((eq (char-before) ?<)
6059 ;; Either an operator starting with '<' or a nested arglist.
6060 (setq pos (point))
6061 (let (id-start id-end subres keyword-match)
6062 (cond
6063 ;; The '<' begins a multi-char operator.
6064 ((looking-at c-<-op-cont-regexp)
6065 (goto-char (match-end 0)))
6066 ;; We're at a nested <.....>
6067 ((progn
6068 (backward-char) ; to the '<'
6069 (and
6070 (save-excursion
6071 ;; There's always an identifier before an angle
6072 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
6073 ;; or `c-<>-arglist-kwds'.
6074 (c-backward-syntactic-ws)
6075 (setq id-end (point))
6076 (c-simple-skip-symbol-backward)
6077 (when (or (setq keyword-match
6078 (looking-at c-opt-<>-sexp-key))
6079 (not (looking-at c-keywords-regexp)))
6080 (setq id-start (point))))
6081 (setq subres
6082 (let ((c-promote-possible-types t)
6083 (c-record-found-types t))
6084 (c-forward-<>-arglist-recur
6085 (and keyword-match
6086 (c-keyword-member
6087 (c-keyword-sym (match-string 1))
6088 'c-<>-type-kwds))))))
6089 (or subres (goto-char pos))
6090 subres)
6091 ;; It was an angle bracket arglist.
6092 (setq c-record-found-types subres)
6093
6094 ;; Record the identifier before the template as a type
6095 ;; or reference depending on whether the arglist is last
6096 ;; in a qualified identifier.
6097 (when (and c-record-type-identifiers
6098 (not keyword-match))
6099 (if (and c-opt-identifier-concat-key
6100 (progn
6101 (c-forward-syntactic-ws)
6102 (looking-at c-opt-identifier-concat-key)))
6103 (c-record-ref-id (cons id-start id-end))
6104 (c-record-type-id (cons id-start id-end)))))
6105
6106 ;; At a "less than" operator.
6107 (t
6108 ;; (forward-char) ; NO! We've already gone over the <.
6109 )))
6110 t) ; carry on looping.
6111
6112 ((and (not c-restricted-<>-arglists)
6113 (or (and (eq (char-before) ?&)
6114 (not (eq (char-after) ?&)))
6115 (eq (char-before) ?,)))
6116 ;; Just another argument. Record the position. The
6117 ;; type check stuff that made us stop at it is at
6118 ;; the top of the loop.
6119 (setq arg-start-pos (cons (point) arg-start-pos)))
6120
6121 (t
6122 ;; Got a character that can't be in an angle bracket
6123 ;; arglist argument. Abort using `throw', since
6124 ;; it's useless to try to find a surrounding arglist
6125 ;; if we're nested.
6126 (throw 'angle-bracket-arglist-escape nil))))))
6127 (if res
6128 (or c-record-found-types t)))))
6129
6130 (defun c-backward-<>-arglist (all-types &optional limit)
6131 ;; The point is assumed to be directly after a ">". Try to treat it
6132 ;; as the close paren of an angle bracket arglist and move back to
6133 ;; the corresponding "<". If successful, the point is left at
6134 ;; the "<" and t is returned, otherwise the point isn't moved and
6135 ;; nil is returned. ALL-TYPES is passed on to
6136 ;; `c-forward-<>-arglist'.
6137 ;;
6138 ;; If the optional LIMIT is given, it bounds the backward search.
6139 ;; It's then assumed to be at a syntactically relevant position.
6140 ;;
6141 ;; This is a wrapper around `c-forward-<>-arglist'. See that
6142 ;; function for more details.
6143
6144 (let ((start (point)))
6145 (backward-char)
6146 (if (and (not c-parse-and-markup-<>-arglists)
6147 (c-get-char-property (point) 'syntax-table))
6148
6149 (if (and (c-go-up-list-backward)
6150 (eq (char-after) ?<))
6151 t
6152 ;; See corresponding note in `c-forward-<>-arglist'.
6153 (goto-char start)
6154 nil)
6155
6156 (while (progn
6157 (c-syntactic-skip-backward "^<;{}" limit t)
6158
6159 (and
6160 (if (eq (char-before) ?<)
6161 t
6162 ;; Stopped at bob or a char that isn't allowed in an
6163 ;; arglist, so we've failed.
6164 (goto-char start)
6165 nil)
6166
6167 (if (> (point)
6168 (progn (c-beginning-of-current-token)
6169 (point)))
6170 ;; If we moved then the "<" was part of some
6171 ;; multicharacter token.
6172 t
6173
6174 (backward-char)
6175 (let ((beg-pos (point)))
6176 (if (c-forward-<>-arglist all-types)
6177 (cond ((= (point) start)
6178 ;; Matched the arglist. Break the while.
6179 (goto-char beg-pos)
6180 nil)
6181 ((> (point) start)
6182 ;; We started from a non-paren ">" inside an
6183 ;; arglist.
6184 (goto-char start)
6185 nil)
6186 (t
6187 ;; Matched a shorter arglist. Can be a nested
6188 ;; one so continue looking.
6189 (goto-char beg-pos)
6190 t))
6191 t))))))
6192
6193 (/= (point) start))))
6194
6195 (defun c-forward-name ()
6196 ;; Move forward over a complete name if at the beginning of one,
6197 ;; stopping at the next following token. A keyword, as such,
6198 ;; doesn't count as a name. If the point is not at something that
6199 ;; is recognized as a name then it stays put.
6200 ;;
6201 ;; A name could be something as simple as "foo" in C or something as
6202 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6203 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6204 ;; int>::*volatile const" in C++ (this function is actually little
6205 ;; more than a `looking-at' call in all modes except those that,
6206 ;; like C++, have `c-recognize-<>-arglists' set).
6207 ;;
6208 ;; Return
6209 ;; o - nil if no name is found;
6210 ;; o - 'template if it's an identifier ending with an angle bracket
6211 ;; arglist;
6212 ;; o - 'operator of it's an operator identifier;
6213 ;; o - t if it's some other kind of name.
6214 ;;
6215 ;; This function records identifier ranges on
6216 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6217 ;; `c-record-type-identifiers' is non-nil.
6218 ;;
6219 ;; This function might do hidden buffer changes.
6220
6221 (let ((pos (point)) (start (point)) res id-start id-end
6222 ;; Turn off `c-promote-possible-types' here since we might
6223 ;; call `c-forward-<>-arglist' and we don't want it to promote
6224 ;; every suspect thing in the arglist to a type. We're
6225 ;; typically called from `c-forward-type' in this case, and
6226 ;; the caller only wants the top level type that it finds to
6227 ;; be promoted.
6228 c-promote-possible-types)
6229 (while
6230 (and
6231 (looking-at c-identifier-key)
6232
6233 (progn
6234 ;; Check for keyword. We go to the last symbol in
6235 ;; `c-identifier-key' first.
6236 (goto-char (setq id-end (match-end 0)))
6237 (c-simple-skip-symbol-backward)
6238 (setq id-start (point))
6239
6240 (if (looking-at c-keywords-regexp)
6241 (when (and (c-major-mode-is 'c++-mode)
6242 (looking-at
6243 (cc-eval-when-compile
6244 (concat "\\(operator\\|\\(template\\)\\)"
6245 "\\(" (c-lang-const c-nonsymbol-key c++)
6246 "\\|$\\)")))
6247 (if (match-beginning 2)
6248 ;; "template" is only valid inside an
6249 ;; identifier if preceded by "::".
6250 (save-excursion
6251 (c-backward-syntactic-ws)
6252 (and (c-safe (backward-char 2) t)
6253 (looking-at "::")))
6254 t))
6255
6256 ;; Handle a C++ operator or template identifier.
6257 (goto-char id-end)
6258 (c-forward-syntactic-ws)
6259 (cond ((eq (char-before id-end) ?e)
6260 ;; Got "... ::template".
6261 (let ((subres (c-forward-name)))
6262 (when subres
6263 (setq pos (point)
6264 res subres))))
6265
6266 ((looking-at c-identifier-start)
6267 ;; Got a cast operator.
6268 (when (c-forward-type)
6269 (setq pos (point)
6270 res 'operator)
6271 ;; Now we should match a sequence of either
6272 ;; '*', '&' or a name followed by ":: *",
6273 ;; where each can be followed by a sequence
6274 ;; of `c-opt-type-modifier-key'.
6275 (while (cond ((looking-at "[*&]")
6276 (goto-char (match-end 0))
6277 t)
6278 ((looking-at c-identifier-start)
6279 (and (c-forward-name)
6280 (looking-at "::")
6281 (progn
6282 (goto-char (match-end 0))
6283 (c-forward-syntactic-ws)
6284 (eq (char-after) ?*))
6285 (progn
6286 (forward-char)
6287 t))))
6288 (while (progn
6289 (c-forward-syntactic-ws)
6290 (setq pos (point))
6291 (looking-at c-opt-type-modifier-key))
6292 (goto-char (match-end 1))))))
6293
6294 ((looking-at c-overloadable-operators-regexp)
6295 ;; Got some other operator.
6296 (setq c-last-identifier-range
6297 (cons (point) (match-end 0)))
6298 (goto-char (match-end 0))
6299 (c-forward-syntactic-ws)
6300 (setq pos (point)
6301 res 'operator)))
6302
6303 nil)
6304
6305 ;; `id-start' is equal to `id-end' if we've jumped over
6306 ;; an identifier that doesn't end with a symbol token.
6307 ;; That can occur e.g. for Java import directives on the
6308 ;; form "foo.bar.*".
6309 (when (and id-start (/= id-start id-end))
6310 (setq c-last-identifier-range
6311 (cons id-start id-end)))
6312 (goto-char id-end)
6313 (c-forward-syntactic-ws)
6314 (setq pos (point)
6315 res t)))
6316
6317 (progn
6318 (goto-char pos)
6319 (when (or c-opt-identifier-concat-key
6320 c-recognize-<>-arglists)
6321
6322 (cond
6323 ((and c-opt-identifier-concat-key
6324 (looking-at c-opt-identifier-concat-key))
6325 ;; Got a concatenated identifier. This handles the
6326 ;; cases with tricky syntactic whitespace that aren't
6327 ;; covered in `c-identifier-key'.
6328 (goto-char (match-end 0))
6329 (c-forward-syntactic-ws)
6330 t)
6331
6332 ((and c-recognize-<>-arglists
6333 (eq (char-after) ?<))
6334 ;; Maybe an angle bracket arglist.
6335 (when (let ((c-record-type-identifiers t)
6336 (c-record-found-types t))
6337 (c-forward-<>-arglist nil))
6338
6339 (c-add-type start (1+ pos))
6340 (c-forward-syntactic-ws)
6341 (setq pos (point)
6342 c-last-identifier-range nil)
6343
6344 (if (and c-opt-identifier-concat-key
6345 (looking-at c-opt-identifier-concat-key))
6346
6347 ;; Continue if there's an identifier concatenation
6348 ;; operator after the template argument.
6349 (progn
6350 (when (and c-record-type-identifiers id-start)
6351 (c-record-ref-id (cons id-start id-end)))
6352 (forward-char 2)
6353 (c-forward-syntactic-ws)
6354 t)
6355
6356 (when (and c-record-type-identifiers id-start)
6357 (c-record-type-id (cons id-start id-end)))
6358 (setq res 'template)
6359 nil)))
6360 )))))
6361
6362 (goto-char pos)
6363 res))
6364
6365 (defun c-forward-type (&optional brace-block-too)
6366 ;; Move forward over a type spec if at the beginning of one,
6367 ;; stopping at the next following token. The keyword "typedef"
6368 ;; isn't part of a type spec here.
6369 ;;
6370 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
6371 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
6372 ;; The current (2009-03-10) intention is to convert all uses of
6373 ;; `c-forward-type' to call with this parameter set, then to
6374 ;; eliminate it.
6375 ;;
6376 ;; Return
6377 ;; o - t if it's a known type that can't be a name or other
6378 ;; expression;
6379 ;; o - 'known if it's an otherwise known type (according to
6380 ;; `*-font-lock-extra-types');
6381 ;; o - 'prefix if it's a known prefix of a type;
6382 ;; o - 'found if it's a type that matches one in `c-found-types';
6383 ;; o - 'maybe if it's an identifier that might be a type;
6384 ;; o - 'decltype if it's a decltype(variable) declaration; - or
6385 ;; o - nil if it can't be a type (the point isn't moved then).
6386 ;;
6387 ;; The point is assumed to be at the beginning of a token.
6388 ;;
6389 ;; Note that this function doesn't skip past the brace definition
6390 ;; that might be considered part of the type, e.g.
6391 ;; "enum {a, b, c} foo".
6392 ;;
6393 ;; This function records identifier ranges on
6394 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6395 ;; `c-record-type-identifiers' is non-nil.
6396 ;;
6397 ;; This function might do hidden buffer changes.
6398 (when (and c-recognize-<>-arglists
6399 (looking-at "<"))
6400 (c-forward-<>-arglist t)
6401 (c-forward-syntactic-ws))
6402
6403 (let ((start (point)) pos res name-res id-start id-end id-range)
6404
6405 ;; Skip leading type modifiers. If any are found we know it's a
6406 ;; prefix of a type.
6407 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
6408 (while (looking-at c-opt-type-modifier-key)
6409 (goto-char (match-end 1))
6410 (c-forward-syntactic-ws)
6411 (setq res 'prefix)))
6412
6413 (cond
6414 ((looking-at c-typeof-key) ; e.g. C++'s "decltype".
6415 (goto-char (match-end 1))
6416 (c-forward-syntactic-ws)
6417 (setq res (and (eq (char-after) ?\()
6418 (c-safe (c-forward-sexp))
6419 'decltype))
6420 (if res
6421 (c-forward-syntactic-ws)
6422 (goto-char start)))
6423
6424 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
6425 ; "typedef".
6426 (goto-char (match-end 1))
6427 (c-forward-syntactic-ws)
6428 (setq pos (point))
6429
6430 (setq name-res (c-forward-name))
6431 (setq res (not (null name-res)))
6432 (when (eq name-res t)
6433 ;; In many languages the name can be used without the
6434 ;; prefix, so we add it to `c-found-types'.
6435 (c-add-type pos (point))
6436 (when (and c-record-type-identifiers
6437 c-last-identifier-range)
6438 (c-record-type-id c-last-identifier-range)))
6439 (when (and brace-block-too
6440 (memq res '(t nil))
6441 (eq (char-after) ?\{)
6442 (save-excursion
6443 (c-safe
6444 (progn (c-forward-sexp)
6445 (c-forward-syntactic-ws)
6446 (setq pos (point))))))
6447 (goto-char pos)
6448 (setq res t))
6449 (unless res (goto-char start))) ; invalid syntax
6450
6451 ((progn
6452 (setq pos nil)
6453 (if (looking-at c-identifier-start)
6454 (save-excursion
6455 (setq id-start (point)
6456 name-res (c-forward-name))
6457 (when name-res
6458 (setq id-end (point)
6459 id-range c-last-identifier-range))))
6460 (and (cond ((looking-at c-primitive-type-key)
6461 (setq res t))
6462 ((c-with-syntax-table c-identifier-syntax-table
6463 (looking-at c-known-type-key))
6464 (setq res 'known)))
6465 (or (not id-end)
6466 (>= (save-excursion
6467 (save-match-data
6468 (goto-char (match-end 1))
6469 (c-forward-syntactic-ws)
6470 (setq pos (point))))
6471 id-end)
6472 (setq res nil))))
6473 ;; Looking at a primitive or known type identifier. We've
6474 ;; checked for a name first so that we don't go here if the
6475 ;; known type match only is a prefix of another name.
6476
6477 (setq id-end (match-end 1))
6478
6479 (when (and c-record-type-identifiers
6480 (or c-promote-possible-types (eq res t)))
6481 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6482
6483 (if (and c-opt-type-component-key
6484 (save-match-data
6485 (looking-at c-opt-type-component-key)))
6486 ;; There might be more keywords for the type.
6487 (let (safe-pos)
6488 (c-forward-keyword-clause 1)
6489 (while (progn
6490 (setq safe-pos (point))
6491 (looking-at c-opt-type-component-key))
6492 (when (and c-record-type-identifiers
6493 (looking-at c-primitive-type-key))
6494 (c-record-type-id (cons (match-beginning 1)
6495 (match-end 1))))
6496 (c-forward-keyword-clause 1))
6497 (if (looking-at c-primitive-type-key)
6498 (progn
6499 (when c-record-type-identifiers
6500 (c-record-type-id (cons (match-beginning 1)
6501 (match-end 1))))
6502 (c-forward-keyword-clause 1)
6503 (setq res t))
6504 (goto-char safe-pos)
6505 (setq res 'prefix)))
6506 (unless (save-match-data (c-forward-keyword-clause 1))
6507 (if pos
6508 (goto-char pos)
6509 (goto-char (match-end 1))
6510 (c-forward-syntactic-ws)))))
6511
6512 (name-res
6513 (cond ((eq name-res t)
6514 ;; A normal identifier.
6515 (goto-char id-end)
6516 (if (or res c-promote-possible-types)
6517 (progn
6518 (c-add-type id-start id-end)
6519 (when (and c-record-type-identifiers id-range)
6520 (c-record-type-id id-range))
6521 (unless res
6522 (setq res 'found)))
6523 (setq res (if (c-check-type id-start id-end)
6524 ;; It's an identifier that has been used as
6525 ;; a type somewhere else.
6526 'found
6527 ;; It's an identifier that might be a type.
6528 'maybe))))
6529 ((eq name-res 'template)
6530 ;; A template is a type.
6531 (goto-char id-end)
6532 (setq res t))
6533 (t
6534 ;; Otherwise it's an operator identifier, which is not a type.
6535 (goto-char start)
6536 (setq res nil)))))
6537
6538 (when res
6539 ;; Skip trailing type modifiers. If any are found we know it's
6540 ;; a type.
6541 (when c-opt-type-modifier-key
6542 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
6543 (goto-char (match-end 1))
6544 (c-forward-syntactic-ws)
6545 (setq res t)))
6546
6547 ;; Step over any type suffix operator. Do not let the existence
6548 ;; of these alter the classification of the found type, since
6549 ;; these operators typically are allowed in normal expressions
6550 ;; too.
6551 (when c-opt-type-suffix-key ; e.g. "..."
6552 (while (looking-at c-opt-type-suffix-key)
6553 (goto-char (match-end 1))
6554 (c-forward-syntactic-ws)))
6555
6556 (when c-opt-type-concat-key ; Only/mainly for pike.
6557 ;; Look for a trailing operator that concatenates the type
6558 ;; with a following one, and if so step past that one through
6559 ;; a recursive call. Note that we don't record concatenated
6560 ;; types in `c-found-types' - it's the component types that
6561 ;; are recorded when appropriate.
6562 (setq pos (point))
6563 (let* ((c-promote-possible-types (or (memq res '(t known))
6564 c-promote-possible-types))
6565 ;; If we can't promote then set `c-record-found-types' so that
6566 ;; we can merge in the types from the second part afterwards if
6567 ;; it turns out to be a known type there.
6568 (c-record-found-types (and c-record-type-identifiers
6569 (not c-promote-possible-types)))
6570 subres)
6571 (if (and (looking-at c-opt-type-concat-key)
6572
6573 (progn
6574 (goto-char (match-end 1))
6575 (c-forward-syntactic-ws)
6576 (setq subres (c-forward-type))))
6577
6578 (progn
6579 ;; If either operand certainly is a type then both are, but we
6580 ;; don't let the existence of the operator itself promote two
6581 ;; uncertain types to a certain one.
6582 (cond ((eq res t))
6583 ((eq subres t)
6584 (unless (eq name-res 'template)
6585 (c-add-type id-start id-end))
6586 (when (and c-record-type-identifiers id-range)
6587 (c-record-type-id id-range))
6588 (setq res t))
6589 ((eq res 'known))
6590 ((eq subres 'known)
6591 (setq res 'known))
6592 ((eq res 'found))
6593 ((eq subres 'found)
6594 (setq res 'found))
6595 (t
6596 (setq res 'maybe)))
6597
6598 (when (and (eq res t)
6599 (consp c-record-found-types))
6600 ;; Merge in the ranges of any types found by the second
6601 ;; `c-forward-type'.
6602 (setq c-record-type-identifiers
6603 ;; `nconc' doesn't mind that the tail of
6604 ;; `c-record-found-types' is t.
6605 (nconc c-record-found-types
6606 c-record-type-identifiers))))
6607
6608 (goto-char pos))))
6609
6610 (when (and c-record-found-types (memq res '(known found)) id-range)
6611 (setq c-record-found-types
6612 (cons id-range c-record-found-types))))
6613
6614 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6615
6616 res))
6617
6618 (defun c-forward-annotation ()
6619 ;; Used for Java code only at the moment. Assumes point is on the
6620 ;; @, moves forward an annotation. returns nil if there is no
6621 ;; annotation at point.
6622 (and (looking-at "@")
6623 (progn (forward-char) t)
6624 (c-forward-type)
6625 (progn (c-forward-syntactic-ws) t)
6626 (if (looking-at "(")
6627 (c-go-list-forward)
6628 t)))
6629
6630 (defmacro c-pull-open-brace (ps)
6631 ;; Pull the next open brace from PS (which has the form of paren-state),
6632 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
6633 `(progn
6634 (while (consp (car ,ps))
6635 (setq ,ps (cdr ,ps)))
6636 (prog1 (car ,ps)
6637 (setq ,ps (cdr ,ps)))))
6638
6639 (defun c-back-over-member-initializer-braces ()
6640 ;; Point is just after a closing brace/parenthesis. Try to parse this as a
6641 ;; C++ member initializer list, going back to just after the introducing ":"
6642 ;; and returning t. Otherwise return nil, leaving point unchanged.
6643 (let ((here (point)) res)
6644 (setq res
6645 (catch 'done
6646 (when (not (c-go-list-backward))
6647 (throw 'done nil))
6648 (c-backward-syntactic-ws)
6649 (when (not (c-simple-skip-symbol-backward))
6650 (throw 'done nil))
6651 (c-backward-syntactic-ws)
6652
6653 (while (eq (char-before) ?,)
6654 (backward-char)
6655 (c-backward-syntactic-ws)
6656 (when (not (memq (char-before) '(?\) ?})))
6657 (throw 'done nil))
6658 (when (not (c-go-list-backward))
6659 (throw 'done nil))
6660 (c-backward-syntactic-ws)
6661 (when (not (c-simple-skip-symbol-backward))
6662 (throw 'done nil))
6663 (c-backward-syntactic-ws))
6664
6665 (eq (char-before) ?:)))
6666 (or res (goto-char here))
6667 res))
6668
6669 (defmacro c-back-over-list-of-member-inits ()
6670 ;; Go back over a list of elements, each looking like:
6671 ;; <symbol> (<expression>) ,
6672 ;; or <symbol> {<expression>} ,
6673 ;; when we are putatively immediately after a comma. Stop when we don't see
6674 ;; a comma. If either of <symbol> or bracketed <expression> is missing,
6675 ;; throw nil to 'level. If the terminating } or ) is unmatched, throw nil
6676 ;; to 'done. This is not a general purpose macro!
6677 `(while (eq (char-before) ?,)
6678 (backward-char)
6679 (c-backward-syntactic-ws)
6680 (when (not (memq (char-before) '(?\) ?})))
6681 (throw 'level nil))
6682 (when (not (c-go-list-backward))
6683 (throw 'done nil))
6684 (c-backward-syntactic-ws)
6685 (when (not (c-simple-skip-symbol-backward))
6686 (throw 'level nil))
6687 (c-backward-syntactic-ws)))
6688
6689 (defun c-back-over-member-initializers ()
6690 ;; Test whether we are in a C++ member initializer list, and if so, go back
6691 ;; to the introducing ":", returning the position of the opening paren of
6692 ;; the function's arglist. Otherwise return nil, leaving point unchanged.
6693 (let ((here (point))
6694 (paren-state (c-parse-state))
6695 pos level-plausible at-top-level res)
6696 ;; Assume tentatively that we're at the top level. Try to go back to the
6697 ;; colon we seek.
6698 (setq res
6699 (catch 'done
6700 (setq level-plausible
6701 (catch 'level
6702 (c-backward-syntactic-ws)
6703 (when (memq (char-before) '(?\) ?}))
6704 (when (not (c-go-list-backward))
6705 (throw 'done nil))
6706 (c-backward-syntactic-ws))
6707 (when (c-simple-skip-symbol-backward)
6708 (c-backward-syntactic-ws))
6709 (c-back-over-list-of-member-inits)
6710 (and (eq (char-before) ?:)
6711 (c-just-after-func-arglist-p))))
6712
6713 (while (and (not (and level-plausible
6714 (setq at-top-level (c-at-toplevel-p))))
6715 (setq pos (c-pull-open-brace paren-state))) ; might be a paren.
6716 (setq level-plausible
6717 (catch 'level
6718 (goto-char pos)
6719 (c-backward-syntactic-ws)
6720 (when (not (c-simple-skip-symbol-backward))
6721 (throw 'level nil))
6722 (c-backward-syntactic-ws)
6723 (c-back-over-list-of-member-inits)
6724 (and (eq (char-before) ?:)
6725 (c-just-after-func-arglist-p)))))
6726
6727 (and at-top-level level-plausible)))
6728 (or res (goto-char here))
6729 res))
6730
6731 \f
6732 ;; Handling of large scale constructs like statements and declarations.
6733
6734 ;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6735 ;; defsubst or perhaps even a defun, but it contains lots of free
6736 ;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6737 (defmacro c-fdoc-shift-type-backward (&optional short)
6738 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6739 ;; of types when parsing a declaration, which means that it
6740 ;; sometimes consumes the identifier in the declaration as a type.
6741 ;; This is used to "backtrack" and make the last type be treated as
6742 ;; an identifier instead.
6743 `(progn
6744 ,(unless short
6745 ;; These identifiers are bound only in the inner let.
6746 '(setq identifier-type at-type
6747 identifier-start type-start
6748 got-parens nil
6749 got-identifier t
6750 got-suffix t
6751 got-suffix-after-parens id-start
6752 paren-depth 0))
6753
6754 (if (setq at-type (if (eq backup-at-type 'prefix)
6755 t
6756 backup-at-type))
6757 (setq type-start backup-type-start
6758 id-start backup-id-start)
6759 (setq type-start start-pos
6760 id-start start-pos))
6761
6762 ;; When these flags already are set we've found specifiers that
6763 ;; unconditionally signal these attributes - backtracking doesn't
6764 ;; change that. So keep them set in that case.
6765 (or at-type-decl
6766 (setq at-type-decl backup-at-type-decl))
6767 (or maybe-typeless
6768 (setq maybe-typeless backup-maybe-typeless))
6769
6770 ,(unless short
6771 ;; This identifier is bound only in the inner let.
6772 '(setq start id-start))))
6773
6774 (defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6775 ;; Move forward over a declaration or a cast if at the start of one.
6776 ;; The point is assumed to be at the start of some token. Nil is
6777 ;; returned if no declaration or cast is recognized, and the point
6778 ;; is clobbered in that case.
6779 ;;
6780 ;; If a declaration is parsed:
6781 ;;
6782 ;; The point is left at the first token after the first complete
6783 ;; declarator, if there is one. The return value is a cons where
6784 ;; the car is the position of the first token in the declarator. (See
6785 ;; below for the cdr.)
6786 ;; Some examples:
6787 ;;
6788 ;; void foo (int a, char *b) stuff ...
6789 ;; car ^ ^ point
6790 ;; float (*a)[], b;
6791 ;; car ^ ^ point
6792 ;; unsigned int a = c_style_initializer, b;
6793 ;; car ^ ^ point
6794 ;; unsigned int a (cplusplus_style_initializer), b;
6795 ;; car ^ ^ point (might change)
6796 ;; class Foo : public Bar {}
6797 ;; car ^ ^ point
6798 ;; class PikeClass (int a, string b) stuff ...
6799 ;; car ^ ^ point
6800 ;; enum bool;
6801 ;; car ^ ^ point
6802 ;; enum bool flag;
6803 ;; car ^ ^ point
6804 ;; void cplusplus_function (int x) throw (Bad);
6805 ;; car ^ ^ point
6806 ;; Foo::Foo (int b) : Base (b) {}
6807 ;; car ^ ^ point
6808 ;;
6809 ;; auto foo = 5;
6810 ;; car ^ ^ point
6811 ;; auto cplusplus_11 (int a, char *b) -> decltype (bar):
6812 ;; car ^ ^ point
6813 ;;
6814 ;;
6815 ;;
6816 ;; The cdr of the return value is non-nil when a
6817 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6818 ;; Specifically it is a dotted pair (A . B) where B is t when a
6819 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6820 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6821 ;; specifier is present. I.e., (some of) the declared
6822 ;; identifier(s) are types.
6823 ;;
6824 ;; If a cast is parsed:
6825 ;;
6826 ;; The point is left at the first token after the closing paren of
6827 ;; the cast. The return value is `cast'. Note that the start
6828 ;; position must be at the first token inside the cast parenthesis
6829 ;; to recognize it.
6830 ;;
6831 ;; PRECEDING-TOKEN-END is the first position after the preceding
6832 ;; token, i.e. on the other side of the syntactic ws from the point.
6833 ;; Use a value less than or equal to (point-min) if the point is at
6834 ;; the first token in (the visible part of) the buffer.
6835 ;;
6836 ;; CONTEXT is a symbol that describes the context at the point:
6837 ;; 'decl In a comma-separated declaration context (typically
6838 ;; inside a function declaration arglist).
6839 ;; '<> In an angle bracket arglist.
6840 ;; 'arglist Some other type of arglist.
6841 ;; nil Some other context or unknown context. Includes
6842 ;; within the parens of an if, for, ... construct.
6843 ;;
6844 ;; LAST-CAST-END is the first token after the closing paren of a
6845 ;; preceding cast, or nil if none is known. If
6846 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6847 ;; the position after the closest preceding call where a cast was
6848 ;; matched. In that case it's used to discover chains of casts like
6849 ;; "(a) (b) c".
6850 ;;
6851 ;; This function records identifier ranges on
6852 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6853 ;; `c-record-type-identifiers' is non-nil.
6854 ;;
6855 ;; This function might do hidden buffer changes.
6856
6857 (let (;; `start-pos' is used below to point to the start of the
6858 ;; first type, i.e. after any leading specifiers. It might
6859 ;; also point at the beginning of the preceding syntactic
6860 ;; whitespace.
6861 (start-pos (point))
6862 ;; Set to the result of `c-forward-type'.
6863 at-type
6864 ;; The position of the first token in what we currently
6865 ;; believe is the type in the declaration or cast, after any
6866 ;; specifiers and their associated clauses.
6867 type-start
6868 ;; The position of the first token in what we currently
6869 ;; believe is the declarator for the first identifier. Set
6870 ;; when the type is found, and moved forward over any
6871 ;; `c-decl-hangon-kwds' and their associated clauses that
6872 ;; occurs after the type.
6873 id-start
6874 ;; These store `at-type', `type-start' and `id-start' of the
6875 ;; identifier before the one in those variables. The previous
6876 ;; identifier might turn out to be the real type in a
6877 ;; declaration if the last one has to be the declarator in it.
6878 ;; If `backup-at-type' is nil then the other variables have
6879 ;; undefined values.
6880 backup-at-type backup-type-start backup-id-start
6881 ;; This stores `kwd-sym' of the symbol before the current one.
6882 ;; This is needed to distinguish the C++11 version of "auto" from
6883 ;; the pre C++11 meaning.
6884 backup-kwd-sym
6885 ;; Set if we've found a specifier (apart from "typedef") that makes
6886 ;; the defined identifier(s) types.
6887 at-type-decl
6888 ;; Set if we've a "typedef" keyword.
6889 at-typedef
6890 ;; Set if we've found a specifier that can start a declaration
6891 ;; where there's no type.
6892 maybe-typeless
6893 ;; Save the value of kwd-sym between loops of the "Check for a
6894 ;; type" loop. Needed to distinguish a C++11 "auto" from a pre
6895 ;; C++11 one.
6896 prev-kwd-sym
6897 ;; If a specifier is found that also can be a type prefix,
6898 ;; these flags are set instead of those above. If we need to
6899 ;; back up an identifier, they are copied to the real flag
6900 ;; variables. Thus they only take effect if we fail to
6901 ;; interpret it as a type.
6902 backup-at-type-decl backup-maybe-typeless
6903 ;; Whether we've found a declaration or a cast. We might know
6904 ;; this before we've found the type in it. It's 'ids if we've
6905 ;; found two consecutive identifiers (usually a sure sign, but
6906 ;; we should allow that in labels too), and t if we've found a
6907 ;; specifier keyword (a 100% sure sign).
6908 at-decl-or-cast
6909 ;; Set when we need to back up to parse this as a declaration
6910 ;; but not as a cast.
6911 backup-if-not-cast
6912 ;; For casts, the return position.
6913 cast-end
6914 ;; Have we got a new-style C++11 "auto"?
6915 new-style-auto
6916 ;; Save `c-record-type-identifiers' and
6917 ;; `c-record-ref-identifiers' since ranges are recorded
6918 ;; speculatively and should be thrown away if it turns out
6919 ;; that it isn't a declaration or cast.
6920 (save-rec-type-ids c-record-type-identifiers)
6921 (save-rec-ref-ids c-record-ref-identifiers))
6922
6923 (while (c-forward-annotation)
6924 (c-forward-syntactic-ws))
6925
6926 ;; Check for a type. Unknown symbols are treated as possible
6927 ;; types, but they could also be specifiers disguised through
6928 ;; macros like __INLINE__, so we recognize both types and known
6929 ;; specifiers after them too.
6930 (while
6931 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6932
6933 ;; Look for a specifier keyword clause.
6934 (when (or (looking-at c-prefix-spec-kwds-re) ;FIXME!!! includes auto
6935 (and (c-major-mode-is 'java-mode)
6936 (looking-at "@[A-Za-z0-9]+")))
6937 (save-match-data
6938 (if (looking-at c-typedef-key)
6939 (setq at-typedef t)))
6940 (setq kwd-sym (c-keyword-sym (match-string 1)))
6941 (save-excursion
6942 (c-forward-keyword-clause 1)
6943 (setq kwd-clause-end (point))))
6944
6945 (when (setq found-type (c-forward-type t)) ; brace-block-too
6946 ;; Found a known or possible type or a prefix of a known type.
6947 (when (and (c-major-mode-is 'c++-mode) ; C++11 style "auto"?
6948 (eq prev-kwd-sym (c-keyword-sym "auto"))
6949 (looking-at "[=(]")) ; FIXME!!! proper regexp.
6950 (setq new-style-auto t)
6951 (setq found-type nil)
6952 (goto-char start)) ; position of foo in "auto foo"
6953
6954 (when at-type
6955 ;; Got two identifiers with nothing but whitespace
6956 ;; between them. That can only happen in declarations.
6957 (setq at-decl-or-cast 'ids)
6958
6959 (when (eq at-type 'found)
6960 ;; If the previous identifier is a found type we
6961 ;; record it as a real one; it might be some sort of
6962 ;; alias for a prefix like "unsigned".
6963 (save-excursion
6964 (goto-char type-start)
6965 (let ((c-promote-possible-types t))
6966 (c-forward-type)))))
6967
6968 (setq backup-at-type at-type
6969 backup-type-start type-start
6970 backup-id-start id-start
6971 backup-kwd-sym kwd-sym
6972 at-type found-type
6973 type-start start
6974 id-start (point)
6975 ;; The previous ambiguous specifier/type turned out
6976 ;; to be a type since we've parsed another one after
6977 ;; it, so clear these backup flags.
6978 backup-at-type-decl nil
6979 backup-maybe-typeless nil))
6980
6981 (if kwd-sym
6982 (progn
6983 ;; Handle known specifier keywords and
6984 ;; `c-decl-hangon-kwds' which can occur after known
6985 ;; types.
6986
6987 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6988 ;; It's a hang-on keyword that can occur anywhere.
6989 (progn
6990 (setq at-decl-or-cast t)
6991 (if at-type
6992 ;; Move the identifier start position if
6993 ;; we've passed a type.
6994 (setq id-start kwd-clause-end)
6995 ;; Otherwise treat this as a specifier and
6996 ;; move the fallback position.
6997 (setq start-pos kwd-clause-end))
6998 (goto-char kwd-clause-end))
6999
7000 ;; It's an ordinary specifier so we know that
7001 ;; anything before this can't be the type.
7002 (setq backup-at-type nil
7003 start-pos kwd-clause-end)
7004
7005 (if found-type
7006 ;; It's ambiguous whether this keyword is a
7007 ;; specifier or a type prefix, so set the backup
7008 ;; flags. (It's assumed that `c-forward-type'
7009 ;; moved further than `c-forward-keyword-clause'.)
7010 (progn
7011 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
7012 (setq backup-at-type-decl t))
7013 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
7014 (setq backup-maybe-typeless t)))
7015
7016 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
7017 ;; This test only happens after we've scanned a type.
7018 ;; So, with valid syntax, kwd-sym can't be 'typedef.
7019 (setq at-type-decl t))
7020 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
7021 (setq maybe-typeless t))
7022
7023 ;; Haven't matched a type so it's an unambiguous
7024 ;; specifier keyword and we know we're in a
7025 ;; declaration.
7026 (setq at-decl-or-cast t)
7027 (setq prev-kwd-sym kwd-sym)
7028
7029 (goto-char kwd-clause-end))))
7030
7031 ;; If the type isn't known we continue so that we'll jump
7032 ;; over all specifiers and type identifiers. The reason
7033 ;; to do this for a known type prefix is to make things
7034 ;; like "unsigned INT16" work.
7035 (and found-type (not (eq found-type t))))))
7036
7037 (cond
7038 ((eq at-type t)
7039 ;; If a known type was found, we still need to skip over any
7040 ;; hangon keyword clauses after it. Otherwise it has already
7041 ;; been done in the loop above.
7042 (while (looking-at c-decl-hangon-key)
7043 (c-forward-keyword-clause 1))
7044 (setq id-start (point)))
7045
7046 ((eq at-type 'prefix)
7047 ;; A prefix type is itself a primitive type when it's not
7048 ;; followed by another type.
7049 (setq at-type t))
7050
7051 ((not at-type)
7052 ;; Got no type but set things up to continue anyway to handle
7053 ;; the various cases when a declaration doesn't start with a
7054 ;; type.
7055 (setq id-start start-pos))
7056
7057 ((and (eq at-type 'maybe)
7058 (c-major-mode-is 'c++-mode))
7059 ;; If it's C++ then check if the last "type" ends on the form
7060 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
7061 ;; (con|de)structor.
7062 (save-excursion
7063 (let (name end-2 end-1)
7064 (goto-char id-start)
7065 (c-backward-syntactic-ws)
7066 (setq end-2 (point))
7067 (when (and
7068 (c-simple-skip-symbol-backward)
7069 (progn
7070 (setq name
7071 (buffer-substring-no-properties (point) end-2))
7072 ;; Cheating in the handling of syntactic ws below.
7073 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
7074 (progn
7075 (setq end-1 (point))
7076 (c-simple-skip-symbol-backward))
7077 (>= (point) type-start)
7078 (equal (buffer-substring-no-properties (point) end-1)
7079 name))
7080 ;; It is a (con|de)structor name. In that case the
7081 ;; declaration is typeless so zap out any preceding
7082 ;; identifier(s) that we might have taken as types.
7083 (goto-char type-start)
7084 (setq at-type nil
7085 backup-at-type nil
7086 id-start type-start))))))
7087
7088 ;; Check for and step over a type decl expression after the thing
7089 ;; that is or might be a type. This can't be skipped since we
7090 ;; need the correct end position of the declarator for
7091 ;; `max-type-decl-end-*'.
7092 (let ((start (point)) (paren-depth 0) pos
7093 ;; True if there's a non-open-paren match of
7094 ;; `c-type-decl-prefix-key'.
7095 got-prefix
7096 ;; True if the declarator is surrounded by a parenthesis pair.
7097 got-parens
7098 ;; True if there is an identifier in the declarator.
7099 got-identifier
7100 ;; True if there's a non-close-paren match of
7101 ;; `c-type-decl-suffix-key'.
7102 got-suffix
7103 ;; True if there's a prefix match outside the outermost
7104 ;; paren pair that surrounds the declarator.
7105 got-prefix-before-parens
7106 ;; True if there's a suffix match outside the outermost
7107 ;; paren pair that surrounds the declarator. The value is
7108 ;; the position of the first suffix match.
7109 got-suffix-after-parens
7110 ;; True if we've parsed the type decl to a token that is
7111 ;; known to end declarations in this context.
7112 at-decl-end
7113 ;; The earlier values of `at-type' and `type-start' if we've
7114 ;; shifted the type backwards.
7115 identifier-type identifier-start
7116 ;; If `c-parse-and-markup-<>-arglists' is set we need to
7117 ;; turn it off during the name skipping below to avoid
7118 ;; getting `c-type' properties that might be bogus. That
7119 ;; can happen since we don't know if
7120 ;; `c-restricted-<>-arglists' will be correct inside the
7121 ;; arglist paren that gets entered.
7122 c-parse-and-markup-<>-arglists
7123 ;; Start of the identifier for which `got-identifier' was set.
7124 name-start)
7125
7126 (goto-char id-start)
7127
7128 ;; Skip over type decl prefix operators. (Note similar code in
7129 ;; `c-font-lock-declarators'.)
7130 (if (and c-recognize-typeless-decls
7131 (equal c-type-decl-prefix-key "\\<\\>"))
7132 (when (eq (char-after) ?\()
7133 (progn
7134 (setq paren-depth (1+ paren-depth))
7135 (forward-char)))
7136 (while (and (looking-at c-type-decl-prefix-key)
7137 (if (and (c-major-mode-is 'c++-mode)
7138 (match-beginning 3))
7139 ;; If the third submatch matches in C++ then
7140 ;; we're looking at an identifier that's a
7141 ;; prefix only if it specifies a member pointer.
7142 (when (progn (setq pos (point))
7143 (setq got-identifier (c-forward-name)))
7144 (setq name-start pos)
7145 (if (looking-at "\\(::\\)")
7146 ;; We only check for a trailing "::" and
7147 ;; let the "*" that should follow be
7148 ;; matched in the next round.
7149 (progn (setq got-identifier nil) t)
7150 ;; It turned out to be the real identifier,
7151 ;; so stop.
7152 nil))
7153 t))
7154
7155 (if (eq (char-after) ?\()
7156 (progn
7157 (setq paren-depth (1+ paren-depth))
7158 (forward-char))
7159 (unless got-prefix-before-parens
7160 (setq got-prefix-before-parens (= paren-depth 0)))
7161 (setq got-prefix t)
7162 (goto-char (match-end 1)))
7163 (c-forward-syntactic-ws)))
7164
7165 (setq got-parens (> paren-depth 0))
7166
7167 ;; Skip over an identifier.
7168 (or got-identifier
7169 (and (looking-at c-identifier-start)
7170 (setq pos (point))
7171 (setq got-identifier (c-forward-name))
7172 (setq name-start pos)))
7173
7174 ;; Skip over type decl suffix operators.
7175 (while (if (looking-at c-type-decl-suffix-key)
7176
7177 (if (eq (char-after) ?\))
7178 (when (> paren-depth 0)
7179 (setq paren-depth (1- paren-depth))
7180 (forward-char)
7181 t)
7182 (when (if (save-match-data (looking-at "\\s("))
7183 (c-safe (c-forward-sexp 1) t)
7184 (goto-char (match-end 1))
7185 t)
7186 (when (and (not got-suffix-after-parens)
7187 (= paren-depth 0))
7188 (setq got-suffix-after-parens (match-beginning 0)))
7189 (setq got-suffix t)))
7190
7191 ;; No suffix matched. We might have matched the
7192 ;; identifier as a type and the open paren of a
7193 ;; function arglist as a type decl prefix. In that
7194 ;; case we should "backtrack": Reinterpret the last
7195 ;; type as the identifier, move out of the arglist and
7196 ;; continue searching for suffix operators.
7197 ;;
7198 ;; Do this even if there's no preceding type, to cope
7199 ;; with old style function declarations in K&R C,
7200 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
7201 ;; style declarations. That isn't applicable in an
7202 ;; arglist context, though.
7203 (when (and (= paren-depth 1)
7204 (not got-prefix-before-parens)
7205 (not (eq at-type t))
7206 (or backup-at-type
7207 maybe-typeless
7208 backup-maybe-typeless
7209 (when c-recognize-typeless-decls
7210 (not context)))
7211 (setq pos (c-up-list-forward (point)))
7212 (eq (char-before pos) ?\)))
7213 (c-fdoc-shift-type-backward)
7214 (goto-char pos)
7215 t))
7216
7217 (c-forward-syntactic-ws))
7218
7219 (when (or (and new-style-auto
7220 (looking-at c-auto-ops-re))
7221 (and (or maybe-typeless backup-maybe-typeless)
7222 (not got-identifier)
7223 (not got-prefix)
7224 at-type))
7225 ;; Have found no identifier but `c-typeless-decl-kwds' has
7226 ;; matched so we know we're inside a declaration. The
7227 ;; preceding type must be the identifier instead.
7228 (c-fdoc-shift-type-backward))
7229
7230 ;; Prepare the "-> type;" for fontification later on.
7231 (when (and new-style-auto
7232 (looking-at c-haskell-op-re))
7233 (save-excursion
7234 (goto-char (match-end 0))
7235 (c-forward-syntactic-ws)
7236 (setq type-start (point))
7237 (setq at-type (c-forward-type))))
7238
7239 (setq
7240 at-decl-or-cast
7241 (catch 'at-decl-or-cast
7242
7243 ;; CASE 1
7244 (when (> paren-depth 0)
7245 ;; Encountered something inside parens that isn't matched by
7246 ;; the `c-type-decl-*' regexps, so it's not a type decl
7247 ;; expression. Try to skip out to the same paren depth to
7248 ;; not confuse the cast check below.
7249 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
7250 ;; If we've found a specifier keyword then it's a
7251 ;; declaration regardless.
7252 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
7253
7254 (setq at-decl-end
7255 (looking-at (cond ((eq context '<>) "[,>]")
7256 (context "[,)]")
7257 (t "[,;]"))))
7258
7259 ;; Now we've collected info about various characteristics of
7260 ;; the construct we're looking at. Below follows a decision
7261 ;; tree based on that. It's ordered to check more certain
7262 ;; signs before less certain ones.
7263
7264 (if got-identifier
7265 (progn
7266
7267 ;; CASE 2
7268 (when (and (or at-type maybe-typeless)
7269 (not (or got-prefix got-parens)))
7270 ;; Got another identifier directly after the type, so it's a
7271 ;; declaration.
7272 (throw 'at-decl-or-cast t))
7273
7274 (when (and got-parens
7275 (not got-prefix)
7276 ;; (not got-suffix-after-parens)
7277 (or backup-at-type
7278 maybe-typeless
7279 backup-maybe-typeless
7280 (eq at-decl-or-cast t)
7281 (save-excursion
7282 (goto-char name-start)
7283 (not (memq (c-forward-type) '(nil maybe))))))
7284 ;; Got a declaration of the form "foo bar (gnu);" or "bar
7285 ;; (gnu);" where we've recognized "bar" as the type and "gnu"
7286 ;; as the declarator. In this case it's however more likely
7287 ;; that "bar" is the declarator and "gnu" a function argument
7288 ;; or initializer (if `c-recognize-paren-inits' is set),
7289 ;; since the parens around "gnu" would be superfluous if it's
7290 ;; a declarator. Shift the type one step backward.
7291 (c-fdoc-shift-type-backward)))
7292
7293 ;; Found no identifier.
7294
7295 (if backup-at-type
7296 (progn
7297
7298 ;; CASE 3
7299 (when (= (point) start)
7300 ;; Got a plain list of identifiers. If a colon follows it's
7301 ;; a valid label, or maybe a bitfield. Otherwise the last
7302 ;; one probably is the declared identifier and we should
7303 ;; back up to the previous type, providing it isn't a cast.
7304 (if (and (eq (char-after) ?:)
7305 (not (c-major-mode-is 'java-mode)))
7306 (cond
7307 ;; If we've found a specifier keyword then it's a
7308 ;; declaration regardless.
7309 ((eq at-decl-or-cast t)
7310 (throw 'at-decl-or-cast t))
7311 ((and c-has-bitfields
7312 (eq at-decl-or-cast 'ids)) ; bitfield.
7313 (setq backup-if-not-cast t)
7314 (throw 'at-decl-or-cast t)))
7315
7316 (setq backup-if-not-cast t)
7317 (throw 'at-decl-or-cast t)))
7318
7319 ;; CASE 4
7320 (when (and got-suffix
7321 (not got-prefix)
7322 (not got-parens))
7323 ;; Got a plain list of identifiers followed by some suffix.
7324 ;; If this isn't a cast then the last identifier probably is
7325 ;; the declared one and we should back up to the previous
7326 ;; type.
7327 (setq backup-if-not-cast t)
7328 (throw 'at-decl-or-cast t)))
7329
7330 ;; CASE 5
7331 (when (eq at-type t)
7332 ;; If the type is known we know that there can't be any
7333 ;; identifier somewhere else, and it's only in declarations in
7334 ;; e.g. function prototypes and in casts that the identifier may
7335 ;; be left out.
7336 (throw 'at-decl-or-cast t))
7337
7338 (when (= (point) start)
7339 ;; Only got a single identifier (parsed as a type so far).
7340 ;; CASE 6
7341 (if (and
7342 ;; Check that the identifier isn't at the start of an
7343 ;; expression.
7344 at-decl-end
7345 (cond
7346 ((eq context 'decl)
7347 ;; Inside an arglist that contains declarations. If K&R
7348 ;; style declarations and parenthesis style initializers
7349 ;; aren't allowed then the single identifier must be a
7350 ;; type, else we require that it's known or found
7351 ;; (primitive types are handled above).
7352 (or (and (not c-recognize-knr-p)
7353 (not c-recognize-paren-inits))
7354 (memq at-type '(known found))))
7355 ((eq context '<>)
7356 ;; Inside a template arglist. Accept known and found
7357 ;; types; other identifiers could just as well be
7358 ;; constants in C++.
7359 (memq at-type '(known found)))))
7360 (throw 'at-decl-or-cast t)
7361 ;; CASE 7
7362 ;; Can't be a valid declaration or cast, but if we've found a
7363 ;; specifier it can't be anything else either, so treat it as
7364 ;; an invalid/unfinished declaration or cast.
7365 (throw 'at-decl-or-cast at-decl-or-cast))))
7366
7367 (if (and got-parens
7368 (not got-prefix)
7369 (not context)
7370 (not (eq at-type t))
7371 (or backup-at-type
7372 maybe-typeless
7373 backup-maybe-typeless
7374 (when c-recognize-typeless-decls
7375 (or (not got-suffix)
7376 (not (looking-at
7377 c-after-suffixed-type-maybe-decl-key))))))
7378 ;; Got an empty paren pair and a preceding type that probably
7379 ;; really is the identifier. Shift the type backwards to make
7380 ;; the last one the identifier. This is analogous to the
7381 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
7382 ;; above.
7383 ;;
7384 ;; Exception: In addition to the conditions in that
7385 ;; "backtracking" code, do not shift backward if we're not
7386 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
7387 ;; Since there's no preceding type, the shift would mean that
7388 ;; the declaration is typeless. But if the regexp doesn't match
7389 ;; then we will simply fall through in the tests below and not
7390 ;; recognize it at all, so it's better to try it as an abstract
7391 ;; declarator instead.
7392 (c-fdoc-shift-type-backward)
7393
7394 ;; Still no identifier.
7395 ;; CASE 8
7396 (when (and got-prefix (or got-parens got-suffix))
7397 ;; Require `got-prefix' together with either `got-parens' or
7398 ;; `got-suffix' to recognize it as an abstract declarator:
7399 ;; `got-parens' only is probably an empty function call.
7400 ;; `got-suffix' only can build an ordinary expression together
7401 ;; with the preceding identifier which we've taken as a type.
7402 ;; We could actually accept on `got-prefix' only, but that can
7403 ;; easily occur temporarily while writing an expression so we
7404 ;; avoid that case anyway. We could do a better job if we knew
7405 ;; the point when the fontification was invoked.
7406 (throw 'at-decl-or-cast t))
7407
7408 ;; CASE 9
7409 (when (and at-type
7410 (not got-prefix)
7411 (not got-parens)
7412 got-suffix-after-parens
7413 (eq (char-after got-suffix-after-parens) ?\())
7414 ;; Got a type, no declarator but a paren suffix. I.e. it's a
7415 ;; normal function call after all (or perhaps a C++ style object
7416 ;; instantiation expression).
7417 (throw 'at-decl-or-cast nil))))
7418
7419 ;; CASE 10
7420 (when at-decl-or-cast
7421 ;; By now we've located the type in the declaration that we know
7422 ;; we're in.
7423 (throw 'at-decl-or-cast t))
7424
7425 ;; CASE 11
7426 (when (and got-identifier
7427 (not context)
7428 (looking-at c-after-suffixed-type-decl-key)
7429 (if (and got-parens
7430 (not got-prefix)
7431 (not got-suffix)
7432 (not (eq at-type t)))
7433 ;; Shift the type backward in the case that there's a
7434 ;; single identifier inside parens. That can only
7435 ;; occur in K&R style function declarations so it's
7436 ;; more likely that it really is a function call.
7437 ;; Therefore we only do this after
7438 ;; `c-after-suffixed-type-decl-key' has matched.
7439 (progn (c-fdoc-shift-type-backward) t)
7440 got-suffix-after-parens))
7441 ;; A declaration according to `c-after-suffixed-type-decl-key'.
7442 (throw 'at-decl-or-cast t))
7443
7444 ;; CASE 12
7445 (when (and (or got-prefix (not got-parens))
7446 (memq at-type '(t known)))
7447 ;; It's a declaration if a known type precedes it and it can't be a
7448 ;; function call.
7449 (throw 'at-decl-or-cast t))
7450
7451 ;; If we get here we can't tell if this is a type decl or a normal
7452 ;; expression by looking at it alone. (That's under the assumption
7453 ;; that normal expressions always can look like type decl expressions,
7454 ;; which isn't really true but the cases where it doesn't hold are so
7455 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
7456 ;; the effort to look for them.)
7457
7458 ;;; 2008-04-16: commented out the next form, to allow the function to recognize
7459 ;;; "foo (int bar)" in CC (an implicit type (in class foo) without a semicolon)
7460 ;;; as a(n almost complete) declaration, enabling it to be fontified.
7461 ;; CASE 13
7462 ;; (unless (or at-decl-end (looking-at "=[^=]"))
7463 ;; If this is a declaration it should end here or its initializer(*)
7464 ;; should start here, so check for allowed separation tokens. Note
7465 ;; that this rule doesn't work e.g. with a K&R arglist after a
7466 ;; function header.
7467 ;;
7468 ;; *) Don't check for C++ style initializers using parens
7469 ;; since those already have been matched as suffixes.
7470 ;;
7471 ;; If `at-decl-or-cast' is then we've found some other sign that
7472 ;; it's a declaration or cast, so then it's probably an
7473 ;; invalid/unfinished one.
7474 ;; (throw 'at-decl-or-cast at-decl-or-cast))
7475
7476 ;; Below are tests that only should be applied when we're certain to
7477 ;; not have parsed halfway through an expression.
7478
7479 ;; CASE 14
7480 (when (memq at-type '(t known))
7481 ;; The expression starts with a known type so treat it as a
7482 ;; declaration.
7483 (throw 'at-decl-or-cast t))
7484
7485 ;; CASE 15
7486 (when (and (c-major-mode-is 'c++-mode)
7487 ;; In C++ we check if the identifier is a known type, since
7488 ;; (con|de)structors use the class name as identifier.
7489 ;; We've always shifted over the identifier as a type and
7490 ;; then backed up again in this case.
7491 identifier-type
7492 (or (memq identifier-type '(found known))
7493 (and (eq (char-after identifier-start) ?~)
7494 ;; `at-type' probably won't be 'found for
7495 ;; destructors since the "~" is then part of the
7496 ;; type name being checked against the list of
7497 ;; known types, so do a check without that
7498 ;; operator.
7499 (or (save-excursion
7500 (goto-char (1+ identifier-start))
7501 (c-forward-syntactic-ws)
7502 (c-with-syntax-table
7503 c-identifier-syntax-table
7504 (looking-at c-known-type-key)))
7505 (save-excursion
7506 (goto-char (1+ identifier-start))
7507 ;; We have already parsed the type earlier,
7508 ;; so it'd be possible to cache the end
7509 ;; position instead of redoing it here, but
7510 ;; then we'd need to keep track of another
7511 ;; position everywhere.
7512 (c-check-type (point)
7513 (progn (c-forward-type)
7514 (point))))))))
7515 (throw 'at-decl-or-cast t))
7516
7517 (if got-identifier
7518 (progn
7519 ;; CASE 16
7520 (when (and got-prefix-before-parens
7521 at-type
7522 (or at-decl-end (looking-at "=[^=]"))
7523 (not context)
7524 (not got-suffix))
7525 ;; Got something like "foo * bar;". Since we're not inside an
7526 ;; arglist it would be a meaningless expression because the
7527 ;; result isn't used. We therefore choose to recognize it as
7528 ;; a declaration. Do not allow a suffix since it could then
7529 ;; be a function call.
7530 (throw 'at-decl-or-cast t))
7531
7532 ;; CASE 17
7533 (when (and (or got-suffix-after-parens
7534 (looking-at "=[^=]"))
7535 (eq at-type 'found)
7536 (not (eq context 'arglist)))
7537 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
7538 ;; be an odd expression or it could be a declaration. Treat
7539 ;; it as a declaration if "a" has been used as a type
7540 ;; somewhere else (if it's a known type we won't get here).
7541 (throw 'at-decl-or-cast t)))
7542
7543 ;; CASE 18
7544 (when (and context
7545 (or got-prefix
7546 (and (eq context 'decl)
7547 (not c-recognize-paren-inits)
7548 (or got-parens got-suffix))))
7549 ;; Got a type followed by an abstract declarator. If `got-prefix'
7550 ;; is set it's something like "a *" without anything after it. If
7551 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
7552 ;; or similar, which we accept only if the context rules out
7553 ;; expressions.
7554 (throw 'at-decl-or-cast t)))
7555
7556 ;; If we had a complete symbol table here (which rules out
7557 ;; `c-found-types') we should return t due to the disambiguation rule
7558 ;; (in at least C++) that anything that can be parsed as a declaration
7559 ;; is a declaration. Now we're being more defensive and prefer to
7560 ;; highlight things like "foo (bar);" as a declaration only if we're
7561 ;; inside an arglist that contains declarations.
7562 ;; CASE 19
7563 (eq context 'decl))))
7564
7565 ;; The point is now after the type decl expression.
7566
7567 (cond
7568 ;; Check for a cast.
7569 ((save-excursion
7570 (and
7571 c-cast-parens
7572
7573 ;; Should be the first type/identifier in a cast paren.
7574 (> preceding-token-end (point-min))
7575 (memq (char-before preceding-token-end) c-cast-parens)
7576
7577 ;; The closing paren should follow.
7578 (progn
7579 (c-forward-syntactic-ws)
7580 (looking-at "\\s)"))
7581
7582 ;; There should be a primary expression after it.
7583 (let (pos)
7584 (forward-char)
7585 (c-forward-syntactic-ws)
7586 (setq cast-end (point))
7587 (and (looking-at c-primary-expr-regexp)
7588 (progn
7589 (setq pos (match-end 0))
7590 (or
7591 ;; Check if the expression begins with a prefix keyword.
7592 (match-beginning 2)
7593 (if (match-beginning 1)
7594 ;; Expression begins with an ambiguous operator. Treat
7595 ;; it as a cast if it's a type decl or if we've
7596 ;; recognized the type somewhere else.
7597 (or at-decl-or-cast
7598 (memq at-type '(t known found)))
7599 ;; Unless it's a keyword, it's the beginning of a primary
7600 ;; expression.
7601 (not (looking-at c-keywords-regexp)))))
7602 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7603 ;; that it matched a whole one so that we don't e.g. confuse
7604 ;; the operator '-' with '->'. It's ok if it matches further,
7605 ;; though, since it e.g. can match the float '.5' while the
7606 ;; operator regexp only matches '.'.
7607 (or (not (looking-at c-nonsymbol-token-regexp))
7608 (<= (match-end 0) pos))))
7609
7610 ;; There should either be a cast before it or something that isn't an
7611 ;; identifier or close paren.
7612 (> preceding-token-end (point-min))
7613 (progn
7614 (goto-char (1- preceding-token-end))
7615 (or (eq (point) last-cast-end)
7616 (progn
7617 (c-backward-syntactic-ws)
7618 (if (< (skip-syntax-backward "w_") 0)
7619 ;; It's a symbol. Accept it only if it's one of the
7620 ;; keywords that can precede an expression (without
7621 ;; surrounding parens).
7622 (looking-at c-simple-stmt-key)
7623 (and
7624 ;; Check that it isn't a close paren (block close is ok,
7625 ;; though).
7626 (not (memq (char-before) '(?\) ?\])))
7627 ;; Check that it isn't a nonsymbol identifier.
7628 (not (c-on-identifier)))))))))
7629
7630 ;; Handle the cast.
7631 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7632 (let ((c-promote-possible-types t))
7633 (goto-char type-start)
7634 (c-forward-type)))
7635
7636 (goto-char cast-end)
7637 'cast)
7638
7639 (at-decl-or-cast
7640 ;; We're at a declaration. Highlight the type and the following
7641 ;; declarators.
7642
7643 (when backup-if-not-cast
7644 (c-fdoc-shift-type-backward t))
7645
7646 (when (and (eq context 'decl) (looking-at ","))
7647 ;; Make sure to propagate the `c-decl-arg-start' property to
7648 ;; the next argument if it's set in this one, to cope with
7649 ;; interactive refontification.
7650 (c-put-c-type-property (point) 'c-decl-arg-start))
7651
7652 ;; Record the type's coordinates in `c-record-type-identifiers' for
7653 ;; later fontification.
7654 (when (and c-record-type-identifiers at-type ;; (not (eq at-type t))
7655 ;; There seems no reason to exclude a token from
7656 ;; fontification just because it's "a known type that can't
7657 ;; be a name or other expression". 2013-09-18.
7658 )
7659 (let ((c-promote-possible-types t))
7660 (save-excursion
7661 (goto-char type-start)
7662 (c-forward-type))))
7663
7664 (cons id-start
7665 (and (or at-type-decl at-typedef)
7666 (cons at-type-decl at-typedef))))
7667
7668 (t
7669 ;; False alarm. Restore the recorded ranges.
7670 (setq c-record-type-identifiers save-rec-type-ids
7671 c-record-ref-identifiers save-rec-ref-ids)
7672 nil))))
7673
7674 (defun c-forward-label (&optional assume-markup preceding-token-end limit)
7675 ;; Assuming that point is at the beginning of a token, check if it starts a
7676 ;; label and if so move over it and return non-nil (t in default situations,
7677 ;; specific symbols (see below) for interesting situations), otherwise don't
7678 ;; move and return nil. "Label" here means "most things with a colon".
7679 ;;
7680 ;; More precisely, a "label" is regarded as one of:
7681 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7682 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7683 ;; bare "case", should the colon be missing. We return t;
7684 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7685 ;; return t;
7686 ;; (iv) One of QT's "extended" C++ variants of
7687 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7688 ;; Returns the symbol `qt-2kwds-colon'.
7689 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
7690 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
7691 ;; colon). Currently (2006-03), this applies only to Objective C's
7692 ;; keywords "@private", "@protected", and "@public". Returns t.
7693 ;;
7694 ;; One of the things which will NOT be recognized as a label is a bit-field
7695 ;; element of a struct, something like "int foo:5".
7696 ;;
7697 ;; The end of the label is taken to be just after the colon, or the end of
7698 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7699 ;; after the end on return. The terminating char gets marked with
7700 ;; `c-decl-end' to improve recognition of the following declaration or
7701 ;; statement.
7702 ;;
7703 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
7704 ;; label, if any, has already been marked up like that.
7705 ;;
7706 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7707 ;; after the preceding token, i.e. on the other side of the
7708 ;; syntactic ws from the point. Use a value less than or equal to
7709 ;; (point-min) if the point is at the first token in (the visible
7710 ;; part of) the buffer.
7711 ;;
7712 ;; The optional LIMIT limits the forward scan for the colon.
7713 ;;
7714 ;; This function records the ranges of the label symbols on
7715 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7716 ;; non-nil.
7717 ;;
7718 ;; This function might do hidden buffer changes.
7719
7720 (let ((start (point))
7721 label-end
7722 qt-symbol-idx
7723 macro-start ; if we're in one.
7724 label-type
7725 kwd)
7726 (cond
7727 ;; "case" or "default" (Doesn't apply to AWK).
7728 ((looking-at c-label-kwds-regexp)
7729 (let ((kwd-end (match-end 1)))
7730 ;; Record only the keyword itself for fontification, since in
7731 ;; case labels the following is a constant expression and not
7732 ;; a label.
7733 (when c-record-type-identifiers
7734 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7735
7736 ;; Find the label end.
7737 (goto-char kwd-end)
7738 (setq label-type
7739 (if (and (c-syntactic-re-search-forward
7740 ;; Stop on chars that aren't allowed in expressions,
7741 ;; and on operator chars that would be meaningless
7742 ;; there. FIXME: This doesn't cope with ?: operators.
7743 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7744 limit t t nil 1)
7745 (match-beginning 2))
7746
7747 (progn ; there's a proper :
7748 (goto-char (match-beginning 2)) ; just after the :
7749 (c-put-c-type-property (1- (point)) 'c-decl-end)
7750 t)
7751
7752 ;; It's an unfinished label. We consider the keyword enough
7753 ;; to recognize it as a label, so that it gets fontified.
7754 ;; Leave the point at the end of it, but don't put any
7755 ;; `c-decl-end' marker.
7756 (goto-char kwd-end)
7757 t))))
7758
7759 ;; @private, @protected, @public, in Objective C, or similar.
7760 ((and c-opt-extra-label-key
7761 (looking-at c-opt-extra-label-key))
7762 ;; For a `c-opt-extra-label-key' match, we record the whole
7763 ;; thing for fontification. That's to get the leading '@' in
7764 ;; Objective-C protection labels fontified.
7765 (goto-char (match-end 1))
7766 (when c-record-type-identifiers
7767 (c-record-ref-id (cons (match-beginning 1) (point))))
7768 (c-put-c-type-property (1- (point)) 'c-decl-end)
7769 (setq label-type t))
7770
7771 ;; All other cases of labels.
7772 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
7773
7774 ;; A colon label must have something before the colon.
7775 (not (eq (char-after) ?:))
7776
7777 ;; Check that we're not after a token that can't precede a label.
7778 (or
7779 ;; Trivially succeeds when there's no preceding token.
7780 ;; Succeeds when we're at a virtual semicolon.
7781 (if preceding-token-end
7782 (<= preceding-token-end (point-min))
7783 (save-excursion
7784 (c-backward-syntactic-ws)
7785 (setq preceding-token-end (point))
7786 (or (bobp)
7787 (c-at-vsemi-p))))
7788
7789 ;; Check if we're after a label, if we're after a closing
7790 ;; paren that belong to statement, and with
7791 ;; `c-label-prefix-re'. It's done in different order
7792 ;; depending on `assume-markup' since the checks have
7793 ;; different expensiveness.
7794 (if assume-markup
7795 (or
7796 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7797 'c-decl-end)
7798
7799 (save-excursion
7800 (goto-char (1- preceding-token-end))
7801 (c-beginning-of-current-token)
7802 (or (looking-at c-label-prefix-re)
7803 (looking-at c-block-stmt-1-key)))
7804
7805 (and (eq (char-before preceding-token-end) ?\))
7806 (c-after-conditional)))
7807
7808 (or
7809 (save-excursion
7810 (goto-char (1- preceding-token-end))
7811 (c-beginning-of-current-token)
7812 (or (looking-at c-label-prefix-re)
7813 (looking-at c-block-stmt-1-key)))
7814
7815 (cond
7816 ((eq (char-before preceding-token-end) ?\))
7817 (c-after-conditional))
7818
7819 ((eq (char-before preceding-token-end) ?:)
7820 ;; Might be after another label, so check it recursively.
7821 (save-restriction
7822 (save-excursion
7823 (goto-char (1- preceding-token-end))
7824 ;; Essentially the same as the
7825 ;; `c-syntactic-re-search-forward' regexp below.
7826 (setq macro-start
7827 (save-excursion (and (c-beginning-of-macro)
7828 (point))))
7829 (if macro-start (narrow-to-region macro-start (point-max)))
7830 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7831 ;; Note: the following should work instead of the
7832 ;; narrow-to-region above. Investigate why not,
7833 ;; sometime. ACM, 2006-03-31.
7834 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7835 ;; macro-start t)
7836 (let ((pte (point))
7837 ;; If the caller turned on recording for us,
7838 ;; it shouldn't apply when we check the
7839 ;; preceding label.
7840 c-record-type-identifiers)
7841 ;; A label can't start at a cpp directive. Check for
7842 ;; this, since c-forward-syntactic-ws would foul up on it.
7843 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7844 (c-forward-syntactic-ws)
7845 (c-forward-label nil pte start))))))))))
7846
7847 ;; Point is still at the beginning of the possible label construct.
7848 ;;
7849 ;; Check that the next nonsymbol token is ":", or that we're in one
7850 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7851 ;; arguments. FIXME: Should build this regexp from the language
7852 ;; constants.
7853 (cond
7854 ;; public: protected: private:
7855 ((and
7856 (c-major-mode-is 'c++-mode)
7857 (search-forward-regexp
7858 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7859 (progn (backward-char)
7860 (c-forward-syntactic-ws limit)
7861 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7862 (forward-char)
7863 (setq label-type t))
7864 ;; QT double keyword like "protected slots:" or goto target.
7865 ((progn (goto-char start) nil))
7866 ((when (c-syntactic-re-search-forward
7867 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7868 (backward-char)
7869 (setq label-end (point))
7870 (setq qt-symbol-idx
7871 (and (c-major-mode-is 'c++-mode)
7872 (string-match
7873 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
7874 (buffer-substring start (point)))))
7875 (c-forward-syntactic-ws limit)
7876 (cond
7877 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7878 (forward-char)
7879 (setq label-type
7880 (if (or (string= "signals" ; Special QT macro
7881 (setq kwd (buffer-substring-no-properties start label-end)))
7882 (string= "Q_SIGNALS" kwd))
7883 'qt-1kwd-colon
7884 'goto-target)))
7885 ((and qt-symbol-idx
7886 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
7887 (progn (c-forward-syntactic-ws limit)
7888 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7889 (forward-char)
7890 (setq label-type 'qt-2kwds-colon)))))))
7891
7892 (save-restriction
7893 (narrow-to-region start (point))
7894
7895 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7896 (catch 'check-label
7897 (goto-char start)
7898 (while (progn
7899 (when (looking-at c-nonlabel-token-key)
7900 (goto-char start)
7901 (setq label-type nil)
7902 (throw 'check-label nil))
7903 (and (c-safe (c-forward-sexp)
7904 (c-forward-syntactic-ws)
7905 t)
7906 (not (eobp)))))
7907
7908 ;; Record the identifiers in the label for fontification, unless
7909 ;; it begins with `c-label-kwds' in which case the following
7910 ;; identifiers are part of a (constant) expression that
7911 ;; shouldn't be fontified.
7912 (when (and c-record-type-identifiers
7913 (progn (goto-char start)
7914 (not (looking-at c-label-kwds-regexp))))
7915 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7916 (c-record-ref-id (cons (match-beginning 0)
7917 (match-end 0)))))
7918
7919 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
7920 (goto-char (point-max)))))
7921
7922 (t
7923 ;; Not a label.
7924 (goto-char start)))
7925 label-type))
7926
7927 (defun c-forward-objc-directive ()
7928 ;; Assuming the point is at the beginning of a token, try to move
7929 ;; forward to the end of the Objective-C directive that starts
7930 ;; there. Return t if a directive was fully recognized, otherwise
7931 ;; the point is moved as far as one could be successfully parsed and
7932 ;; nil is returned.
7933 ;;
7934 ;; This function records identifier ranges on
7935 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7936 ;; `c-record-type-identifiers' is non-nil.
7937 ;;
7938 ;; This function might do hidden buffer changes.
7939
7940 (let ((start (point))
7941 start-char
7942 (c-promote-possible-types t)
7943 lim
7944 ;; Turn off recognition of angle bracket arglists while parsing
7945 ;; types here since the protocol reference list might then be
7946 ;; considered part of the preceding name or superclass-name.
7947 c-recognize-<>-arglists)
7948
7949 (if (or
7950 (when (looking-at
7951 (eval-when-compile
7952 (c-make-keywords-re t
7953 (append (c-lang-const c-protection-kwds objc)
7954 '("@end"))
7955 'objc-mode)))
7956 (goto-char (match-end 1))
7957 t)
7958
7959 (and
7960 (looking-at
7961 (eval-when-compile
7962 (c-make-keywords-re t
7963 '("@interface" "@implementation" "@protocol")
7964 'objc-mode)))
7965
7966 ;; Handle the name of the class itself.
7967 (progn
7968 ;; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7969 ;; at EOB.
7970 (goto-char (match-end 0))
7971 (setq lim (point))
7972 (c-skip-ws-forward)
7973 (c-forward-type))
7974
7975 (catch 'break
7976 ;; Look for ": superclass-name" or "( category-name )".
7977 (when (looking-at "[:(]")
7978 (setq start-char (char-after))
7979 (forward-char)
7980 (c-forward-syntactic-ws)
7981 (unless (c-forward-type) (throw 'break nil))
7982 (when (eq start-char ?\()
7983 (unless (eq (char-after) ?\)) (throw 'break nil))
7984 (forward-char)
7985 (c-forward-syntactic-ws)))
7986
7987 ;; Look for a protocol reference list.
7988 (if (eq (char-after) ?<)
7989 (let ((c-recognize-<>-arglists t)
7990 (c-parse-and-markup-<>-arglists t)
7991 c-restricted-<>-arglists)
7992 (c-forward-<>-arglist t))
7993 t))))
7994
7995 (progn
7996 (c-backward-syntactic-ws lim)
7997 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7998 (c-put-c-type-property (1- (point)) 'c-decl-end)
7999 t)
8000
8001 (c-clear-c-type-property start (point) 'c-decl-end)
8002 nil)))
8003
8004 (defun c-beginning-of-inheritance-list (&optional lim)
8005 ;; Go to the first non-whitespace after the colon that starts a
8006 ;; multiple inheritance introduction. Optional LIM is the farthest
8007 ;; back we should search.
8008 ;;
8009 ;; This function might do hidden buffer changes.
8010 (c-with-syntax-table c++-template-syntax-table
8011 (c-backward-token-2 0 t lim)
8012 (while (and (or (looking-at c-symbol-start)
8013 (looking-at "[<,]\\|::"))
8014 (zerop (c-backward-token-2 1 t lim))))))
8015
8016 (defun c-in-method-def-p ()
8017 ;; Return nil if we aren't in a method definition, otherwise the
8018 ;; position of the initial [+-].
8019 ;;
8020 ;; This function might do hidden buffer changes.
8021 (save-excursion
8022 (beginning-of-line)
8023 (and c-opt-method-key
8024 (looking-at c-opt-method-key)
8025 (point))
8026 ))
8027
8028 ;; Contributed by Kevin Ryde <user42@zip.com.au>.
8029 (defun c-in-gcc-asm-p ()
8030 ;; Return non-nil if point is within a gcc \"asm\" block.
8031 ;;
8032 ;; This should be called with point inside an argument list.
8033 ;;
8034 ;; Only one level of enclosing parentheses is considered, so for
8035 ;; instance nil is returned when in a function call within an asm
8036 ;; operand.
8037 ;;
8038 ;; This function might do hidden buffer changes.
8039
8040 (and c-opt-asm-stmt-key
8041 (save-excursion
8042 (beginning-of-line)
8043 (backward-up-list 1)
8044 (c-beginning-of-statement-1 (point-min) nil t)
8045 (looking-at c-opt-asm-stmt-key))))
8046
8047 (defun c-at-toplevel-p ()
8048 "Return a determination as to whether point is \"at the top level\".
8049 Informally, \"at the top level\" is anywhere where you can write
8050 a function.
8051
8052 More precisely, being at the top-level means that point is either
8053 outside any enclosing block (such as a function definition), or
8054 directly inside a class, namespace or other block that contains
8055 another declaration level.
8056
8057 If point is not at the top-level (e.g. it is inside a method
8058 definition), then nil is returned. Otherwise, if point is at a
8059 top-level not enclosed within a class definition, t is returned.
8060 Otherwise, a 2-vector is returned where the zeroth element is the
8061 buffer position of the start of the class declaration, and the first
8062 element is the buffer position of the enclosing class's opening
8063 brace.
8064
8065 Note that this function might do hidden buffer changes. See the
8066 comment at the start of cc-engine.el for more info."
8067 ;; Note to maintainers: this function consumes a great mass of CPU cycles.
8068 ;; Its use should thus be minimized as far as possible.
8069 (let ((paren-state (c-parse-state)))
8070 (or (not (c-most-enclosing-brace paren-state))
8071 (c-search-uplist-for-classkey paren-state))))
8072
8073 (defun c-just-after-func-arglist-p (&optional lim)
8074 ;; Return non-nil if the point is in the region after the argument
8075 ;; list of a function and its opening brace (or semicolon in case it
8076 ;; got no body). If there are K&R style argument declarations in
8077 ;; that region, the point has to be inside the first one for this
8078 ;; function to recognize it.
8079 ;;
8080 ;; If successful, the point is moved to the first token after the
8081 ;; function header (see `c-forward-decl-or-cast-1' for details) and
8082 ;; the position of the opening paren of the function arglist is
8083 ;; returned.
8084 ;;
8085 ;; The point is clobbered if not successful.
8086 ;;
8087 ;; LIM is used as bound for backward buffer searches.
8088 ;;
8089 ;; This function might do hidden buffer changes.
8090
8091 (let ((beg (point)) id-start)
8092 (and
8093 (eq (c-beginning-of-statement-1 lim) 'same)
8094
8095 (not (and (c-major-mode-is 'objc-mode)
8096 (c-forward-objc-directive)))
8097
8098 (setq id-start
8099 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
8100 (< id-start beg)
8101
8102 ;; There should not be a '=' or ',' between beg and the
8103 ;; start of the declaration since that means we were in the
8104 ;; "expression part" of the declaration.
8105 (or (> (point) beg)
8106 (not (looking-at "[=,]")))
8107
8108 (save-excursion
8109 ;; Check that there's an arglist paren in the
8110 ;; declaration.
8111 (goto-char id-start)
8112 (cond ((eq (char-after) ?\()
8113 ;; The declarator is a paren expression, so skip past it
8114 ;; so that we don't get stuck on that instead of the
8115 ;; function arglist.
8116 (c-forward-sexp))
8117 ((and c-opt-op-identifier-prefix
8118 (looking-at c-opt-op-identifier-prefix))
8119 ;; Don't trip up on "operator ()".
8120 (c-forward-token-2 2 t)))
8121 (and (< (point) beg)
8122 (c-syntactic-re-search-forward "(" beg t t)
8123 (1- (point)))))))
8124
8125 (defun c-in-knr-argdecl (&optional lim)
8126 ;; Return the position of the first argument declaration if point is
8127 ;; inside a K&R style argument declaration list, nil otherwise.
8128 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
8129 ;; position that bounds the backward search for the argument list.
8130 ;;
8131 ;; Point must be within a possible K&R region, e.g. just before a top-level
8132 ;; "{". It must be outside of parens and brackets. The test can return
8133 ;; false positives otherwise.
8134 ;;
8135 ;; This function might do hidden buffer changes.
8136
8137 (save-excursion
8138 (save-restriction
8139 ;; If we're in a macro, our search range is restricted to it. Narrow to
8140 ;; the searchable range.
8141 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
8142 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
8143 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
8144 before-lparen after-rparen
8145 (pp-count-out 20)) ; Max number of paren/brace constructs before
8146 ; we give up
8147 (narrow-to-region low-lim (or macro-end (point-max)))
8148
8149 ;; Search backwards for the defun's argument list. We give up if we
8150 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
8151 ;; a knr region) or BOB.
8152 ;;
8153 ;; The criterion for a paren structure being the arg list is:
8154 ;; o - there is non-WS stuff after it but before any "{"; AND
8155 ;; o - the token after it isn't a ";" AND
8156 ;; o - it is preceded by either an identifier (the function name) or
8157 ;; a macro expansion like "DEFUN (...)"; AND
8158 ;; o - its content is a non-empty comma-separated list of identifiers
8159 ;; (an empty arg list won't have a knr region).
8160 ;;
8161 ;; The following snippet illustrates these rules:
8162 ;; int foo (bar, baz, yuk)
8163 ;; int bar [] ;
8164 ;; int (*baz) (my_type) ;
8165 ;; int (*) (void) (*yuk) (void) ;
8166 ;; {
8167
8168 (catch 'knr
8169 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
8170 (setq pp-count-out (1- pp-count-out))
8171 (c-syntactic-skip-backward "^)]}=")
8172 (cond ((eq (char-before) ?\))
8173 (setq after-rparen (point)))
8174 ((eq (char-before) ?\])
8175 (setq after-rparen nil))
8176 (t ; either } (hit previous defun) or = or no more
8177 ; parens/brackets.
8178 (throw 'knr nil)))
8179
8180 (if after-rparen
8181 ;; We're inside a paren. Could it be our argument list....?
8182 (if
8183 (and
8184 (progn
8185 (goto-char after-rparen)
8186 (unless (c-go-list-backward) (throw 'knr nil)) ;
8187 ;; FIXME!!! What about macros between the parens? 2007/01/20
8188 (setq before-lparen (point)))
8189
8190 ;; It can't be the arg list if next token is ; or {
8191 (progn (goto-char after-rparen)
8192 (c-forward-syntactic-ws)
8193 (not (memq (char-after) '(?\; ?\{ ?\=))))
8194
8195 ;; Is the thing preceding the list an identifier (the
8196 ;; function name), or a macro expansion?
8197 (progn
8198 (goto-char before-lparen)
8199 (eq (c-backward-token-2) 0)
8200 (or (eq (c-on-identifier) (point))
8201 (and (eq (char-after) ?\))
8202 (c-go-up-list-backward)
8203 (eq (c-backward-token-2) 0)
8204 (eq (c-on-identifier) (point)))))
8205
8206 ;; Have we got a non-empty list of comma-separated
8207 ;; identifiers?
8208 (progn
8209 (goto-char before-lparen)
8210 (c-forward-token-2) ; to first token inside parens
8211 (and
8212 (c-on-identifier)
8213 (c-forward-token-2)
8214 (catch 'id-list
8215 (while (eq (char-after) ?\,)
8216 (c-forward-token-2)
8217 (unless (c-on-identifier) (throw 'id-list nil))
8218 (c-forward-token-2))
8219 (eq (char-after) ?\))))))
8220
8221 ;; ...Yes. We've identified the function's argument list.
8222 (throw 'knr
8223 (progn (goto-char after-rparen)
8224 (c-forward-syntactic-ws)
8225 (point)))
8226
8227 ;; ...No. The current parens aren't the function's arg list.
8228 (goto-char before-lparen))
8229
8230 (or (c-go-list-backward) ; backwards over [ .... ]
8231 (throw 'knr nil)))))))))
8232
8233 (defun c-skip-conditional ()
8234 ;; skip forward over conditional at point, including any predicate
8235 ;; statements in parentheses. No error checking is performed.
8236 ;;
8237 ;; This function might do hidden buffer changes.
8238 (c-forward-sexp (cond
8239 ;; else if()
8240 ((looking-at (concat "\\<else"
8241 "\\([ \t\n]\\|\\\\\n\\)+"
8242 "if\\>\\([^_]\\|$\\)"))
8243 3)
8244 ;; do, else, try, finally
8245 ((looking-at (concat "\\<\\("
8246 "do\\|else\\|try\\|finally"
8247 "\\)\\>\\([^_]\\|$\\)"))
8248 1)
8249 ;; for, if, while, switch, catch, synchronized, foreach
8250 (t 2))))
8251
8252 (defun c-after-conditional (&optional lim)
8253 ;; If looking at the token after a conditional then return the
8254 ;; position of its start, otherwise return nil.
8255 ;;
8256 ;; This function might do hidden buffer changes.
8257 (save-excursion
8258 (and (zerop (c-backward-token-2 1 t lim))
8259 (or (looking-at c-block-stmt-1-key)
8260 (and (eq (char-after) ?\()
8261 (zerop (c-backward-token-2 1 t lim))
8262 (or (looking-at c-block-stmt-2-key)
8263 (looking-at c-block-stmt-1-2-key))))
8264 (point))))
8265
8266 (defun c-after-special-operator-id (&optional lim)
8267 ;; If the point is after an operator identifier that isn't handled
8268 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
8269 ;; position of the start of that identifier is returned. nil is
8270 ;; returned otherwise. The point may be anywhere in the syntactic
8271 ;; whitespace after the last token of the operator identifier.
8272 ;;
8273 ;; This function might do hidden buffer changes.
8274 (save-excursion
8275 (and c-overloadable-operators-regexp
8276 (zerop (c-backward-token-2 1 nil lim))
8277 (looking-at c-overloadable-operators-regexp)
8278 (or (not c-opt-op-identifier-prefix)
8279 (and
8280 (zerop (c-backward-token-2 1 nil lim))
8281 (looking-at c-opt-op-identifier-prefix)))
8282 (point))))
8283
8284 (defsubst c-backward-to-block-anchor (&optional lim)
8285 ;; Assuming point is at a brace that opens a statement block of some
8286 ;; kind, move to the proper anchor point for that block. It might
8287 ;; need to be adjusted further by c-add-stmt-syntax, but the
8288 ;; position at return is suitable as start position for that
8289 ;; function.
8290 ;;
8291 ;; This function might do hidden buffer changes.
8292 (unless (= (point) (c-point 'boi))
8293 (let ((start (c-after-conditional lim)))
8294 (if start
8295 (goto-char start)))))
8296
8297 (defsubst c-backward-to-decl-anchor (&optional lim)
8298 ;; Assuming point is at a brace that opens the block of a top level
8299 ;; declaration of some kind, move to the proper anchor point for
8300 ;; that block.
8301 ;;
8302 ;; This function might do hidden buffer changes.
8303 (unless (= (point) (c-point 'boi))
8304 (c-beginning-of-statement-1 lim)))
8305
8306 (defun c-search-decl-header-end ()
8307 ;; Search forward for the end of the "header" of the current
8308 ;; declaration. That's the position where the definition body
8309 ;; starts, or the first variable initializer, or the ending
8310 ;; semicolon. I.e. search forward for the closest following
8311 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
8312 ;; _after_ the first found token, or at point-max if none is found.
8313 ;;
8314 ;; This function might do hidden buffer changes.
8315
8316 (let ((base (point)))
8317 (if (c-major-mode-is 'c++-mode)
8318
8319 ;; In C++ we need to take special care to handle operator
8320 ;; tokens and those pesky template brackets.
8321 (while (and
8322 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
8323 (or
8324 (c-end-of-current-token base)
8325 ;; Handle operator identifiers, i.e. ignore any
8326 ;; operator token preceded by "operator".
8327 (save-excursion
8328 (and (c-safe (c-backward-sexp) t)
8329 (looking-at c-opt-op-identifier-prefix)))
8330 (and (eq (char-before) ?<)
8331 (c-with-syntax-table c++-template-syntax-table
8332 (if (c-safe (goto-char (c-up-list-forward (point))))
8333 t
8334 (goto-char (point-max))
8335 nil)))))
8336 (setq base (point)))
8337
8338 (while (and
8339 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
8340 (c-end-of-current-token base))
8341 (setq base (point))))))
8342
8343 (defun c-beginning-of-decl-1 (&optional lim)
8344 ;; Go to the beginning of the current declaration, or the beginning
8345 ;; of the previous one if already at the start of it. Point won't
8346 ;; be moved out of any surrounding paren. Return a cons cell of the
8347 ;; form (MOVE . KNR-POS). MOVE is like the return value from
8348 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
8349 ;; style argument declarations (and they are to be recognized) then
8350 ;; KNR-POS is set to the start of the first such argument
8351 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
8352 ;; position that bounds the backward search.
8353 ;;
8354 ;; NB: Cases where the declaration continues after the block, as in
8355 ;; "struct foo { ... } bar;", are currently recognized as two
8356 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
8357 ;;
8358 ;; This function might do hidden buffer changes.
8359 (catch 'return
8360 (let* ((start (point))
8361 (last-stmt-start (point))
8362 (move (c-beginning-of-statement-1 lim nil t)))
8363
8364 ;; `c-beginning-of-statement-1' stops at a block start, but we
8365 ;; want to continue if the block doesn't begin a top level
8366 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
8367 ;; or an open paren.
8368 (let ((beg (point)) tentative-move)
8369 ;; Go back one "statement" each time round the loop until we're just
8370 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
8371 ;; an ObjC method. This will move over a multiple declaration whose
8372 ;; components are comma separated.
8373 (while (and
8374 ;; Must check with c-opt-method-key in ObjC mode.
8375 (not (and c-opt-method-key
8376 (looking-at c-opt-method-key)))
8377 (/= last-stmt-start (point))
8378 (progn
8379 (c-backward-syntactic-ws lim)
8380 (not (memq (char-before) '(?\; ?} ?: nil))))
8381 (save-excursion
8382 (backward-char)
8383 (not (looking-at "\\s(")))
8384 ;; Check that we don't move from the first thing in a
8385 ;; macro to its header.
8386 (not (eq (setq tentative-move
8387 (c-beginning-of-statement-1 lim nil t))
8388 'macro)))
8389 (setq last-stmt-start beg
8390 beg (point)
8391 move tentative-move))
8392 (goto-char beg))
8393
8394 (when c-recognize-knr-p
8395 (let ((fallback-pos (point)) knr-argdecl-start)
8396 ;; Handle K&R argdecls. Back up after the "statement" jumped
8397 ;; over by `c-beginning-of-statement-1', unless it was the
8398 ;; function body, in which case we're sitting on the opening
8399 ;; brace now. Then test if we're in a K&R argdecl region and
8400 ;; that we started at the other side of the first argdecl in
8401 ;; it.
8402 (unless (eq (char-after) ?{)
8403 (goto-char last-stmt-start))
8404 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
8405 (< knr-argdecl-start start)
8406 (progn
8407 (goto-char knr-argdecl-start)
8408 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
8409 (throw 'return
8410 (cons (if (eq (char-after fallback-pos) ?{)
8411 'previous
8412 'same)
8413 knr-argdecl-start))
8414 (goto-char fallback-pos))))
8415
8416 ;; `c-beginning-of-statement-1' counts each brace block as a separate
8417 ;; statement, so the result will be 'previous if we've moved over any.
8418 ;; So change our result back to 'same if necessary.
8419 ;;
8420 ;; If they were brace list initializers we might not have moved over a
8421 ;; declaration boundary though, so change it to 'same if we've moved
8422 ;; past a '=' before '{', but not ';'. (This ought to be integrated
8423 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
8424 ;; potentially can search over a large amount of text.). Take special
8425 ;; pains not to get mislead by C++'s "operator=", and the like.
8426 (if (and (eq move 'previous)
8427 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
8428 c++-template-syntax-table
8429 (syntax-table))
8430 (save-excursion
8431 (and
8432 (progn
8433 (while ; keep going back to "[;={"s until we either find
8434 ; no more, or get to one which isn't an "operator ="
8435 (and (c-syntactic-re-search-forward "[;={]" start t t t)
8436 (eq (char-before) ?=)
8437 c-overloadable-operators-regexp
8438 c-opt-op-identifier-prefix
8439 (save-excursion
8440 (eq (c-backward-token-2) 0)
8441 (looking-at c-overloadable-operators-regexp)
8442 (eq (c-backward-token-2) 0)
8443 (looking-at c-opt-op-identifier-prefix))))
8444 (eq (char-before) ?=))
8445 (c-syntactic-re-search-forward "[;{]" start t t)
8446 (eq (char-before) ?{)
8447 (c-safe (goto-char (c-up-list-forward (point))) t)
8448 (not (c-syntactic-re-search-forward ";" start t t))))))
8449 (cons 'same nil)
8450 (cons move nil)))))
8451
8452 (defun c-end-of-decl-1 ()
8453 ;; Assuming point is at the start of a declaration (as detected by
8454 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
8455 ;; `c-beginning-of-decl-1', this function handles the case when a
8456 ;; block is followed by identifiers in e.g. struct declarations in C
8457 ;; or C++. If a proper end was found then t is returned, otherwise
8458 ;; point is moved as far as possible within the current sexp and nil
8459 ;; is returned. This function doesn't handle macros; use
8460 ;; `c-end-of-macro' instead in those cases.
8461 ;;
8462 ;; This function might do hidden buffer changes.
8463 (let ((start (point))
8464 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
8465 c++-template-syntax-table
8466 (syntax-table))))
8467 (catch 'return
8468 (c-search-decl-header-end)
8469
8470 (when (and c-recognize-knr-p
8471 (eq (char-before) ?\;)
8472 (c-in-knr-argdecl start))
8473 ;; Stopped at the ';' in a K&R argdecl section which is
8474 ;; detected using the same criteria as in
8475 ;; `c-beginning-of-decl-1'. Move to the following block
8476 ;; start.
8477 (c-syntactic-re-search-forward "{" nil 'move t))
8478
8479 (when (eq (char-before) ?{)
8480 ;; Encountered a block in the declaration. Jump over it.
8481 (condition-case nil
8482 (goto-char (c-up-list-forward (point)))
8483 (error (goto-char (point-max))
8484 (throw 'return nil)))
8485 (if (or (not c-opt-block-decls-with-vars-key)
8486 (save-excursion
8487 (c-with-syntax-table decl-syntax-table
8488 (let ((lim (point)))
8489 (goto-char start)
8490 (not (and
8491 ;; Check for `c-opt-block-decls-with-vars-key'
8492 ;; before the first paren.
8493 (c-syntactic-re-search-forward
8494 (concat "[;=([{]\\|\\("
8495 c-opt-block-decls-with-vars-key
8496 "\\)")
8497 lim t t t)
8498 (match-beginning 1)
8499 (not (eq (char-before) ?_))
8500 ;; Check that the first following paren is
8501 ;; the block.
8502 (c-syntactic-re-search-forward "[;=([{]"
8503 lim t t t)
8504 (eq (char-before) ?{)))))))
8505 ;; The declaration doesn't have any of the
8506 ;; `c-opt-block-decls-with-vars' keywords in the
8507 ;; beginning, so it ends here at the end of the block.
8508 (throw 'return t)))
8509
8510 (c-with-syntax-table decl-syntax-table
8511 (while (progn
8512 (if (eq (char-before) ?\;)
8513 (throw 'return t))
8514 (c-syntactic-re-search-forward ";" nil 'move t))))
8515 nil)))
8516
8517 (defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
8518 ;; Assuming the point is at an open brace, check if it starts a
8519 ;; block that contains another declaration level, i.e. that isn't a
8520 ;; statement block or a brace list, and if so return non-nil.
8521 ;;
8522 ;; If the check is successful, the return value is the start of the
8523 ;; keyword that tells what kind of construct it is, i.e. typically
8524 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
8525 ;; the point will be at the start of the construct, before any
8526 ;; leading specifiers, otherwise it's at the returned position.
8527 ;;
8528 ;; The point is clobbered if the check is unsuccessful.
8529 ;;
8530 ;; CONTAINING-SEXP is the position of the open of the surrounding
8531 ;; paren, or nil if none.
8532 ;;
8533 ;; The optional LIMIT limits the backward search for the start of
8534 ;; the construct. It's assumed to be at a syntactically relevant
8535 ;; position.
8536 ;;
8537 ;; If any template arglists are found in the searched region before
8538 ;; the open brace, they get marked with paren syntax.
8539 ;;
8540 ;; This function might do hidden buffer changes.
8541
8542 (let ((open-brace (point)) kwd-start first-specifier-pos)
8543 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8544
8545 (when (and c-recognize-<>-arglists
8546 (eq (char-before) ?>))
8547 ;; Could be at the end of a template arglist.
8548 (let ((c-parse-and-markup-<>-arglists t))
8549 (while (and
8550 (c-backward-<>-arglist nil limit)
8551 (progn
8552 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8553 (eq (char-before) ?>))))))
8554
8555 ;; Note: Can't get bogus hits inside template arglists below since they
8556 ;; have gotten paren syntax above.
8557 (when (and
8558 ;; If `goto-start' is set we begin by searching for the
8559 ;; first possible position of a leading specifier list.
8560 ;; The `c-decl-block-key' search continues from there since
8561 ;; we know it can't match earlier.
8562 (if goto-start
8563 (when (c-syntactic-re-search-forward c-symbol-start
8564 open-brace t t)
8565 (goto-char (setq first-specifier-pos (match-beginning 0)))
8566 t)
8567 t)
8568
8569 (cond
8570 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
8571 (goto-char (setq kwd-start (match-beginning 0)))
8572 (and
8573 ;; Exclude cases where we matched what would ordinarily
8574 ;; be a block declaration keyword, except where it's not
8575 ;; legal because it's part of a "compound keyword" like
8576 ;; "enum class". Of course, if c-after-brace-list-key
8577 ;; is nil, we can skip the test.
8578 (or (equal c-after-brace-list-key "\\<\\>")
8579 (save-match-data
8580 (save-excursion
8581 (not
8582 (and
8583 (looking-at c-after-brace-list-key)
8584 (= (c-backward-token-2 1 t) 0)
8585 (looking-at c-brace-list-key))))))
8586 (or
8587 ;; Found a keyword that can't be a type?
8588 (match-beginning 1)
8589
8590 ;; Can be a type too, in which case it's the return type of a
8591 ;; function (under the assumption that no declaration level
8592 ;; block construct starts with a type).
8593 (not (c-forward-type))
8594
8595 ;; Jumped over a type, but it could be a declaration keyword
8596 ;; followed by the declared identifier that we've jumped over
8597 ;; instead (e.g. in "class Foo {"). If it indeed is a type
8598 ;; then we should be at the declarator now, so check for a
8599 ;; valid declarator start.
8600 ;;
8601 ;; Note: This doesn't cope with the case when a declared
8602 ;; identifier is followed by e.g. '(' in a language where '('
8603 ;; also might be part of a declarator expression. Currently
8604 ;; there's no such language.
8605 (not (or (looking-at c-symbol-start)
8606 (looking-at c-type-decl-prefix-key))))))
8607
8608 ;; In Pike a list of modifiers may be followed by a brace
8609 ;; to make them apply to many identifiers. Note that the
8610 ;; match data will be empty on return in this case.
8611 ((and (c-major-mode-is 'pike-mode)
8612 (progn
8613 (goto-char open-brace)
8614 (= (c-backward-token-2) 0))
8615 (looking-at c-specifier-key)
8616 ;; Use this variant to avoid yet another special regexp.
8617 (c-keyword-member (c-keyword-sym (match-string 1))
8618 'c-modifier-kwds))
8619 (setq kwd-start (point))
8620 t)))
8621
8622 ;; Got a match.
8623
8624 (if goto-start
8625 ;; Back up over any preceding specifiers and their clauses
8626 ;; by going forward from `first-specifier-pos', which is the
8627 ;; earliest possible position where the specifier list can
8628 ;; start.
8629 (progn
8630 (goto-char first-specifier-pos)
8631
8632 (while (< (point) kwd-start)
8633 (if (looking-at c-symbol-key)
8634 ;; Accept any plain symbol token on the ground that
8635 ;; it's a specifier masked through a macro (just
8636 ;; like `c-forward-decl-or-cast-1' skip forward over
8637 ;; such tokens).
8638 ;;
8639 ;; Could be more restrictive wrt invalid keywords,
8640 ;; but that'd only occur in invalid code so there's
8641 ;; no use spending effort on it.
8642 (let ((end (match-end 0)))
8643 (unless (c-forward-keyword-clause 0)
8644 (goto-char end)
8645 (c-forward-syntactic-ws)))
8646
8647 ;; Can't parse a declaration preamble and is still
8648 ;; before `kwd-start'. That means `first-specifier-pos'
8649 ;; was in some earlier construct. Search again.
8650 (if (c-syntactic-re-search-forward c-symbol-start
8651 kwd-start 'move t)
8652 (goto-char (setq first-specifier-pos (match-beginning 0)))
8653 ;; Got no preamble before the block declaration keyword.
8654 (setq first-specifier-pos kwd-start))))
8655
8656 (goto-char first-specifier-pos))
8657 (goto-char kwd-start))
8658
8659 kwd-start)))
8660
8661 (defun c-search-uplist-for-classkey (paren-state)
8662 ;; Check if the closest containing paren sexp is a declaration
8663 ;; block, returning a 2 element vector in that case. Aref 0
8664 ;; contains the bufpos at boi of the class key line, and aref 1
8665 ;; contains the bufpos of the open brace. This function is an
8666 ;; obsolete wrapper for `c-looking-at-decl-block'.
8667 ;;
8668 ;; This function might do hidden buffer changes.
8669 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8670 (when open-paren-pos
8671 (save-excursion
8672 (goto-char open-paren-pos)
8673 (when (and (eq (char-after) ?{)
8674 (c-looking-at-decl-block
8675 (c-safe-position open-paren-pos paren-state)
8676 nil))
8677 (back-to-indentation)
8678 (vector (point) open-paren-pos))))))
8679
8680 (defun c-most-enclosing-decl-block (paren-state)
8681 ;; Return the buffer position of the most enclosing decl-block brace (in the
8682 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8683 ;; none was found.
8684 (let* ((open-brace (c-pull-open-brace paren-state))
8685 (next-open-brace (c-pull-open-brace paren-state)))
8686 (while (and open-brace
8687 (save-excursion
8688 (goto-char open-brace)
8689 (not (c-looking-at-decl-block next-open-brace nil))))
8690 (setq open-brace next-open-brace
8691 next-open-brace (c-pull-open-brace paren-state)))
8692 open-brace))
8693
8694 (defun c-cheap-inside-bracelist-p (paren-state)
8695 ;; Return the position of the L-brace if point is inside a brace list
8696 ;; initialization of an array, etc. This is an approximate function,
8697 ;; designed for speed over accuracy. It will not find every bracelist, but
8698 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
8699 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
8700 ;; is everywhere else.
8701 (let (b-pos)
8702 (save-excursion
8703 (while
8704 (and (setq b-pos (c-pull-open-brace paren-state))
8705 (progn (goto-char b-pos)
8706 (c-backward-sws)
8707 (c-backward-token-2)
8708 (not (looking-at "=")))))
8709 b-pos)))
8710
8711 (defun c-backward-colon-prefixed-type ()
8712 ;; We're at the token after what might be a type prefixed with a colon. Try
8713 ;; moving backward over this type and the colon. On success, return t and
8714 ;; leave point before colon; on failure, leave point unchanged. Will clobber
8715 ;; match data.
8716 (let ((here (point))
8717 (colon-pos nil))
8718 (save-excursion
8719 (while
8720 (and (eql (c-backward-token-2) 0)
8721 (or (not (looking-at "\\s)"))
8722 (c-go-up-list-backward))
8723 (cond
8724 ((eql (char-after) ?:)
8725 (setq colon-pos (point))
8726 (forward-char)
8727 (c-forward-syntactic-ws)
8728 (or (and (c-forward-type)
8729 (progn (c-forward-syntactic-ws)
8730 (eq (point) here)))
8731 (setq colon-pos nil))
8732 nil)
8733 ((eql (char-after) ?\()
8734 t)
8735 ((looking-at c-symbol-key)
8736 t)
8737 (t nil)))))
8738 (when colon-pos
8739 (goto-char colon-pos)
8740 t)))
8741
8742 (defun c-backward-over-enum-header ()
8743 ;; We're at a "{". Move back to the enum-like keyword that starts this
8744 ;; declaration and return t, otherwise don't move and return nil.
8745 (let ((here (point))
8746 up-sexp-pos before-identifier)
8747 (when c-recognize-post-brace-list-type-p
8748 (c-backward-colon-prefixed-type))
8749 (while
8750 (and
8751 (eq (c-backward-token-2) 0)
8752 (or (not (looking-at "\\s)"))
8753 (c-go-up-list-backward))
8754 (cond
8755 ((and (looking-at c-symbol-key) (c-on-identifier)
8756 (not before-identifier))
8757 (setq before-identifier t))
8758 ((and before-identifier
8759 (or (eql (char-after) ?,)
8760 (looking-at c-postfix-decl-spec-key)))
8761 (setq before-identifier nil)
8762 t)
8763 ((looking-at c-after-brace-list-key) t)
8764 ((looking-at c-brace-list-key) nil)
8765 ((and c-recognize-<>-arglists
8766 (eq (char-after) ?<)
8767 (looking-at "\\s("))
8768 t)
8769 (t nil))))
8770 (or (looking-at c-brace-list-key)
8771 (progn (goto-char here) nil))))
8772
8773 (defun c-inside-bracelist-p (containing-sexp paren-state)
8774 ;; return the buffer position of the beginning of the brace list
8775 ;; statement if we're inside a brace list, otherwise return nil.
8776 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
8777 ;; paren. PAREN-STATE is the remainder of the state of enclosing
8778 ;; braces
8779 ;;
8780 ;; N.B.: This algorithm can potentially get confused by cpp macros
8781 ;; placed in inconvenient locations. It's a trade-off we make for
8782 ;; speed.
8783 ;;
8784 ;; This function might do hidden buffer changes.
8785 (or
8786 ;; This will pick up brace list declarations.
8787 (save-excursion
8788 (goto-char containing-sexp)
8789 (c-backward-over-enum-header))
8790 ;; this will pick up array/aggregate init lists, even if they are nested.
8791 (save-excursion
8792 (let ((class-key
8793 ;; Pike can have class definitions anywhere, so we must
8794 ;; check for the class key here.
8795 (and (c-major-mode-is 'pike-mode)
8796 c-decl-block-key))
8797 bufpos braceassignp lim next-containing macro-start)
8798 (while (and (not bufpos)
8799 containing-sexp)
8800 (when paren-state
8801 (if (consp (car paren-state))
8802 (setq lim (cdr (car paren-state))
8803 paren-state (cdr paren-state))
8804 (setq lim (car paren-state)))
8805 (when paren-state
8806 (setq next-containing (car paren-state)
8807 paren-state (cdr paren-state))))
8808 (goto-char containing-sexp)
8809 (if (c-looking-at-inexpr-block next-containing next-containing)
8810 ;; We're in an in-expression block of some kind. Do not
8811 ;; check nesting. We deliberately set the limit to the
8812 ;; containing sexp, so that c-looking-at-inexpr-block
8813 ;; doesn't check for an identifier before it.
8814 (setq containing-sexp nil)
8815 ;; see if the open brace is preceded by = or [...] in
8816 ;; this statement, but watch out for operator=
8817 (setq braceassignp 'dontknow)
8818 (c-backward-token-2 1 t lim)
8819 ;; Checks to do only on the first sexp before the brace.
8820 (when (and c-opt-inexpr-brace-list-key
8821 (eq (char-after) ?\[))
8822 ;; In Java, an initialization brace list may follow
8823 ;; directly after "new Foo[]", so check for a "new"
8824 ;; earlier.
8825 (while (eq braceassignp 'dontknow)
8826 (setq braceassignp
8827 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8828 ((looking-at c-opt-inexpr-brace-list-key) t)
8829 ((looking-at "\\sw\\|\\s_\\|[.[]")
8830 ;; Carry on looking if this is an
8831 ;; identifier (may contain "." in Java)
8832 ;; or another "[]" sexp.
8833 'dontknow)
8834 (t nil)))))
8835 ;; Checks to do on all sexps before the brace, up to the
8836 ;; beginning of the statement.
8837 (while (eq braceassignp 'dontknow)
8838 (cond ((eq (char-after) ?\;)
8839 (setq braceassignp nil))
8840 ((and class-key
8841 (looking-at class-key))
8842 (setq braceassignp nil))
8843 ((eq (char-after) ?=)
8844 ;; We've seen a =, but must check earlier tokens so
8845 ;; that it isn't something that should be ignored.
8846 (setq braceassignp 'maybe)
8847 (while (and (eq braceassignp 'maybe)
8848 (zerop (c-backward-token-2 1 t lim)))
8849 (setq braceassignp
8850 (cond
8851 ;; Check for operator =
8852 ((and c-opt-op-identifier-prefix
8853 (looking-at c-opt-op-identifier-prefix))
8854 nil)
8855 ;; Check for `<opchar>= in Pike.
8856 ((and (c-major-mode-is 'pike-mode)
8857 (or (eq (char-after) ?`)
8858 ;; Special case for Pikes
8859 ;; `[]=, since '[' is not in
8860 ;; the punctuation class.
8861 (and (eq (char-after) ?\[)
8862 (eq (char-before) ?`))))
8863 nil)
8864 ((looking-at "\\s.") 'maybe)
8865 ;; make sure we're not in a C++ template
8866 ;; argument assignment
8867 ((and
8868 (c-major-mode-is 'c++-mode)
8869 (save-excursion
8870 (let ((here (point))
8871 (pos< (progn
8872 (skip-chars-backward "^<>")
8873 (point))))
8874 (and (eq (char-before) ?<)
8875 (not (c-crosses-statement-barrier-p
8876 pos< here))
8877 (not (c-in-literal))
8878 ))))
8879 nil)
8880 (t t))))))
8881 (if (and (eq braceassignp 'dontknow)
8882 (/= (c-backward-token-2 1 t lim) 0))
8883 (setq braceassignp nil)))
8884 (cond
8885 (braceassignp
8886 ;; We've hit the beginning of the aggregate list.
8887 (c-beginning-of-statement-1
8888 (c-most-enclosing-brace paren-state))
8889 (setq bufpos (point)))
8890 ((eq (char-after) ?\;)
8891 ;; Brace lists can't contain a semicolon, so we're done.
8892 (setq containing-sexp nil))
8893 ((and (setq macro-start (point))
8894 (c-forward-to-cpp-define-body)
8895 (eq (point) containing-sexp))
8896 ;; We've a macro whose expansion starts with the '{'.
8897 ;; Heuristically, if we have a ';' in it we've not got a
8898 ;; brace list, otherwise we have.
8899 (let ((macro-end (progn (c-end-of-macro) (point))))
8900 (goto-char containing-sexp)
8901 (forward-char)
8902 (if (and (c-syntactic-re-search-forward "[;,]" macro-end t t)
8903 (eq (char-before) ?\;))
8904 (setq bufpos nil
8905 containing-sexp nil)
8906 (setq bufpos macro-start))))
8907 (t
8908 ;; Go up one level
8909 (setq containing-sexp next-containing
8910 lim nil
8911 next-containing nil)))))
8912
8913 bufpos))
8914 ))
8915
8916 (defun c-looking-at-special-brace-list (&optional lim)
8917 ;; If we're looking at the start of a pike-style list, i.e., `({ })',
8918 ;; `([ ])', `(< >)', etc., a cons of a cons of its starting and ending
8919 ;; positions and its entry in c-special-brace-lists is returned, nil
8920 ;; otherwise. The ending position is nil if the list is still open.
8921 ;; LIM is the limit for forward search. The point may either be at
8922 ;; the `(' or at the following paren character. Tries to check the
8923 ;; matching closer, but assumes it's correct if no balanced paren is
8924 ;; found (i.e. the case `({ ... } ... )' is detected as _not_ being
8925 ;; a special brace list).
8926 ;;
8927 ;; This function might do hidden buffer changes.
8928 (if c-special-brace-lists
8929 (condition-case ()
8930 (save-excursion
8931 (let ((beg (point))
8932 inner-beg end type)
8933 (c-forward-syntactic-ws)
8934 (if (eq (char-after) ?\()
8935 (progn
8936 (forward-char 1)
8937 (c-forward-syntactic-ws)
8938 (setq inner-beg (point))
8939 (setq type (assq (char-after) c-special-brace-lists)))
8940 (if (setq type (assq (char-after) c-special-brace-lists))
8941 (progn
8942 (setq inner-beg (point))
8943 (c-backward-syntactic-ws)
8944 (forward-char -1)
8945 (setq beg (if (eq (char-after) ?\()
8946 (point)
8947 nil)))))
8948 (if (and beg type)
8949 (if (and (c-safe
8950 (goto-char beg)
8951 (c-forward-sexp 1)
8952 (setq end (point))
8953 (= (char-before) ?\)))
8954 (c-safe
8955 (goto-char inner-beg)
8956 (if (looking-at "\\s(")
8957 ;; Check balancing of the inner paren
8958 ;; below.
8959 (progn
8960 (c-forward-sexp 1)
8961 t)
8962 ;; If the inner char isn't a paren then
8963 ;; we can't check balancing, so just
8964 ;; check the char before the outer
8965 ;; closing paren.
8966 (goto-char end)
8967 (backward-char)
8968 (c-backward-syntactic-ws)
8969 (= (char-before) (cdr type)))))
8970 (if (or (/= (char-syntax (char-before)) ?\))
8971 (= (progn
8972 (c-forward-syntactic-ws)
8973 (point))
8974 (1- end)))
8975 (cons (cons beg end) type))
8976 (cons (list beg) type)))))
8977 (error nil))))
8978
8979 (defun c-looking-at-bos (&optional lim)
8980 ;; Return non-nil if between two statements or declarations, assuming
8981 ;; point is not inside a literal or comment.
8982 ;;
8983 ;; Obsolete - `c-at-statement-start-p' or `c-at-expression-start-p'
8984 ;; are recommended instead.
8985 ;;
8986 ;; This function might do hidden buffer changes.
8987 (c-at-statement-start-p))
8988 (make-obsolete 'c-looking-at-bos 'c-at-statement-start-p "22.1")
8989
8990 (defun c-looking-at-inexpr-block (lim containing-sexp &optional check-at-end)
8991 ;; Return non-nil if we're looking at the beginning of a block
8992 ;; inside an expression. The value returned is actually a cons of
8993 ;; either 'inlambda, 'inexpr-statement or 'inexpr-class and the
8994 ;; position of the beginning of the construct.
8995 ;;
8996 ;; LIM limits the backward search. CONTAINING-SEXP is the start
8997 ;; position of the closest containing list. If it's nil, the
8998 ;; containing paren isn't used to decide whether we're inside an
8999 ;; expression or not. If both LIM and CONTAINING-SEXP are used, LIM
9000 ;; needs to be farther back.
9001 ;;
9002 ;; If CHECK-AT-END is non-nil then extra checks at the end of the
9003 ;; brace block might be done. It should only be used when the
9004 ;; construct can be assumed to be complete, i.e. when the original
9005 ;; starting position was further down than that.
9006 ;;
9007 ;; This function might do hidden buffer changes.
9008
9009 (save-excursion
9010 (let ((res 'maybe) passed-paren
9011 (closest-lim (or containing-sexp lim (point-min)))
9012 ;; Look at the character after point only as a last resort
9013 ;; when we can't disambiguate.
9014 (block-follows (and (eq (char-after) ?{) (point))))
9015
9016 (while (and (eq res 'maybe)
9017 (progn (c-backward-syntactic-ws)
9018 (> (point) closest-lim))
9019 (not (bobp))
9020 (progn (backward-char)
9021 (looking-at "[]).]\\|\\w\\|\\s_"))
9022 (c-safe (forward-char)
9023 (goto-char (scan-sexps (point) -1))))
9024
9025 (setq res
9026 (if (looking-at c-keywords-regexp)
9027 (let ((kw-sym (c-keyword-sym (match-string 1))))
9028 (cond
9029 ((and block-follows
9030 (c-keyword-member kw-sym 'c-inexpr-class-kwds))
9031 (and (not (eq passed-paren ?\[))
9032 (or (not (looking-at c-class-key))
9033 ;; If the class definition is at the start of
9034 ;; a statement, we don't consider it an
9035 ;; in-expression class.
9036 (let ((prev (point)))
9037 (while (and
9038 (= (c-backward-token-2 1 nil closest-lim) 0)
9039 (eq (char-syntax (char-after)) ?w))
9040 (setq prev (point)))
9041 (goto-char prev)
9042 (not (c-at-statement-start-p)))
9043 ;; Also, in Pike we treat it as an
9044 ;; in-expression class if it's used in an
9045 ;; object clone expression.
9046 (save-excursion
9047 (and check-at-end
9048 (c-major-mode-is 'pike-mode)
9049 (progn (goto-char block-follows)
9050 (zerop (c-forward-token-2 1 t)))
9051 (eq (char-after) ?\())))
9052 (cons 'inexpr-class (point))))
9053 ((c-keyword-member kw-sym 'c-inexpr-block-kwds)
9054 (when (not passed-paren)
9055 (cons 'inexpr-statement (point))))
9056 ((c-keyword-member kw-sym 'c-lambda-kwds)
9057 (when (or (not passed-paren)
9058 (eq passed-paren ?\())
9059 (cons 'inlambda (point))))
9060 ((c-keyword-member kw-sym 'c-block-stmt-kwds)
9061 nil)
9062 (t
9063 'maybe)))
9064
9065 (if (looking-at "\\s(")
9066 (if passed-paren
9067 (if (and (eq passed-paren ?\[)
9068 (eq (char-after) ?\[))
9069 ;; Accept several square bracket sexps for
9070 ;; Java array initializations.
9071 'maybe)
9072 (setq passed-paren (char-after))
9073 'maybe)
9074 'maybe))))
9075
9076 (if (eq res 'maybe)
9077 (when (and c-recognize-paren-inexpr-blocks
9078 block-follows
9079 containing-sexp
9080 (eq (char-after containing-sexp) ?\())
9081 (goto-char containing-sexp)
9082 (if (or (save-excursion
9083 (c-backward-syntactic-ws lim)
9084 (and (> (point) (or lim (point-min)))
9085 (c-on-identifier)))
9086 (and c-special-brace-lists
9087 (c-looking-at-special-brace-list)))
9088 nil
9089 (cons 'inexpr-statement (point))))
9090
9091 res))))
9092
9093 (defun c-looking-at-inexpr-block-backward (paren-state)
9094 ;; Returns non-nil if we're looking at the end of an in-expression
9095 ;; block, otherwise the same as `c-looking-at-inexpr-block'.
9096 ;; PAREN-STATE is the paren state relevant at the current position.
9097 ;;
9098 ;; This function might do hidden buffer changes.
9099 (save-excursion
9100 ;; We currently only recognize a block.
9101 (let ((here (point))
9102 (elem (car-safe paren-state))
9103 containing-sexp)
9104 (when (and (consp elem)
9105 (progn (goto-char (cdr elem))
9106 (c-forward-syntactic-ws here)
9107 (= (point) here)))
9108 (goto-char (car elem))
9109 (if (setq paren-state (cdr paren-state))
9110 (setq containing-sexp (car-safe paren-state)))
9111 (c-looking-at-inexpr-block (c-safe-position containing-sexp
9112 paren-state)
9113 containing-sexp)))))
9114
9115 (defun c-at-macro-vsemi-p (&optional pos)
9116 ;; Is there a "virtual semicolon" at POS or point?
9117 ;; (See cc-defs.el for full details of "virtual semicolons".)
9118 ;;
9119 ;; This is true when point is at the last non syntactic WS position on the
9120 ;; line, there is a macro call last on the line, and this particular macro's
9121 ;; name is defined by the regexp `c-vs-macro-regexp' as not needing a
9122 ;; semicolon.
9123 (save-excursion
9124 (save-restriction
9125 (widen)
9126 (if pos
9127 (goto-char pos)
9128 (setq pos (point)))
9129 (and
9130 c-macro-with-semi-re
9131 (eq (skip-chars-backward " \t") 0)
9132
9133 ;; Check we've got nothing after this except comments and empty lines
9134 ;; joined by escaped EOLs.
9135 (skip-chars-forward " \t") ; always returns non-nil.
9136 (progn
9137 (while ; go over 1 block comment per iteration.
9138 (and
9139 (looking-at "\\(\\\\[\n\r][ \t]*\\)*")
9140 (goto-char (match-end 0))
9141 (cond
9142 ((looking-at c-block-comment-start-regexp)
9143 (and (forward-comment 1)
9144 (skip-chars-forward " \t"))) ; always returns non-nil
9145 ((looking-at c-line-comment-start-regexp)
9146 (end-of-line)
9147 nil)
9148 (t nil))))
9149 (eolp))
9150
9151 (goto-char pos)
9152 (progn (c-backward-syntactic-ws)
9153 (eq (point) pos))
9154
9155 ;; Check for one of the listed macros being before point.
9156 (or (not (eq (char-before) ?\)))
9157 (when (c-go-list-backward)
9158 (c-backward-syntactic-ws)
9159 t))
9160 (c-simple-skip-symbol-backward)
9161 (looking-at c-macro-with-semi-re)
9162 (goto-char pos)
9163 (not (c-in-literal)))))) ; The most expensive check last.
9164
9165 (defun c-macro-vsemi-status-unknown-p () t) ; See cc-defs.el.
9166
9167 \f
9168 ;; `c-guess-basic-syntax' and the functions that precedes it below
9169 ;; implements the main decision tree for determining the syntactic
9170 ;; analysis of the current line of code.
9171
9172 ;; Dynamically bound to t when `c-guess-basic-syntax' is called during
9173 ;; auto newline analysis.
9174 (defvar c-auto-newline-analysis nil)
9175
9176 (defun c-brace-anchor-point (bracepos)
9177 ;; BRACEPOS is the position of a brace in a construct like "namespace
9178 ;; Bar {". Return the anchor point in this construct; this is the
9179 ;; earliest symbol on the brace's line which isn't earlier than
9180 ;; "namespace".
9181 ;;
9182 ;; Currently (2007-08-17), "like namespace" means "matches
9183 ;; c-other-block-decl-kwds". It doesn't work with "class" or "struct"
9184 ;; or anything like that.
9185 (save-excursion
9186 (let ((boi (c-point 'boi bracepos)))
9187 (goto-char bracepos)
9188 (while (and (> (point) boi)
9189 (not (looking-at c-other-decl-block-key)))
9190 (c-backward-token-2))
9191 (if (> (point) boi) (point) boi))))
9192
9193 (defsubst c-add-syntax (symbol &rest args)
9194 ;; A simple function to prepend a new syntax element to
9195 ;; `c-syntactic-context'. Using `setq' on it is unsafe since it
9196 ;; should always be dynamically bound but since we read it first
9197 ;; we'll fail properly anyway if this function is misused.
9198 (setq c-syntactic-context (cons (cons symbol args)
9199 c-syntactic-context)))
9200
9201 (defsubst c-append-syntax (symbol &rest args)
9202 ;; Like `c-add-syntax' but appends to the end of the syntax list.
9203 ;; (Normally not necessary.)
9204 (setq c-syntactic-context (nconc c-syntactic-context
9205 (list (cons symbol args)))))
9206
9207 (defun c-add-stmt-syntax (syntax-symbol
9208 syntax-extra-args
9209 stop-at-boi-only
9210 containing-sexp
9211 paren-state)
9212 ;; Add the indicated SYNTAX-SYMBOL to `c-syntactic-context', extending it as
9213 ;; needed with further syntax elements of the types `substatement',
9214 ;; `inexpr-statement', `arglist-cont-nonempty', `statement-block-intro', and
9215 ;; `defun-block-intro'.
9216 ;;
9217 ;; Do the generic processing to anchor the given syntax symbol on
9218 ;; the preceding statement: Skip over any labels and containing
9219 ;; statements on the same line, and then search backward until we
9220 ;; find a statement or block start that begins at boi without a
9221 ;; label or comment.
9222 ;;
9223 ;; Point is assumed to be at the prospective anchor point for the
9224 ;; given SYNTAX-SYMBOL. More syntax entries are added if we need to
9225 ;; skip past open parens and containing statements. Most of the added
9226 ;; syntax elements will get the same anchor point - the exception is
9227 ;; for an anchor in a construct like "namespace"[*] - this is as early
9228 ;; as possible in the construct but on the same line as the {.
9229 ;;
9230 ;; [*] i.e. with a keyword matching c-other-block-decl-kwds.
9231 ;;
9232 ;; SYNTAX-EXTRA-ARGS are a list of the extra arguments for the
9233 ;; syntax symbol. They are appended after the anchor point.
9234 ;;
9235 ;; If STOP-AT-BOI-ONLY is nil, we can stop in the middle of the line
9236 ;; if the current statement starts there.
9237 ;;
9238 ;; Note: It's not a problem if PAREN-STATE "overshoots"
9239 ;; CONTAINING-SEXP, i.e. contains info about parens further down.
9240 ;;
9241 ;; This function might do hidden buffer changes.
9242
9243 (if (= (point) (c-point 'boi))
9244 ;; This is by far the most common case, so let's give it special
9245 ;; treatment.
9246 (apply 'c-add-syntax syntax-symbol (point) syntax-extra-args)
9247
9248 (let ((syntax-last c-syntactic-context)
9249 (boi (c-point 'boi))
9250 ;; Set when we're on a label, so that we don't stop there.
9251 ;; FIXME: To be complete we should check if we're on a label
9252 ;; now at the start.
9253 on-label)
9254
9255 ;; Use point as the anchor point for "namespace", "extern", etc.
9256 (apply 'c-add-syntax syntax-symbol
9257 (if (rassq syntax-symbol c-other-decl-block-key-in-symbols-alist)
9258 (point) nil)
9259 syntax-extra-args)
9260
9261 ;; Loop while we have to back out of containing blocks.
9262 (while
9263 (and
9264 (catch 'back-up-block
9265
9266 ;; Loop while we have to back up statements.
9267 (while (or (/= (point) boi)
9268 on-label
9269 (looking-at c-comment-start-regexp))
9270
9271 ;; Skip past any comments that stands between the
9272 ;; statement start and boi.
9273 (let ((savepos (point)))
9274 (while (and (/= savepos boi)
9275 (c-backward-single-comment))
9276 (setq savepos (point)
9277 boi (c-point 'boi)))
9278 (goto-char savepos))
9279
9280 ;; Skip to the beginning of this statement or backward
9281 ;; another one.
9282 (let ((old-pos (point))
9283 (old-boi boi)
9284 (step-type (c-beginning-of-statement-1 containing-sexp)))
9285 (setq boi (c-point 'boi)
9286 on-label (eq step-type 'label))
9287
9288 (cond ((= (point) old-pos)
9289 ;; If we didn't move we're at the start of a block and
9290 ;; have to continue outside it.
9291 (throw 'back-up-block t))
9292
9293 ((and (eq step-type 'up)
9294 (>= (point) old-boi)
9295 (looking-at "else\\>[^_]")
9296 (save-excursion
9297 (goto-char old-pos)
9298 (looking-at "if\\>[^_]")))
9299 ;; Special case to avoid deeper and deeper indentation
9300 ;; of "else if" clauses.
9301 )
9302
9303 ((and (not stop-at-boi-only)
9304 (/= old-pos old-boi)
9305 (memq step-type '(up previous)))
9306 ;; If stop-at-boi-only is nil, we shouldn't back up
9307 ;; over previous or containing statements to try to
9308 ;; reach boi, so go back to the last position and
9309 ;; exit.
9310 (goto-char old-pos)
9311 (throw 'back-up-block nil))
9312
9313 (t
9314 (if (and (not stop-at-boi-only)
9315 (memq step-type '(up previous beginning)))
9316 ;; If we've moved into another statement then we
9317 ;; should no longer try to stop in the middle of a
9318 ;; line.
9319 (setq stop-at-boi-only t))
9320
9321 ;; Record this as a substatement if we skipped up one
9322 ;; level.
9323 (when (eq step-type 'up)
9324 (c-add-syntax 'substatement nil))))
9325 )))
9326
9327 containing-sexp)
9328
9329 ;; Now we have to go out of this block.
9330 (goto-char containing-sexp)
9331
9332 ;; Don't stop in the middle of a special brace list opener
9333 ;; like "({".
9334 (when c-special-brace-lists
9335 (let ((special-list (c-looking-at-special-brace-list)))
9336 (when (and special-list
9337 (< (car (car special-list)) (point)))
9338 (setq containing-sexp (car (car special-list)))
9339 (goto-char containing-sexp))))
9340
9341 (setq paren-state (c-whack-state-after containing-sexp paren-state)
9342 containing-sexp (c-most-enclosing-brace paren-state)
9343 boi (c-point 'boi))
9344
9345 ;; Analyze the construct in front of the block we've stepped out
9346 ;; from and add the right syntactic element for it.
9347 (let ((paren-pos (point))
9348 (paren-char (char-after))
9349 step-type)
9350
9351 (if (eq paren-char ?\()
9352 ;; Stepped out of a parenthesis block, so we're in an
9353 ;; expression now.
9354 (progn
9355 (when (/= paren-pos boi)
9356 (if (and c-recognize-paren-inexpr-blocks
9357 (progn
9358 (c-backward-syntactic-ws containing-sexp)
9359 (or (not (looking-at "\\>"))
9360 (not (c-on-identifier))))
9361 (save-excursion
9362 (goto-char (1+ paren-pos))
9363 (c-forward-syntactic-ws)
9364 (eq (char-after) ?{)))
9365 ;; Stepped out of an in-expression statement. This
9366 ;; syntactic element won't get an anchor pos.
9367 (c-add-syntax 'inexpr-statement)
9368
9369 ;; A parenthesis normally belongs to an arglist.
9370 (c-add-syntax 'arglist-cont-nonempty nil paren-pos)))
9371
9372 (goto-char (max boi
9373 (if containing-sexp
9374 (1+ containing-sexp)
9375 (point-min))))
9376 (setq step-type 'same
9377 on-label nil))
9378
9379 ;; Stepped out of a brace block.
9380 (setq step-type (c-beginning-of-statement-1 containing-sexp)
9381 on-label (eq step-type 'label))
9382
9383 (if (and (eq step-type 'same)
9384 (/= paren-pos (point)))
9385 (let (inexpr)
9386 (cond
9387 ((save-excursion
9388 (goto-char paren-pos)
9389 (setq inexpr (c-looking-at-inexpr-block
9390 (c-safe-position containing-sexp paren-state)
9391 containing-sexp)))
9392 (c-add-syntax (if (eq (car inexpr) 'inlambda)
9393 'defun-block-intro
9394 'statement-block-intro)
9395 nil))
9396 ((looking-at c-other-decl-block-key)
9397 (c-add-syntax
9398 (cdr (assoc (match-string 1)
9399 c-other-decl-block-key-in-symbols-alist))
9400 (max (c-point 'boi paren-pos) (point))))
9401 (t (c-add-syntax 'defun-block-intro nil))))
9402
9403 (c-add-syntax 'statement-block-intro nil)))
9404
9405 (if (= paren-pos boi)
9406 ;; Always done if the open brace was at boi. The
9407 ;; c-beginning-of-statement-1 call above is necessary
9408 ;; anyway, to decide the type of block-intro to add.
9409 (goto-char paren-pos)
9410 (setq boi (c-point 'boi)))
9411 ))
9412
9413 ;; Fill in the current point as the anchor for all the symbols
9414 ;; added above.
9415 (let ((p c-syntactic-context) q)
9416 (while (not (eq p syntax-last))
9417 (setq q (cdr (car p))) ; e.g. (nil 28) [from (arglist-cont-nonempty nil 28)]
9418 (while q
9419 (unless (car q)
9420 (setcar q (point)))
9421 (setq q (cdr q)))
9422 (setq p (cdr p))))
9423 )))
9424
9425 (defun c-add-class-syntax (symbol
9426 containing-decl-open
9427 containing-decl-start
9428 containing-decl-kwd
9429 paren-state)
9430 ;; The inclass and class-close syntactic symbols are added in
9431 ;; several places and some work is needed to fix everything.
9432 ;; Therefore it's collected here.
9433 ;;
9434 ;; This function might do hidden buffer changes.
9435 (goto-char containing-decl-open)
9436 (if (and (eq symbol 'inclass) (= (point) (c-point 'boi)))
9437 (progn
9438 (c-add-syntax symbol containing-decl-open)
9439 containing-decl-open)
9440 (goto-char containing-decl-start)
9441 ;; Ought to use `c-add-stmt-syntax' instead of backing up to boi
9442 ;; here, but we have to do like this for compatibility.
9443 (back-to-indentation)
9444 (c-add-syntax symbol (point))
9445 (if (and (c-keyword-member containing-decl-kwd
9446 'c-inexpr-class-kwds)
9447 (/= containing-decl-start (c-point 'boi containing-decl-start)))
9448 (c-add-syntax 'inexpr-class))
9449 (point)))
9450
9451 (defun c-guess-continued-construct (indent-point
9452 char-after-ip
9453 beg-of-same-or-containing-stmt
9454 containing-sexp
9455 paren-state)
9456 ;; This function contains the decision tree reached through both
9457 ;; cases 18 and 10. It's a continued statement or top level
9458 ;; construct of some kind.
9459 ;;
9460 ;; This function might do hidden buffer changes.
9461
9462 (let (special-brace-list placeholder)
9463 (goto-char indent-point)
9464 (skip-chars-forward " \t")
9465
9466 (cond
9467 ;; (CASE A removed.)
9468 ;; CASE B: open braces for class or brace-lists
9469 ((setq special-brace-list
9470 (or (and c-special-brace-lists
9471 (c-looking-at-special-brace-list))
9472 (eq char-after-ip ?{)))
9473
9474 (cond
9475 ;; CASE B.1: class-open
9476 ((save-excursion
9477 (and (eq (char-after) ?{)
9478 (c-looking-at-decl-block containing-sexp t)
9479 (setq beg-of-same-or-containing-stmt (point))))
9480 (c-add-syntax 'class-open beg-of-same-or-containing-stmt))
9481
9482 ;; CASE B.2: brace-list-open
9483 ((or (consp special-brace-list)
9484 (save-excursion
9485 (goto-char beg-of-same-or-containing-stmt)
9486 (c-syntactic-re-search-forward "=\\([^=]\\|$\\)"
9487 indent-point t t t)))
9488 ;; The most semantically accurate symbol here is
9489 ;; brace-list-open, but we normally report it simply as a
9490 ;; statement-cont. The reason is that one normally adjusts
9491 ;; brace-list-open for brace lists as top-level constructs,
9492 ;; and brace lists inside statements is a completely different
9493 ;; context. C.f. case 5A.3.
9494 (c-beginning-of-statement-1 containing-sexp)
9495 (c-add-stmt-syntax (if c-auto-newline-analysis
9496 ;; Turn off the dwim above when we're
9497 ;; analyzing the nature of the brace
9498 ;; for the auto newline feature.
9499 'brace-list-open
9500 'statement-cont)
9501 nil nil
9502 containing-sexp paren-state))
9503
9504 ;; CASE B.3: The body of a function declared inside a normal
9505 ;; block. Can occur e.g. in Pike and when using gcc
9506 ;; extensions, but watch out for macros followed by blocks.
9507 ;; C.f. cases E, 16F and 17G.
9508 ((and (not (c-at-statement-start-p))
9509 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
9510 'same)
9511 (save-excursion
9512 (let ((c-recognize-typeless-decls nil))
9513 ;; Turn off recognition of constructs that lacks a
9514 ;; type in this case, since that's more likely to be
9515 ;; a macro followed by a block.
9516 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9517 (c-add-stmt-syntax 'defun-open nil t
9518 containing-sexp paren-state))
9519
9520 ;; CASE B.4: Continued statement with block open. The most
9521 ;; accurate analysis is perhaps `statement-cont' together with
9522 ;; `block-open' but we play DWIM and use `substatement-open'
9523 ;; instead. The rationale is that this typically is a macro
9524 ;; followed by a block which makes it very similar to a
9525 ;; statement with a substatement block.
9526 (t
9527 (c-add-stmt-syntax 'substatement-open nil nil
9528 containing-sexp paren-state))
9529 ))
9530
9531 ;; CASE C: iostream insertion or extraction operator
9532 ((and (looking-at "\\(<<\\|>>\\)\\([^=]\\|$\\)")
9533 (save-excursion
9534 (goto-char beg-of-same-or-containing-stmt)
9535 ;; If there is no preceding streamop in the statement
9536 ;; then indent this line as a normal statement-cont.
9537 (when (c-syntactic-re-search-forward
9538 "\\(<<\\|>>\\)\\([^=]\\|$\\)" indent-point 'move t t)
9539 (c-add-syntax 'stream-op (c-point 'boi))
9540 t))))
9541
9542 ;; CASE E: In the "K&R region" of a function declared inside a
9543 ;; normal block. C.f. case B.3.
9544 ((and (save-excursion
9545 ;; Check that the next token is a '{'. This works as
9546 ;; long as no language that allows nested function
9547 ;; definitions allows stuff like member init lists, K&R
9548 ;; declarations or throws clauses there.
9549 ;;
9550 ;; Note that we do a forward search for something ahead
9551 ;; of the indentation line here. That's not good since
9552 ;; the user might not have typed it yet. Unfortunately
9553 ;; it's exceedingly tricky to recognize a function
9554 ;; prototype in a code block without resorting to this.
9555 (c-forward-syntactic-ws)
9556 (eq (char-after) ?{))
9557 (not (c-at-statement-start-p))
9558 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
9559 'same)
9560 (save-excursion
9561 (let ((c-recognize-typeless-decls nil))
9562 ;; Turn off recognition of constructs that lacks a
9563 ;; type in this case, since that's more likely to be
9564 ;; a macro followed by a block.
9565 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9566 (c-add-stmt-syntax 'func-decl-cont nil t
9567 containing-sexp paren-state))
9568
9569 ;;CASE F: continued statement and the only preceding items are
9570 ;;annotations.
9571 ((and (c-major-mode-is 'java-mode)
9572 (setq placeholder (point))
9573 (c-beginning-of-statement-1)
9574 (progn
9575 (while (and (c-forward-annotation)
9576 (< (point) placeholder))
9577 (c-forward-syntactic-ws))
9578 t)
9579 (prog1
9580 (>= (point) placeholder)
9581 (goto-char placeholder)))
9582 (c-beginning-of-statement-1 containing-sexp)
9583 (c-add-syntax 'annotation-var-cont (point)))
9584
9585 ;; CASE G: a template list continuation?
9586 ;; Mostly a duplication of case 5D.3 to fix templates-19:
9587 ((and (c-major-mode-is 'c++-mode)
9588 (save-excursion
9589 (goto-char indent-point)
9590 (c-with-syntax-table c++-template-syntax-table
9591 (setq placeholder (c-up-list-backward)))
9592 (and placeholder
9593 (eq (char-after placeholder) ?<)
9594 (/= (char-before placeholder) ?<)
9595 (progn
9596 (goto-char (1+ placeholder))
9597 (not (looking-at c-<-op-cont-regexp))))))
9598 (c-with-syntax-table c++-template-syntax-table
9599 (goto-char placeholder)
9600 (c-beginning-of-statement-1 containing-sexp t))
9601 (if (save-excursion
9602 (c-backward-syntactic-ws containing-sexp)
9603 (eq (char-before) ?<))
9604 ;; In a nested template arglist.
9605 (progn
9606 (goto-char placeholder)
9607 (c-syntactic-skip-backward "^,;" containing-sexp t)
9608 (c-forward-syntactic-ws))
9609 (back-to-indentation))
9610 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
9611 ;; template aware.
9612 (c-add-syntax 'template-args-cont (point) placeholder))
9613
9614 ;; CASE D: continued statement.
9615 (t
9616 (c-beginning-of-statement-1 containing-sexp)
9617 (c-add-stmt-syntax 'statement-cont nil nil
9618 containing-sexp paren-state))
9619 )))
9620
9621 ;; The next autoload was added by RMS on 2005/8/9 - don't know why (ACM,
9622 ;; 2005/11/29).
9623 ;;;###autoload
9624 (defun c-guess-basic-syntax ()
9625 "Return the syntactic context of the current line."
9626 (save-excursion
9627 (beginning-of-line)
9628 (c-save-buffer-state
9629 ((indent-point (point))
9630 (case-fold-search nil)
9631 ;; A whole ugly bunch of various temporary variables. Have
9632 ;; to declare them here since it's not possible to declare
9633 ;; a variable with only the scope of a cond test and the
9634 ;; following result clauses, and most of this function is a
9635 ;; single gigantic cond. :P
9636 literal char-before-ip before-ws-ip char-after-ip macro-start
9637 in-macro-expr c-syntactic-context placeholder c-in-literal-cache
9638 step-type tmpsymbol keyword injava-inher special-brace-list tmp-pos
9639 containing-<
9640 ;; The following record some positions for the containing
9641 ;; declaration block if we're directly within one:
9642 ;; `containing-decl-open' is the position of the open
9643 ;; brace. `containing-decl-start' is the start of the
9644 ;; declaration. `containing-decl-kwd' is the keyword
9645 ;; symbol of the keyword that tells what kind of block it
9646 ;; is.
9647 containing-decl-open
9648 containing-decl-start
9649 containing-decl-kwd
9650 ;; The open paren of the closest surrounding sexp or nil if
9651 ;; there is none.
9652 containing-sexp
9653 ;; The position after the closest preceding brace sexp
9654 ;; (nested sexps are ignored), or the position after
9655 ;; `containing-sexp' if there is none, or (point-min) if
9656 ;; `containing-sexp' is nil.
9657 lim
9658 ;; The paren state outside `containing-sexp', or at
9659 ;; `indent-point' if `containing-sexp' is nil.
9660 (paren-state (c-parse-state))
9661 ;; There's always at most one syntactic element which got
9662 ;; an anchor pos. It's stored in syntactic-relpos.
9663 syntactic-relpos
9664 (c-stmt-delim-chars c-stmt-delim-chars))
9665
9666 ;; Check if we're directly inside an enclosing declaration
9667 ;; level block.
9668 (when (and (setq containing-sexp
9669 (c-most-enclosing-brace paren-state))
9670 (progn
9671 (goto-char containing-sexp)
9672 (eq (char-after) ?{))
9673 (setq placeholder
9674 (c-looking-at-decl-block
9675 (c-most-enclosing-brace paren-state
9676 containing-sexp)
9677 t)))
9678 (setq containing-decl-open containing-sexp
9679 containing-decl-start (point)
9680 containing-sexp nil)
9681 (goto-char placeholder)
9682 (setq containing-decl-kwd (and (looking-at c-keywords-regexp)
9683 (c-keyword-sym (match-string 1)))))
9684
9685 ;; Init some position variables.
9686 (if paren-state
9687 (progn
9688 (setq containing-sexp (car paren-state)
9689 paren-state (cdr paren-state))
9690 (if (consp containing-sexp)
9691 (save-excursion
9692 (goto-char (cdr containing-sexp))
9693 (if (and (c-major-mode-is 'c++-mode)
9694 (c-back-over-member-initializer-braces))
9695 (c-syntactic-skip-backward "^}" nil t))
9696 (setq lim (point))
9697 (if paren-state
9698 ;; Ignore balanced paren. The next entry
9699 ;; can't be another one.
9700 (setq containing-sexp (car paren-state)
9701 paren-state (cdr paren-state))
9702 ;; If there is no surrounding open paren then
9703 ;; put the last balanced pair back on paren-state.
9704 (setq paren-state (cons containing-sexp paren-state)
9705 containing-sexp nil)))
9706 (setq lim (1+ containing-sexp))))
9707 (setq lim (point-min)))
9708
9709 ;; If we're in a parenthesis list then ',' delimits the
9710 ;; "statements" rather than being an operator (with the
9711 ;; exception of the "for" clause). This difference is
9712 ;; typically only noticeable when statements are used in macro
9713 ;; arglists.
9714 (when (and containing-sexp
9715 (eq (char-after containing-sexp) ?\())
9716 (setq c-stmt-delim-chars c-stmt-delim-chars-with-comma))
9717 ;; cache char before and after indent point, and move point to
9718 ;; the most likely position to perform the majority of tests
9719 (goto-char indent-point)
9720 (c-backward-syntactic-ws lim)
9721 (setq before-ws-ip (point)
9722 char-before-ip (char-before))
9723 (goto-char indent-point)
9724 (skip-chars-forward " \t")
9725 (setq char-after-ip (char-after))
9726
9727 ;; are we in a literal?
9728 (setq literal (c-in-literal lim))
9729
9730 ;; now figure out syntactic qualities of the current line
9731 (cond
9732
9733 ;; CASE 1: in a string.
9734 ((eq literal 'string)
9735 (c-add-syntax 'string (c-point 'bopl)))
9736
9737 ;; CASE 2: in a C or C++ style comment.
9738 ((and (memq literal '(c c++))
9739 ;; This is a kludge for XEmacs where we use
9740 ;; `buffer-syntactic-context', which doesn't correctly
9741 ;; recognize "\*/" to end a block comment.
9742 ;; `parse-partial-sexp' which is used by
9743 ;; `c-literal-limits' will however do that in most
9744 ;; versions, which results in that we get nil from
9745 ;; `c-literal-limits' even when `c-in-literal' claims
9746 ;; we're inside a comment.
9747 (setq placeholder (c-literal-limits lim)))
9748 (c-add-syntax literal (car placeholder)))
9749
9750 ;; CASE 3: in a cpp preprocessor macro continuation.
9751 ((and (save-excursion
9752 (when (c-beginning-of-macro)
9753 (setq macro-start (point))))
9754 (/= macro-start (c-point 'boi))
9755 (progn
9756 (setq tmpsymbol 'cpp-macro-cont)
9757 (or (not c-syntactic-indentation-in-macros)
9758 (save-excursion
9759 (goto-char macro-start)
9760 ;; If at the beginning of the body of a #define
9761 ;; directive then analyze as cpp-define-intro
9762 ;; only. Go on with the syntactic analysis
9763 ;; otherwise. in-macro-expr is set if we're in a
9764 ;; cpp expression, i.e. before the #define body
9765 ;; or anywhere in a non-#define directive.
9766 (if (c-forward-to-cpp-define-body)
9767 (let ((indent-boi (c-point 'boi indent-point)))
9768 (setq in-macro-expr (> (point) indent-boi)
9769 tmpsymbol 'cpp-define-intro)
9770 (= (point) indent-boi))
9771 (setq in-macro-expr t)
9772 nil)))))
9773 (c-add-syntax tmpsymbol macro-start)
9774 (setq macro-start nil))
9775
9776 ;; CASE 11: an else clause?
9777 ((looking-at "else\\>[^_]")
9778 (c-beginning-of-statement-1 containing-sexp)
9779 (c-add-stmt-syntax 'else-clause nil t
9780 containing-sexp paren-state))
9781
9782 ;; CASE 12: while closure of a do/while construct?
9783 ((and (looking-at "while\\>[^_]")
9784 (save-excursion
9785 (prog1 (eq (c-beginning-of-statement-1 containing-sexp)
9786 'beginning)
9787 (setq placeholder (point)))))
9788 (goto-char placeholder)
9789 (c-add-stmt-syntax 'do-while-closure nil t
9790 containing-sexp paren-state))
9791
9792 ;; CASE 13: A catch or finally clause? This case is simpler
9793 ;; than if-else and do-while, because a block is required
9794 ;; after every try, catch and finally.
9795 ((save-excursion
9796 (and (cond ((c-major-mode-is 'c++-mode)
9797 (looking-at "catch\\>[^_]"))
9798 ((c-major-mode-is 'java-mode)
9799 (looking-at "\\(catch\\|finally\\)\\>[^_]")))
9800 (and (c-safe (c-backward-syntactic-ws)
9801 (c-backward-sexp)
9802 t)
9803 (eq (char-after) ?{)
9804 (c-safe (c-backward-syntactic-ws)
9805 (c-backward-sexp)
9806 t)
9807 (if (eq (char-after) ?\()
9808 (c-safe (c-backward-sexp) t)
9809 t))
9810 (looking-at "\\(try\\|catch\\)\\>[^_]")
9811 (setq placeholder (point))))
9812 (goto-char placeholder)
9813 (c-add-stmt-syntax 'catch-clause nil t
9814 containing-sexp paren-state))
9815
9816 ;; CASE 18: A substatement we can recognize by keyword.
9817 ((save-excursion
9818 (and c-opt-block-stmt-key
9819 (not (eq char-before-ip ?\;))
9820 (not (c-at-vsemi-p before-ws-ip))
9821 (not (memq char-after-ip '(?\) ?\] ?,)))
9822 (or (not (eq char-before-ip ?}))
9823 (c-looking-at-inexpr-block-backward c-state-cache))
9824 (> (point)
9825 (progn
9826 ;; Ought to cache the result from the
9827 ;; c-beginning-of-statement-1 calls here.
9828 (setq placeholder (point))
9829 (while (eq (setq step-type
9830 (c-beginning-of-statement-1 lim))
9831 'label))
9832 (if (eq step-type 'previous)
9833 (goto-char placeholder)
9834 (setq placeholder (point))
9835 (if (and (eq step-type 'same)
9836 (not (looking-at c-opt-block-stmt-key)))
9837 ;; Step up to the containing statement if we
9838 ;; stayed in the same one.
9839 (let (step)
9840 (while (eq
9841 (setq step
9842 (c-beginning-of-statement-1 lim))
9843 'label))
9844 (if (eq step 'up)
9845 (setq placeholder (point))
9846 ;; There was no containing statement after all.
9847 (goto-char placeholder)))))
9848 placeholder))
9849 (if (looking-at c-block-stmt-2-key)
9850 ;; Require a parenthesis after these keywords.
9851 ;; Necessary to catch e.g. synchronized in Java,
9852 ;; which can be used both as statement and
9853 ;; modifier.
9854 (and (zerop (c-forward-token-2 1 nil))
9855 (eq (char-after) ?\())
9856 (looking-at c-opt-block-stmt-key))))
9857
9858 (if (eq step-type 'up)
9859 ;; CASE 18A: Simple substatement.
9860 (progn
9861 (goto-char placeholder)
9862 (cond
9863 ((eq char-after-ip ?{)
9864 (c-add-stmt-syntax 'substatement-open nil nil
9865 containing-sexp paren-state))
9866 ((save-excursion
9867 (goto-char indent-point)
9868 (back-to-indentation)
9869 (c-forward-label))
9870 (c-add-stmt-syntax 'substatement-label nil nil
9871 containing-sexp paren-state))
9872 (t
9873 (c-add-stmt-syntax 'substatement nil nil
9874 containing-sexp paren-state))))
9875
9876 ;; CASE 18B: Some other substatement. This is shared
9877 ;; with case 10.
9878 (c-guess-continued-construct indent-point
9879 char-after-ip
9880 placeholder
9881 lim
9882 paren-state)))
9883
9884 ;; CASE 14: A case or default label
9885 ((looking-at c-label-kwds-regexp)
9886 (if containing-sexp
9887 (progn
9888 (goto-char containing-sexp)
9889 (setq lim (c-most-enclosing-brace c-state-cache
9890 containing-sexp))
9891 (c-backward-to-block-anchor lim)
9892 (c-add-stmt-syntax 'case-label nil t lim paren-state))
9893 ;; Got a bogus label at the top level. In lack of better
9894 ;; alternatives, anchor it on (point-min).
9895 (c-add-syntax 'case-label (point-min))))
9896
9897 ;; CASE 15: any other label
9898 ((save-excursion
9899 (back-to-indentation)
9900 (and (not (looking-at c-syntactic-ws-start))
9901 (c-forward-label)))
9902 (cond (containing-decl-open
9903 (setq placeholder (c-add-class-syntax 'inclass
9904 containing-decl-open
9905 containing-decl-start
9906 containing-decl-kwd
9907 paren-state))
9908 ;; Append access-label with the same anchor point as
9909 ;; inclass gets.
9910 (c-append-syntax 'access-label placeholder))
9911
9912 (containing-sexp
9913 (goto-char containing-sexp)
9914 (setq lim (c-most-enclosing-brace c-state-cache
9915 containing-sexp))
9916 (save-excursion
9917 (setq tmpsymbol
9918 (if (and (eq (c-beginning-of-statement-1 lim) 'up)
9919 (looking-at "switch\\>[^_]"))
9920 ;; If the surrounding statement is a switch then
9921 ;; let's analyze all labels as switch labels, so
9922 ;; that they get lined up consistently.
9923 'case-label
9924 'label)))
9925 (c-backward-to-block-anchor lim)
9926 (c-add-stmt-syntax tmpsymbol nil t lim paren-state))
9927
9928 (t
9929 ;; A label on the top level. Treat it as a class
9930 ;; context. (point-min) is the closest we get to the
9931 ;; class open brace.
9932 (c-add-syntax 'access-label (point-min)))))
9933
9934 ;; CASE 4: In-expression statement. C.f. cases 7B, 16A and
9935 ;; 17E.
9936 ((setq placeholder (c-looking-at-inexpr-block
9937 (c-safe-position containing-sexp paren-state)
9938 containing-sexp
9939 ;; Have to turn on the heuristics after
9940 ;; the point even though it doesn't work
9941 ;; very well. C.f. test case class-16.pike.
9942 t))
9943 (setq tmpsymbol (assq (car placeholder)
9944 '((inexpr-class . class-open)
9945 (inexpr-statement . block-open))))
9946 (if tmpsymbol
9947 ;; It's a statement block or an anonymous class.
9948 (setq tmpsymbol (cdr tmpsymbol))
9949 ;; It's a Pike lambda. Check whether we are between the
9950 ;; lambda keyword and the argument list or at the defun
9951 ;; opener.
9952 (setq tmpsymbol (if (eq char-after-ip ?{)
9953 'inline-open
9954 'lambda-intro-cont)))
9955 (goto-char (cdr placeholder))
9956 (back-to-indentation)
9957 (c-add-stmt-syntax tmpsymbol nil t
9958 (c-most-enclosing-brace c-state-cache (point))
9959 paren-state)
9960 (unless (eq (point) (cdr placeholder))
9961 (c-add-syntax (car placeholder))))
9962
9963 ;; CASE 5: Line is inside a declaration level block or at top level.
9964 ((or containing-decl-open (null containing-sexp))
9965 (cond
9966
9967 ;; CASE 5A: we are looking at a defun, brace list, class,
9968 ;; or inline-inclass method opening brace
9969 ((setq special-brace-list
9970 (or (and c-special-brace-lists
9971 (c-looking-at-special-brace-list))
9972 (eq char-after-ip ?{)))
9973 (cond
9974
9975 ;; CASE 5A.1: Non-class declaration block open.
9976 ((save-excursion
9977 (let (tmp)
9978 (and (eq char-after-ip ?{)
9979 (setq tmp (c-looking-at-decl-block containing-sexp t))
9980 (progn
9981 (setq placeholder (point))
9982 (goto-char tmp)
9983 (looking-at c-symbol-key))
9984 (c-keyword-member
9985 (c-keyword-sym (setq keyword (match-string 0)))
9986 'c-other-block-decl-kwds))))
9987 (goto-char placeholder)
9988 (c-add-stmt-syntax
9989 (if (string-equal keyword "extern")
9990 ;; Special case for extern-lang-open.
9991 'extern-lang-open
9992 (intern (concat keyword "-open")))
9993 nil t containing-sexp paren-state))
9994
9995 ;; CASE 5A.2: we are looking at a class opening brace
9996 ((save-excursion
9997 (goto-char indent-point)
9998 (skip-chars-forward " \t")
9999 (and (eq (char-after) ?{)
10000 (c-looking-at-decl-block containing-sexp t)
10001 (setq placeholder (point))))
10002 (c-add-syntax 'class-open placeholder))
10003
10004 ;; CASE 5A.3: brace list open
10005 ((save-excursion
10006 (c-beginning-of-decl-1 lim)
10007 (while (looking-at c-specifier-key)
10008 (goto-char (match-end 1))
10009 (c-forward-syntactic-ws indent-point))
10010 (setq placeholder (c-point 'boi))
10011 (or (consp special-brace-list)
10012 (and (or (save-excursion
10013 (goto-char indent-point)
10014 (setq tmpsymbol nil)
10015 (while (and (> (point) placeholder)
10016 (zerop (c-backward-token-2 1 t))
10017 (not (looking-at "=\\([^=]\\|$\\)")))
10018 (and c-opt-inexpr-brace-list-key
10019 (not tmpsymbol)
10020 (looking-at c-opt-inexpr-brace-list-key)
10021 (setq tmpsymbol 'topmost-intro-cont)))
10022 (looking-at "=\\([^=]\\|$\\)"))
10023 (looking-at c-brace-list-key))
10024 (save-excursion
10025 (while (and (< (point) indent-point)
10026 (zerop (c-forward-token-2 1 t))
10027 (not (memq (char-after) '(?\; ?\()))))
10028 (not (memq (char-after) '(?\; ?\()))
10029 ))))
10030 (if (and (not c-auto-newline-analysis)
10031 (c-major-mode-is 'java-mode)
10032 (eq tmpsymbol 'topmost-intro-cont))
10033 ;; We're in Java and have found that the open brace
10034 ;; belongs to a "new Foo[]" initialization list,
10035 ;; which means the brace list is part of an
10036 ;; expression and not a top level definition. We
10037 ;; therefore treat it as any topmost continuation
10038 ;; even though the semantically correct symbol still
10039 ;; is brace-list-open, on the same grounds as in
10040 ;; case B.2.
10041 (progn
10042 (c-beginning-of-statement-1 lim)
10043 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
10044 (c-add-syntax 'brace-list-open placeholder)))
10045
10046 ;; CASE 5A.4: inline defun open
10047 ((and containing-decl-open
10048 (not (c-keyword-member containing-decl-kwd
10049 'c-other-block-decl-kwds)))
10050 (c-add-syntax 'inline-open)
10051 (c-add-class-syntax 'inclass
10052 containing-decl-open
10053 containing-decl-start
10054 containing-decl-kwd
10055 paren-state))
10056
10057 ;; CASE 5A.5: ordinary defun open
10058 (t
10059 (save-excursion
10060 (c-beginning-of-decl-1 lim)
10061 (while (looking-at c-specifier-key)
10062 (goto-char (match-end 1))
10063 (c-forward-syntactic-ws indent-point))
10064 (c-add-syntax 'defun-open (c-point 'boi))
10065 ;; Bogus to use bol here, but it's the legacy. (Resolved,
10066 ;; 2007-11-09)
10067 ))))
10068
10069 ;; CASE 5R: Member init list. (Used to be part of CASE 5B.1)
10070 ;; Note there is no limit on the backward search here, since member
10071 ;; init lists can, in practice, be very large.
10072 ((save-excursion
10073 (when (and (c-major-mode-is 'c++-mode)
10074 (setq placeholder (c-back-over-member-initializers)))
10075 (setq tmp-pos (point))))
10076 (if (= (c-point 'bosws) (1+ tmp-pos))
10077 (progn
10078 ;; There is no preceding member init clause.
10079 ;; Indent relative to the beginning of indentation
10080 ;; for the topmost-intro line that contains the
10081 ;; prototype's open paren.
10082 (goto-char placeholder)
10083 (c-add-syntax 'member-init-intro (c-point 'boi)))
10084 ;; Indent relative to the first member init clause.
10085 (goto-char (1+ tmp-pos))
10086 (c-forward-syntactic-ws)
10087 (c-add-syntax 'member-init-cont (point))))
10088
10089 ;; CASE 5B: After a function header but before the body (or
10090 ;; the ending semicolon if there's no body).
10091 ((save-excursion
10092 (when (setq placeholder (c-just-after-func-arglist-p
10093 (max lim (c-determine-limit 500))))
10094 (setq tmp-pos (point))))
10095 (cond
10096
10097 ;; CASE 5B.1: Member init list.
10098 ((eq (char-after tmp-pos) ?:)
10099 ;; There is no preceding member init clause.
10100 ;; Indent relative to the beginning of indentation
10101 ;; for the topmost-intro line that contains the
10102 ;; prototype's open paren.
10103 (goto-char placeholder)
10104 (c-add-syntax 'member-init-intro (c-point 'boi)))
10105
10106 ;; CASE 5B.2: K&R arg decl intro
10107 ((and c-recognize-knr-p
10108 (c-in-knr-argdecl lim))
10109 (c-beginning-of-statement-1 lim)
10110 (c-add-syntax 'knr-argdecl-intro (c-point 'boi))
10111 (if containing-decl-open
10112 (c-add-class-syntax 'inclass
10113 containing-decl-open
10114 containing-decl-start
10115 containing-decl-kwd
10116 paren-state)))
10117
10118 ;; CASE 5B.4: Nether region after a C++ or Java func
10119 ;; decl, which could include a `throws' declaration.
10120 (t
10121 (c-beginning-of-statement-1 lim)
10122 (c-add-syntax 'func-decl-cont (c-point 'boi))
10123 )))
10124
10125 ;; CASE 5C: inheritance line. could be first inheritance
10126 ;; line, or continuation of a multiple inheritance
10127 ((or (and (c-major-mode-is 'c++-mode)
10128 (progn
10129 (when (eq char-after-ip ?,)
10130 (skip-chars-forward " \t")
10131 (forward-char))
10132 (looking-at c-opt-postfix-decl-spec-key)))
10133 (and (or (eq char-before-ip ?:)
10134 ;; watch out for scope operator
10135 (save-excursion
10136 (and (eq char-after-ip ?:)
10137 (c-safe (forward-char 1) t)
10138 (not (eq (char-after) ?:))
10139 )))
10140 (save-excursion
10141 (c-beginning-of-statement-1 lim)
10142 (when (looking-at c-opt-<>-sexp-key)
10143 (goto-char (match-end 1))
10144 (c-forward-syntactic-ws)
10145 (c-forward-<>-arglist nil)
10146 (c-forward-syntactic-ws))
10147 (looking-at c-class-key)))
10148 ;; for Java
10149 (and (c-major-mode-is 'java-mode)
10150 (let ((fence (save-excursion
10151 (c-beginning-of-statement-1 lim)
10152 (point)))
10153 cont done)
10154 (save-excursion
10155 (while (not done)
10156 (cond ((looking-at c-opt-postfix-decl-spec-key)
10157 (setq injava-inher (cons cont (point))
10158 done t))
10159 ((or (not (c-safe (c-forward-sexp -1) t))
10160 (<= (point) fence))
10161 (setq done t))
10162 )
10163 (setq cont t)))
10164 injava-inher)
10165 (not (c-crosses-statement-barrier-p (cdr injava-inher)
10166 (point)))
10167 ))
10168 (cond
10169
10170 ;; CASE 5C.1: non-hanging colon on an inher intro
10171 ((eq char-after-ip ?:)
10172 (c-beginning-of-statement-1 lim)
10173 (c-add-syntax 'inher-intro (c-point 'boi))
10174 ;; don't add inclass symbol since relative point already
10175 ;; contains any class offset
10176 )
10177
10178 ;; CASE 5C.2: hanging colon on an inher intro
10179 ((eq char-before-ip ?:)
10180 (c-beginning-of-statement-1 lim)
10181 (c-add-syntax 'inher-intro (c-point 'boi))
10182 (if containing-decl-open
10183 (c-add-class-syntax 'inclass
10184 containing-decl-open
10185 containing-decl-start
10186 containing-decl-kwd
10187 paren-state)))
10188
10189 ;; CASE 5C.3: in a Java implements/extends
10190 (injava-inher
10191 (let ((where (cdr injava-inher))
10192 (cont (car injava-inher)))
10193 (goto-char where)
10194 (cond ((looking-at "throws\\>[^_]")
10195 (c-add-syntax 'func-decl-cont
10196 (progn (c-beginning-of-statement-1 lim)
10197 (c-point 'boi))))
10198 (cont (c-add-syntax 'inher-cont where))
10199 (t (c-add-syntax 'inher-intro
10200 (progn (goto-char (cdr injava-inher))
10201 (c-beginning-of-statement-1 lim)
10202 (point))))
10203 )))
10204
10205 ;; CASE 5C.4: a continued inheritance line
10206 (t
10207 (c-beginning-of-inheritance-list lim)
10208 (c-add-syntax 'inher-cont (point))
10209 ;; don't add inclass symbol since relative point already
10210 ;; contains any class offset
10211 )))
10212
10213 ;; CASE 5P: AWK pattern or function or continuation
10214 ;; thereof.
10215 ((c-major-mode-is 'awk-mode)
10216 (setq placeholder (point))
10217 (c-add-stmt-syntax
10218 (if (and (eq (c-beginning-of-statement-1) 'same)
10219 (/= (point) placeholder))
10220 'topmost-intro-cont
10221 'topmost-intro)
10222 nil nil
10223 containing-sexp paren-state))
10224
10225 ;; CASE 5D: this could be a top-level initialization, a
10226 ;; member init list continuation, or a template argument
10227 ;; list continuation.
10228 ((save-excursion
10229 ;; Note: We use the fact that lim is always after any
10230 ;; preceding brace sexp.
10231 (if c-recognize-<>-arglists
10232 (while (and
10233 (progn
10234 (c-syntactic-skip-backward "^;,=<>" lim t)
10235 (> (point) lim))
10236 (or
10237 (when c-overloadable-operators-regexp
10238 (when (setq placeholder (c-after-special-operator-id lim))
10239 (goto-char placeholder)
10240 t))
10241 (cond
10242 ((eq (char-before) ?>)
10243 (or (c-backward-<>-arglist nil lim)
10244 (backward-char))
10245 t)
10246 ((eq (char-before) ?<)
10247 (backward-char)
10248 (if (save-excursion
10249 (c-forward-<>-arglist nil))
10250 (progn (forward-char)
10251 nil)
10252 t))
10253 (t nil)))))
10254 ;; NB: No c-after-special-operator-id stuff in this
10255 ;; clause - we assume only C++ needs it.
10256 (c-syntactic-skip-backward "^;,=" lim t))
10257 (memq (char-before) '(?, ?= ?<)))
10258 (cond
10259
10260 ;; CASE 5D.3: perhaps a template list continuation?
10261 ((and (c-major-mode-is 'c++-mode)
10262 (save-excursion
10263 (save-restriction
10264 (c-with-syntax-table c++-template-syntax-table
10265 (goto-char indent-point)
10266 (setq placeholder (c-up-list-backward))
10267 (and placeholder
10268 (eq (char-after placeholder) ?<))))))
10269 (c-with-syntax-table c++-template-syntax-table
10270 (goto-char placeholder)
10271 (c-beginning-of-statement-1 lim t))
10272 (if (save-excursion
10273 (c-backward-syntactic-ws lim)
10274 (eq (char-before) ?<))
10275 ;; In a nested template arglist.
10276 (progn
10277 (goto-char placeholder)
10278 (c-syntactic-skip-backward "^,;" lim t)
10279 (c-forward-syntactic-ws))
10280 (back-to-indentation))
10281 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
10282 ;; template aware.
10283 (c-add-syntax 'template-args-cont (point) placeholder))
10284
10285 ;; CASE 5D.4: perhaps a multiple inheritance line?
10286 ((and (c-major-mode-is 'c++-mode)
10287 (save-excursion
10288 (c-beginning-of-statement-1 lim)
10289 (setq placeholder (point))
10290 (if (looking-at "static\\>[^_]")
10291 (c-forward-token-2 1 nil indent-point))
10292 (and (looking-at c-class-key)
10293 (zerop (c-forward-token-2 2 nil indent-point))
10294 (if (eq (char-after) ?<)
10295 (c-with-syntax-table c++-template-syntax-table
10296 (zerop (c-forward-token-2 1 t indent-point)))
10297 t)
10298 (eq (char-after) ?:))))
10299 (goto-char placeholder)
10300 (c-add-syntax 'inher-cont (c-point 'boi)))
10301
10302 ;; CASE 5D.5: Continuation of the "expression part" of a
10303 ;; top level construct. Or, perhaps, an unrecognized construct.
10304 (t
10305 (while (and (setq placeholder (point))
10306 (eq (car (c-beginning-of-decl-1 containing-sexp)) ; Can't use `lim' here.
10307 'same)
10308 (save-excursion
10309 (c-backward-syntactic-ws)
10310 (eq (char-before) ?}))
10311 (< (point) placeholder)))
10312 (c-add-stmt-syntax
10313 (cond
10314 ((eq (point) placeholder) 'statement) ; unrecognized construct
10315 ;; A preceding comma at the top level means that a
10316 ;; new variable declaration starts here. Use
10317 ;; topmost-intro-cont for it, for consistency with
10318 ;; the first variable declaration. C.f. case 5N.
10319 ((eq char-before-ip ?,) 'topmost-intro-cont)
10320 (t 'statement-cont))
10321 nil nil containing-sexp paren-state))
10322 ))
10323
10324 ;; CASE 5F: Close of a non-class declaration level block.
10325 ((and (eq char-after-ip ?})
10326 (c-keyword-member containing-decl-kwd
10327 'c-other-block-decl-kwds))
10328 ;; This is inconsistent: Should use `containing-decl-open'
10329 ;; here if it's at boi, like in case 5J.
10330 (goto-char containing-decl-start)
10331 (c-add-stmt-syntax
10332 (if (string-equal (symbol-name containing-decl-kwd) "extern")
10333 ;; Special case for compatibility with the
10334 ;; extern-lang syntactic symbols.
10335 'extern-lang-close
10336 (intern (concat (symbol-name containing-decl-kwd)
10337 "-close")))
10338 nil t
10339 (c-most-enclosing-brace paren-state (point))
10340 paren-state))
10341
10342 ;; CASE 5G: we are looking at the brace which closes the
10343 ;; enclosing nested class decl
10344 ((and containing-sexp
10345 (eq char-after-ip ?})
10346 (eq containing-decl-open containing-sexp))
10347 (c-add-class-syntax 'class-close
10348 containing-decl-open
10349 containing-decl-start
10350 containing-decl-kwd
10351 paren-state))
10352
10353 ;; CASE 5H: we could be looking at subsequent knr-argdecls
10354 ((and c-recognize-knr-p
10355 (not containing-sexp) ; can't be knr inside braces.
10356 (not (eq char-before-ip ?}))
10357 (save-excursion
10358 (setq placeholder (cdr (c-beginning-of-decl-1 lim)))
10359 (and placeholder
10360 ;; Do an extra check to avoid tripping up on
10361 ;; statements that occur in invalid contexts
10362 ;; (e.g. in macro bodies where we don't really
10363 ;; know the context of what we're looking at).
10364 (not (and c-opt-block-stmt-key
10365 (looking-at c-opt-block-stmt-key)))))
10366 (< placeholder indent-point))
10367 (goto-char placeholder)
10368 (c-add-syntax 'knr-argdecl (point)))
10369
10370 ;; CASE 5I: ObjC method definition.
10371 ((and c-opt-method-key
10372 (looking-at c-opt-method-key))
10373 (c-beginning-of-statement-1 nil t)
10374 (if (= (point) indent-point)
10375 ;; Handle the case when it's the first (non-comment)
10376 ;; thing in the buffer. Can't look for a 'same return
10377 ;; value from cbos1 since ObjC directives currently
10378 ;; aren't recognized fully, so that we get 'same
10379 ;; instead of 'previous if it moved over a preceding
10380 ;; directive.
10381 (goto-char (point-min)))
10382 (c-add-syntax 'objc-method-intro (c-point 'boi)))
10383
10384 ;; CASE 5N: At a variable declaration that follows a class
10385 ;; definition or some other block declaration that doesn't
10386 ;; end at the closing '}'. C.f. case 5D.5.
10387 ((progn
10388 (c-backward-syntactic-ws lim)
10389 (and (eq (char-before) ?})
10390 (save-excursion
10391 (let ((start (point)))
10392 (if (and c-state-cache
10393 (consp (car c-state-cache))
10394 (eq (cdar c-state-cache) (point)))
10395 ;; Speed up the backward search a bit.
10396 (goto-char (caar c-state-cache)))
10397 (c-beginning-of-decl-1 containing-sexp) ; Can't use `lim' here.
10398 (setq placeholder (point))
10399 (if (= start (point))
10400 ;; The '}' is unbalanced.
10401 nil
10402 (c-end-of-decl-1)
10403 (>= (point) indent-point))))))
10404 (goto-char placeholder)
10405 (c-add-stmt-syntax 'topmost-intro-cont nil nil
10406 containing-sexp paren-state))
10407
10408 ;; NOTE: The point is at the end of the previous token here.
10409
10410 ;; CASE 5J: we are at the topmost level, make
10411 ;; sure we skip back past any access specifiers
10412 ((and
10413 ;; A macro continuation line is never at top level.
10414 (not (and macro-start
10415 (> indent-point macro-start)))
10416 (save-excursion
10417 (setq placeholder (point))
10418 (or (memq char-before-ip '(?\; ?{ ?} nil))
10419 (c-at-vsemi-p before-ws-ip)
10420 (when (and (eq char-before-ip ?:)
10421 (eq (c-beginning-of-statement-1 lim)
10422 'label))
10423 (c-backward-syntactic-ws lim)
10424 (setq placeholder (point)))
10425 (and (c-major-mode-is 'objc-mode)
10426 (catch 'not-in-directive
10427 (c-beginning-of-statement-1 lim)
10428 (setq placeholder (point))
10429 (while (and (c-forward-objc-directive)
10430 (< (point) indent-point))
10431 (c-forward-syntactic-ws)
10432 (if (>= (point) indent-point)
10433 (throw 'not-in-directive t))
10434 (setq placeholder (point)))
10435 nil)))))
10436 ;; For historic reasons we anchor at bol of the last
10437 ;; line of the previous declaration. That's clearly
10438 ;; highly bogus and useless, and it makes our lives hard
10439 ;; to remain compatible. :P
10440 (goto-char placeholder)
10441 (c-add-syntax 'topmost-intro (c-point 'bol))
10442 (if containing-decl-open
10443 (if (c-keyword-member containing-decl-kwd
10444 'c-other-block-decl-kwds)
10445 (progn
10446 (goto-char (c-brace-anchor-point containing-decl-open))
10447 (c-add-stmt-syntax
10448 (if (string-equal (symbol-name containing-decl-kwd)
10449 "extern")
10450 ;; Special case for compatibility with the
10451 ;; extern-lang syntactic symbols.
10452 'inextern-lang
10453 (intern (concat "in"
10454 (symbol-name containing-decl-kwd))))
10455 nil t
10456 (c-most-enclosing-brace paren-state (point))
10457 paren-state))
10458 (c-add-class-syntax 'inclass
10459 containing-decl-open
10460 containing-decl-start
10461 containing-decl-kwd
10462 paren-state)))
10463 (when (and c-syntactic-indentation-in-macros
10464 macro-start
10465 (/= macro-start (c-point 'boi indent-point)))
10466 (c-add-syntax 'cpp-define-intro)
10467 (setq macro-start nil)))
10468
10469 ;; CASE 5K: we are at an ObjC method definition
10470 ;; continuation line.
10471 ((and c-opt-method-key
10472 (save-excursion
10473 (c-beginning-of-statement-1 lim)
10474 (beginning-of-line)
10475 (when (looking-at c-opt-method-key)
10476 (setq placeholder (point)))))
10477 (c-add-syntax 'objc-method-args-cont placeholder))
10478
10479 ;; CASE 5L: we are at the first argument of a template
10480 ;; arglist that begins on the previous line.
10481 ((and c-recognize-<>-arglists
10482 (eq (char-before) ?<)
10483 (not (and c-overloadable-operators-regexp
10484 (c-after-special-operator-id lim))))
10485 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
10486 (c-add-syntax 'template-args-cont (c-point 'boi)))
10487
10488 ;; CASE 5Q: we are at a statement within a macro.
10489 (macro-start
10490 (c-beginning-of-statement-1 containing-sexp)
10491 (c-add-stmt-syntax 'statement nil t containing-sexp paren-state))
10492
10493 ;;CASE 5N: We are at a topmost continuation line and the only
10494 ;;preceding items are annotations.
10495 ((and (c-major-mode-is 'java-mode)
10496 (setq placeholder (point))
10497 (c-beginning-of-statement-1)
10498 (progn
10499 (while (and (c-forward-annotation))
10500 (c-forward-syntactic-ws))
10501 t)
10502 (prog1
10503 (>= (point) placeholder)
10504 (goto-char placeholder)))
10505 (c-add-syntax 'annotation-top-cont (c-point 'boi)))
10506
10507 ;; CASE 5M: we are at a topmost continuation line
10508 (t
10509 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
10510 (when (c-major-mode-is 'objc-mode)
10511 (setq placeholder (point))
10512 (while (and (c-forward-objc-directive)
10513 (< (point) indent-point))
10514 (c-forward-syntactic-ws)
10515 (setq placeholder (point)))
10516 (goto-char placeholder))
10517 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
10518 ))
10519
10520 ;; (CASE 6 has been removed.)
10521
10522 ;; CASE 7: line is an expression, not a statement. Most
10523 ;; likely we are either in a function prototype or a function
10524 ;; call argument list
10525 ((not (or (and c-special-brace-lists
10526 (save-excursion
10527 (goto-char containing-sexp)
10528 (c-looking-at-special-brace-list)))
10529 (eq (char-after containing-sexp) ?{)))
10530 (cond
10531
10532 ;; CASE 7A: we are looking at the arglist closing paren.
10533 ;; C.f. case 7F.
10534 ((memq char-after-ip '(?\) ?\]))
10535 (goto-char containing-sexp)
10536 (setq placeholder (c-point 'boi))
10537 (if (and (c-safe (backward-up-list 1) t)
10538 (>= (point) placeholder))
10539 (progn
10540 (forward-char)
10541 (skip-chars-forward " \t"))
10542 (goto-char placeholder))
10543 (c-add-stmt-syntax 'arglist-close (list containing-sexp) t
10544 (c-most-enclosing-brace paren-state (point))
10545 paren-state))
10546
10547 ;; CASE 7B: Looking at the opening brace of an
10548 ;; in-expression block or brace list. C.f. cases 4, 16A
10549 ;; and 17E.
10550 ((and (eq char-after-ip ?{)
10551 (progn
10552 (setq placeholder (c-inside-bracelist-p (point)
10553 paren-state))
10554 (if placeholder
10555 (setq tmpsymbol '(brace-list-open . inexpr-class))
10556 (setq tmpsymbol '(block-open . inexpr-statement)
10557 placeholder
10558 (cdr-safe (c-looking-at-inexpr-block
10559 (c-safe-position containing-sexp
10560 paren-state)
10561 containing-sexp)))
10562 ;; placeholder is nil if it's a block directly in
10563 ;; a function arglist. That makes us skip out of
10564 ;; this case.
10565 )))
10566 (goto-char placeholder)
10567 (back-to-indentation)
10568 (c-add-stmt-syntax (car tmpsymbol) nil t
10569 (c-most-enclosing-brace paren-state (point))
10570 paren-state)
10571 (if (/= (point) placeholder)
10572 (c-add-syntax (cdr tmpsymbol))))
10573
10574 ;; CASE 7C: we are looking at the first argument in an empty
10575 ;; argument list. Use arglist-close if we're actually
10576 ;; looking at a close paren or bracket.
10577 ((memq char-before-ip '(?\( ?\[))
10578 (goto-char containing-sexp)
10579 (setq placeholder (c-point 'boi))
10580 (if (and (c-safe (backward-up-list 1) t)
10581 (>= (point) placeholder))
10582 (progn
10583 (forward-char)
10584 (skip-chars-forward " \t"))
10585 (goto-char placeholder))
10586 (c-add-stmt-syntax 'arglist-intro (list containing-sexp) t
10587 (c-most-enclosing-brace paren-state (point))
10588 paren-state))
10589
10590 ;; CASE 7D: we are inside a conditional test clause. treat
10591 ;; these things as statements
10592 ((progn
10593 (goto-char containing-sexp)
10594 (and (c-safe (c-forward-sexp -1) t)
10595 (looking-at "\\<for\\>[^_]")))
10596 (goto-char (1+ containing-sexp))
10597 (c-forward-syntactic-ws indent-point)
10598 (if (eq char-before-ip ?\;)
10599 (c-add-syntax 'statement (point))
10600 (c-add-syntax 'statement-cont (point))
10601 ))
10602
10603 ;; CASE 7E: maybe a continued ObjC method call. This is the
10604 ;; case when we are inside a [] bracketed exp, and what
10605 ;; precede the opening bracket is not an identifier.
10606 ((and c-opt-method-key
10607 (eq (char-after containing-sexp) ?\[)
10608 (progn
10609 (goto-char (1- containing-sexp))
10610 (c-backward-syntactic-ws (c-point 'bod))
10611 (if (not (looking-at c-symbol-key))
10612 (c-add-syntax 'objc-method-call-cont containing-sexp))
10613 )))
10614
10615 ;; CASE 7F: we are looking at an arglist continuation line,
10616 ;; but the preceding argument is on the same line as the
10617 ;; opening paren. This case includes multi-line
10618 ;; mathematical paren groupings, but we could be on a
10619 ;; for-list continuation line. C.f. case 7A.
10620 ((progn
10621 (goto-char (1+ containing-sexp))
10622 (< (save-excursion
10623 (c-forward-syntactic-ws)
10624 (point))
10625 (c-point 'bonl)))
10626 (goto-char containing-sexp) ; paren opening the arglist
10627 (setq placeholder (c-point 'boi))
10628 (if (and (c-safe (backward-up-list 1) t)
10629 (>= (point) placeholder))
10630 (progn
10631 (forward-char)
10632 (skip-chars-forward " \t"))
10633 (goto-char placeholder))
10634 (c-add-stmt-syntax 'arglist-cont-nonempty (list containing-sexp) t
10635 (c-most-enclosing-brace c-state-cache (point))
10636 paren-state))
10637
10638 ;; CASE 7G: we are looking at just a normal arglist
10639 ;; continuation line
10640 (t (c-forward-syntactic-ws indent-point)
10641 (c-add-syntax 'arglist-cont (c-point 'boi)))
10642 ))
10643
10644 ;; CASE 8: func-local multi-inheritance line
10645 ((and (c-major-mode-is 'c++-mode)
10646 (save-excursion
10647 (goto-char indent-point)
10648 (skip-chars-forward " \t")
10649 (looking-at c-opt-postfix-decl-spec-key)))
10650 (goto-char indent-point)
10651 (skip-chars-forward " \t")
10652 (cond
10653
10654 ;; CASE 8A: non-hanging colon on an inher intro
10655 ((eq char-after-ip ?:)
10656 (c-backward-syntactic-ws lim)
10657 (c-add-syntax 'inher-intro (c-point 'boi)))
10658
10659 ;; CASE 8B: hanging colon on an inher intro
10660 ((eq char-before-ip ?:)
10661 (c-add-syntax 'inher-intro (c-point 'boi)))
10662
10663 ;; CASE 8C: a continued inheritance line
10664 (t
10665 (c-beginning-of-inheritance-list lim)
10666 (c-add-syntax 'inher-cont (point))
10667 )))
10668
10669 ;; CASE 9: we are inside a brace-list
10670 ((and (not (c-major-mode-is 'awk-mode)) ; Maybe this isn't needed (ACM, 2002/3/29)
10671 (setq special-brace-list
10672 (or (and c-special-brace-lists ;;;; ALWAYS NIL FOR AWK!!
10673 (save-excursion
10674 (goto-char containing-sexp)
10675 (c-looking-at-special-brace-list)))
10676 (c-inside-bracelist-p containing-sexp paren-state))))
10677 (cond
10678
10679 ;; CASE 9A: In the middle of a special brace list opener.
10680 ((and (consp special-brace-list)
10681 (save-excursion
10682 (goto-char containing-sexp)
10683 (eq (char-after) ?\())
10684 (eq char-after-ip (car (cdr special-brace-list))))
10685 (goto-char (car (car special-brace-list)))
10686 (skip-chars-backward " \t")
10687 (if (and (bolp)
10688 (assoc 'statement-cont
10689 (setq placeholder (c-guess-basic-syntax))))
10690 (setq c-syntactic-context placeholder)
10691 (c-beginning-of-statement-1
10692 (c-safe-position (1- containing-sexp) paren-state))
10693 (c-forward-token-2 0)
10694 (while (looking-at c-specifier-key)
10695 (goto-char (match-end 1))
10696 (c-forward-syntactic-ws))
10697 (c-add-syntax 'brace-list-open (c-point 'boi))))
10698
10699 ;; CASE 9B: brace-list-close brace
10700 ((if (consp special-brace-list)
10701 ;; Check special brace list closer.
10702 (progn
10703 (goto-char (car (car special-brace-list)))
10704 (save-excursion
10705 (goto-char indent-point)
10706 (back-to-indentation)
10707 (or
10708 ;; We were between the special close char and the `)'.
10709 (and (eq (char-after) ?\))
10710 (eq (1+ (point)) (cdr (car special-brace-list))))
10711 ;; We were before the special close char.
10712 (and (eq (char-after) (cdr (cdr special-brace-list)))
10713 (zerop (c-forward-token-2))
10714 (eq (1+ (point)) (cdr (car special-brace-list)))))))
10715 ;; Normal brace list check.
10716 (and (eq char-after-ip ?})
10717 (c-safe (goto-char (c-up-list-backward (point))) t)
10718 (= (point) containing-sexp)))
10719 (if (eq (point) (c-point 'boi))
10720 (c-add-syntax 'brace-list-close (point))
10721 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10722 (c-beginning-of-statement-1 lim nil nil t)
10723 (c-add-stmt-syntax 'brace-list-close nil t lim paren-state)))
10724
10725 (t
10726 ;; Prepare for the rest of the cases below by going to the
10727 ;; token following the opening brace
10728 (if (consp special-brace-list)
10729 (progn
10730 (goto-char (car (car special-brace-list)))
10731 (c-forward-token-2 1 nil indent-point))
10732 (goto-char containing-sexp))
10733 (forward-char)
10734 (let ((start (point)))
10735 (c-forward-syntactic-ws indent-point)
10736 (goto-char (max start (c-point 'bol))))
10737 (c-skip-ws-forward indent-point)
10738 (cond
10739
10740 ;; CASE 9C: we're looking at the first line in a brace-list
10741 ((= (point) indent-point)
10742 (if (consp special-brace-list)
10743 (goto-char (car (car special-brace-list)))
10744 (goto-char containing-sexp))
10745 (if (eq (point) (c-point 'boi))
10746 (c-add-syntax 'brace-list-intro (point))
10747 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10748 (c-beginning-of-statement-1 lim)
10749 (c-add-stmt-syntax 'brace-list-intro nil t lim paren-state)))
10750
10751 ;; CASE 9D: this is just a later brace-list-entry or
10752 ;; brace-entry-open
10753 (t (if (or (eq char-after-ip ?{)
10754 (and c-special-brace-lists
10755 (save-excursion
10756 (goto-char indent-point)
10757 (c-forward-syntactic-ws (c-point 'eol))
10758 (c-looking-at-special-brace-list (point)))))
10759 (c-add-syntax 'brace-entry-open (point))
10760 (c-add-syntax 'brace-list-entry (point))
10761 ))
10762 ))))
10763
10764 ;; CASE 10: A continued statement or top level construct.
10765 ((and (not (memq char-before-ip '(?\; ?:)))
10766 (not (c-at-vsemi-p before-ws-ip))
10767 (or (not (eq char-before-ip ?}))
10768 (c-looking-at-inexpr-block-backward c-state-cache))
10769 (> (point)
10770 (save-excursion
10771 (c-beginning-of-statement-1 containing-sexp)
10772 (setq placeholder (point))))
10773 (/= placeholder containing-sexp))
10774 ;; This is shared with case 18.
10775 (c-guess-continued-construct indent-point
10776 char-after-ip
10777 placeholder
10778 containing-sexp
10779 paren-state))
10780
10781 ;; CASE 16: block close brace, possibly closing the defun or
10782 ;; the class
10783 ((eq char-after-ip ?})
10784 ;; From here on we have the next containing sexp in lim.
10785 (setq lim (c-most-enclosing-brace paren-state))
10786 (goto-char containing-sexp)
10787 (cond
10788
10789 ;; CASE 16E: Closing a statement block? This catches
10790 ;; cases where it's preceded by a statement keyword,
10791 ;; which works even when used in an "invalid" context,
10792 ;; e.g. a macro argument.
10793 ((c-after-conditional)
10794 (c-backward-to-block-anchor lim)
10795 (c-add-stmt-syntax 'block-close nil t lim paren-state))
10796
10797 ;; CASE 16A: closing a lambda defun or an in-expression
10798 ;; block? C.f. cases 4, 7B and 17E.
10799 ((setq placeholder (c-looking-at-inexpr-block
10800 (c-safe-position containing-sexp paren-state)
10801 nil))
10802 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10803 'inline-close
10804 'block-close))
10805 (goto-char containing-sexp)
10806 (back-to-indentation)
10807 (if (= containing-sexp (point))
10808 (c-add-syntax tmpsymbol (point))
10809 (goto-char (cdr placeholder))
10810 (back-to-indentation)
10811 (c-add-stmt-syntax tmpsymbol nil t
10812 (c-most-enclosing-brace paren-state (point))
10813 paren-state)
10814 (if (/= (point) (cdr placeholder))
10815 (c-add-syntax (car placeholder)))))
10816
10817 ;; CASE 16B: does this close an inline or a function in
10818 ;; a non-class declaration level block?
10819 ((save-excursion
10820 (and lim
10821 (progn
10822 (goto-char lim)
10823 (c-looking-at-decl-block
10824 (c-most-enclosing-brace paren-state lim)
10825 nil))
10826 (setq placeholder (point))))
10827 (c-backward-to-decl-anchor lim)
10828 (back-to-indentation)
10829 (if (save-excursion
10830 (goto-char placeholder)
10831 (looking-at c-other-decl-block-key))
10832 (c-add-syntax 'defun-close (point))
10833 (c-add-syntax 'inline-close (point))))
10834
10835 ;; CASE 16F: Can be a defun-close of a function declared
10836 ;; in a statement block, e.g. in Pike or when using gcc
10837 ;; extensions, but watch out for macros followed by
10838 ;; blocks. Let it through to be handled below.
10839 ;; C.f. cases B.3 and 17G.
10840 ((save-excursion
10841 (and (not (c-at-statement-start-p))
10842 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
10843 (setq placeholder (point))
10844 (let ((c-recognize-typeless-decls nil))
10845 ;; Turn off recognition of constructs that
10846 ;; lacks a type in this case, since that's more
10847 ;; likely to be a macro followed by a block.
10848 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10849 (back-to-indentation)
10850 (if (/= (point) containing-sexp)
10851 (goto-char placeholder))
10852 (c-add-stmt-syntax 'defun-close nil t lim paren-state))
10853
10854 ;; CASE 16C: If there is an enclosing brace then this is
10855 ;; a block close since defun closes inside declaration
10856 ;; level blocks have been handled above.
10857 (lim
10858 ;; If the block is preceded by a case/switch label on
10859 ;; the same line, we anchor at the first preceding label
10860 ;; at boi. The default handling in c-add-stmt-syntax
10861 ;; really fixes it better, but we do like this to keep
10862 ;; the indentation compatible with version 5.28 and
10863 ;; earlier. C.f. case 17H.
10864 (while (and (/= (setq placeholder (point)) (c-point 'boi))
10865 (eq (c-beginning-of-statement-1 lim) 'label)))
10866 (goto-char placeholder)
10867 (if (looking-at c-label-kwds-regexp)
10868 (c-add-syntax 'block-close (point))
10869 (goto-char containing-sexp)
10870 ;; c-backward-to-block-anchor not necessary here; those
10871 ;; situations are handled in case 16E above.
10872 (c-add-stmt-syntax 'block-close nil t lim paren-state)))
10873
10874 ;; CASE 16D: Only top level defun close left.
10875 (t
10876 (goto-char containing-sexp)
10877 (c-backward-to-decl-anchor lim)
10878 (c-add-stmt-syntax 'defun-close nil nil
10879 (c-most-enclosing-brace paren-state)
10880 paren-state))
10881 ))
10882
10883 ;; CASE 19: line is an expression, not a statement, and is directly
10884 ;; contained by a template delimiter. Most likely, we are in a
10885 ;; template arglist within a statement. This case is based on CASE
10886 ;; 7. At some point in the future, we may wish to create more
10887 ;; syntactic symbols such as `template-intro',
10888 ;; `template-cont-nonempty', etc., and distinguish between them as we
10889 ;; do for `arglist-intro' etc. (2009-12-07).
10890 ((and c-recognize-<>-arglists
10891 (setq containing-< (c-up-list-backward indent-point containing-sexp))
10892 (eq (char-after containing-<) ?\<))
10893 (setq placeholder (c-point 'boi containing-<))
10894 (goto-char containing-sexp) ; Most nested Lbrace/Lparen (but not
10895 ; '<') before indent-point.
10896 (if (>= (point) placeholder)
10897 (progn
10898 (forward-char)
10899 (skip-chars-forward " \t"))
10900 (goto-char placeholder))
10901 (c-add-stmt-syntax 'template-args-cont (list containing-<) t
10902 (c-most-enclosing-brace c-state-cache (point))
10903 paren-state))
10904
10905 ;; CASE 17: Statement or defun catchall.
10906 (t
10907 (goto-char indent-point)
10908 ;; Back up statements until we find one that starts at boi.
10909 (while (let* ((prev-point (point))
10910 (last-step-type (c-beginning-of-statement-1
10911 containing-sexp)))
10912 (if (= (point) prev-point)
10913 (progn
10914 (setq step-type (or step-type last-step-type))
10915 nil)
10916 (setq step-type last-step-type)
10917 (/= (point) (c-point 'boi)))))
10918 (cond
10919
10920 ;; CASE 17B: continued statement
10921 ((and (eq step-type 'same)
10922 (/= (point) indent-point))
10923 (c-add-stmt-syntax 'statement-cont nil nil
10924 containing-sexp paren-state))
10925
10926 ;; CASE 17A: After a case/default label?
10927 ((progn
10928 (while (and (eq step-type 'label)
10929 (not (looking-at c-label-kwds-regexp)))
10930 (setq step-type
10931 (c-beginning-of-statement-1 containing-sexp)))
10932 (eq step-type 'label))
10933 (c-add-stmt-syntax (if (eq char-after-ip ?{)
10934 'statement-case-open
10935 'statement-case-intro)
10936 nil t containing-sexp paren-state))
10937
10938 ;; CASE 17D: any old statement
10939 ((progn
10940 (while (eq step-type 'label)
10941 (setq step-type
10942 (c-beginning-of-statement-1 containing-sexp)))
10943 (eq step-type 'previous))
10944 (c-add-stmt-syntax 'statement nil t
10945 containing-sexp paren-state)
10946 (if (eq char-after-ip ?{)
10947 (c-add-syntax 'block-open)))
10948
10949 ;; CASE 17I: Inside a substatement block.
10950 ((progn
10951 ;; The following tests are all based on containing-sexp.
10952 (goto-char containing-sexp)
10953 ;; From here on we have the next containing sexp in lim.
10954 (setq lim (c-most-enclosing-brace paren-state containing-sexp))
10955 (c-after-conditional))
10956 (c-backward-to-block-anchor lim)
10957 (c-add-stmt-syntax 'statement-block-intro nil t
10958 lim paren-state)
10959 (if (eq char-after-ip ?{)
10960 (c-add-syntax 'block-open)))
10961
10962 ;; CASE 17E: first statement in an in-expression block.
10963 ;; C.f. cases 4, 7B and 16A.
10964 ((setq placeholder (c-looking-at-inexpr-block
10965 (c-safe-position containing-sexp paren-state)
10966 nil))
10967 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10968 'defun-block-intro
10969 'statement-block-intro))
10970 (back-to-indentation)
10971 (if (= containing-sexp (point))
10972 (c-add-syntax tmpsymbol (point))
10973 (goto-char (cdr placeholder))
10974 (back-to-indentation)
10975 (c-add-stmt-syntax tmpsymbol nil t
10976 (c-most-enclosing-brace c-state-cache (point))
10977 paren-state)
10978 (if (/= (point) (cdr placeholder))
10979 (c-add-syntax (car placeholder))))
10980 (if (eq char-after-ip ?{)
10981 (c-add-syntax 'block-open)))
10982
10983 ;; CASE 17F: first statement in an inline, or first
10984 ;; statement in a top-level defun. we can tell this is it
10985 ;; if there are no enclosing braces that haven't been
10986 ;; narrowed out by a class (i.e. don't use bod here).
10987 ((save-excursion
10988 (or (not (setq placeholder (c-most-enclosing-brace
10989 paren-state)))
10990 (and (progn
10991 (goto-char placeholder)
10992 (eq (char-after) ?{))
10993 (c-looking-at-decl-block (c-most-enclosing-brace
10994 paren-state (point))
10995 nil))))
10996 (c-backward-to-decl-anchor lim)
10997 (back-to-indentation)
10998 (c-add-syntax 'defun-block-intro (point)))
10999
11000 ;; CASE 17G: First statement in a function declared inside
11001 ;; a normal block. This can occur in Pike and with
11002 ;; e.g. the gcc extensions, but watch out for macros
11003 ;; followed by blocks. C.f. cases B.3 and 16F.
11004 ((save-excursion
11005 (and (not (c-at-statement-start-p))
11006 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
11007 (setq placeholder (point))
11008 (let ((c-recognize-typeless-decls nil))
11009 ;; Turn off recognition of constructs that lacks
11010 ;; a type in this case, since that's more likely
11011 ;; to be a macro followed by a block.
11012 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
11013 (back-to-indentation)
11014 (if (/= (point) containing-sexp)
11015 (goto-char placeholder))
11016 (c-add-stmt-syntax 'defun-block-intro nil t
11017 lim paren-state))
11018
11019 ;; CASE 17H: First statement in a block.
11020 (t
11021 ;; If the block is preceded by a case/switch label on the
11022 ;; same line, we anchor at the first preceding label at
11023 ;; boi. The default handling in c-add-stmt-syntax is
11024 ;; really fixes it better, but we do like this to keep the
11025 ;; indentation compatible with version 5.28 and earlier.
11026 ;; C.f. case 16C.
11027 (while (and (/= (setq placeholder (point)) (c-point 'boi))
11028 (eq (c-beginning-of-statement-1 lim) 'label)))
11029 (goto-char placeholder)
11030 (if (looking-at c-label-kwds-regexp)
11031 (c-add-syntax 'statement-block-intro (point))
11032 (goto-char containing-sexp)
11033 ;; c-backward-to-block-anchor not necessary here; those
11034 ;; situations are handled in case 17I above.
11035 (c-add-stmt-syntax 'statement-block-intro nil t
11036 lim paren-state))
11037 (if (eq char-after-ip ?{)
11038 (c-add-syntax 'block-open)))
11039 ))
11040 )
11041
11042 ;; now we need to look at any modifiers
11043 (goto-char indent-point)
11044 (skip-chars-forward " \t")
11045
11046 ;; are we looking at a comment only line?
11047 (when (and (looking-at c-comment-start-regexp)
11048 (/= (c-forward-token-2 0 nil (c-point 'eol)) 0))
11049 (c-append-syntax 'comment-intro))
11050
11051 ;; we might want to give additional offset to friends (in C++).
11052 (when (and c-opt-friend-key
11053 (looking-at c-opt-friend-key))
11054 (c-append-syntax 'friend))
11055
11056 ;; Set syntactic-relpos.
11057 (let ((p c-syntactic-context))
11058 (while (and p
11059 (if (integerp (c-langelem-pos (car p)))
11060 (progn
11061 (setq syntactic-relpos (c-langelem-pos (car p)))
11062 nil)
11063 t))
11064 (setq p (cdr p))))
11065
11066 ;; Start of or a continuation of a preprocessor directive?
11067 (if (and macro-start
11068 (eq macro-start (c-point 'boi))
11069 (not (and (c-major-mode-is 'pike-mode)
11070 (eq (char-after (1+ macro-start)) ?\"))))
11071 (c-append-syntax 'cpp-macro)
11072 (when (and c-syntactic-indentation-in-macros macro-start)
11073 (if in-macro-expr
11074 (when (or
11075 (< syntactic-relpos macro-start)
11076 (not (or
11077 (assq 'arglist-intro c-syntactic-context)
11078 (assq 'arglist-cont c-syntactic-context)
11079 (assq 'arglist-cont-nonempty c-syntactic-context)
11080 (assq 'arglist-close c-syntactic-context))))
11081 ;; If inside a cpp expression, i.e. anywhere in a
11082 ;; cpp directive except a #define body, we only let
11083 ;; through the syntactic analysis that is internal
11084 ;; in the expression. That means the arglist
11085 ;; elements, if they are anchored inside the cpp
11086 ;; expression.
11087 (setq c-syntactic-context nil)
11088 (c-add-syntax 'cpp-macro-cont macro-start))
11089 (when (and (eq macro-start syntactic-relpos)
11090 (not (assq 'cpp-define-intro c-syntactic-context))
11091 (save-excursion
11092 (goto-char macro-start)
11093 (or (not (c-forward-to-cpp-define-body))
11094 (<= (point) (c-point 'boi indent-point)))))
11095 ;; Inside a #define body and the syntactic analysis is
11096 ;; anchored on the start of the #define. In this case
11097 ;; we add cpp-define-intro to get the extra
11098 ;; indentation of the #define body.
11099 (c-add-syntax 'cpp-define-intro)))))
11100
11101 ;; return the syntax
11102 c-syntactic-context)))
11103
11104 \f
11105 ;; Indentation calculation.
11106
11107 (defun c-evaluate-offset (offset langelem symbol)
11108 ;; offset can be a number, a function, a variable, a list, or one of
11109 ;; the symbols + or -
11110 ;;
11111 ;; This function might do hidden buffer changes.
11112 (let ((res
11113 (cond
11114 ((numberp offset) offset)
11115 ((vectorp offset) offset)
11116 ((null offset) nil)
11117
11118 ((eq offset '+) c-basic-offset)
11119 ((eq offset '-) (- c-basic-offset))
11120 ((eq offset '++) (* 2 c-basic-offset))
11121 ((eq offset '--) (* 2 (- c-basic-offset)))
11122 ((eq offset '*) (/ c-basic-offset 2))
11123 ((eq offset '/) (/ (- c-basic-offset) 2))
11124
11125 ((functionp offset)
11126 (c-evaluate-offset
11127 (funcall offset
11128 (cons (c-langelem-sym langelem)
11129 (c-langelem-pos langelem)))
11130 langelem symbol))
11131
11132 ((listp offset)
11133 (cond
11134 ((eq (car offset) 'quote)
11135 (c-benign-error "The offset %S for %s was mistakenly quoted"
11136 offset symbol)
11137 nil)
11138
11139 ((memq (car offset) '(min max))
11140 (let (res val (method (car offset)))
11141 (setq offset (cdr offset))
11142 (while offset
11143 (setq val (c-evaluate-offset (car offset) langelem symbol))
11144 (cond
11145 ((not val))
11146 ((not res)
11147 (setq res val))
11148 ((integerp val)
11149 (if (vectorp res)
11150 (c-benign-error "\
11151 Error evaluating offset %S for %s: \
11152 Cannot combine absolute offset %S with relative %S in `%s' method"
11153 (car offset) symbol res val method)
11154 (setq res (funcall method res val))))
11155 (t
11156 (if (integerp res)
11157 (c-benign-error "\
11158 Error evaluating offset %S for %s: \
11159 Cannot combine relative offset %S with absolute %S in `%s' method"
11160 (car offset) symbol res val method)
11161 (setq res (vector (funcall method (aref res 0)
11162 (aref val 0)))))))
11163 (setq offset (cdr offset)))
11164 res))
11165
11166 ((eq (car offset) 'add)
11167 (let (res val)
11168 (setq offset (cdr offset))
11169 (while offset
11170 (setq val (c-evaluate-offset (car offset) langelem symbol))
11171 (cond
11172 ((not val))
11173 ((not res)
11174 (setq res val))
11175 ((integerp val)
11176 (if (vectorp res)
11177 (setq res (vector (+ (aref res 0) val)))
11178 (setq res (+ res val))))
11179 (t
11180 (if (vectorp res)
11181 (c-benign-error "\
11182 Error evaluating offset %S for %s: \
11183 Cannot combine absolute offsets %S and %S in `add' method"
11184 (car offset) symbol res val)
11185 (setq res val)))) ; Override.
11186 (setq offset (cdr offset)))
11187 res))
11188
11189 (t
11190 (let (res)
11191 (when (eq (car offset) 'first)
11192 (setq offset (cdr offset)))
11193 (while (and (not res) offset)
11194 (setq res (c-evaluate-offset (car offset) langelem symbol)
11195 offset (cdr offset)))
11196 res))))
11197
11198 ((and (symbolp offset) (boundp offset))
11199 (symbol-value offset))
11200
11201 (t
11202 (c-benign-error "Unknown offset format %S for %s" offset symbol)
11203 nil))))
11204
11205 (if (or (null res) (integerp res)
11206 (and (vectorp res) (= (length res) 1) (integerp (aref res 0))))
11207 res
11208 (c-benign-error "Error evaluating offset %S for %s: Got invalid value %S"
11209 offset symbol res)
11210 nil)))
11211
11212 (defun c-calc-offset (langelem)
11213 ;; Get offset from LANGELEM which is a list beginning with the
11214 ;; syntactic symbol and followed by any analysis data it provides.
11215 ;; That data may be zero or more elements, but if at least one is
11216 ;; given then the first is the anchor position (or nil). The symbol
11217 ;; is matched against `c-offsets-alist' and the offset calculated
11218 ;; from that is returned.
11219 ;;
11220 ;; This function might do hidden buffer changes.
11221 (let* ((symbol (c-langelem-sym langelem))
11222 (match (assq symbol c-offsets-alist))
11223 (offset (cdr-safe match)))
11224 (if match
11225 (setq offset (c-evaluate-offset offset langelem symbol))
11226 (if c-strict-syntax-p
11227 (c-benign-error "No offset found for syntactic symbol %s" symbol))
11228 (setq offset 0))
11229 (if (vectorp offset)
11230 offset
11231 (or (and (numberp offset) offset)
11232 (and (symbolp offset) (symbol-value offset))
11233 0))
11234 ))
11235
11236 (defun c-get-offset (langelem)
11237 ;; This is a compatibility wrapper for `c-calc-offset' in case
11238 ;; someone is calling it directly. It takes an old style syntactic
11239 ;; element on the form (SYMBOL . ANCHOR-POS) and converts it to the
11240 ;; new list form.
11241 ;;
11242 ;; This function might do hidden buffer changes.
11243 (if (c-langelem-pos langelem)
11244 (c-calc-offset (list (c-langelem-sym langelem)
11245 (c-langelem-pos langelem)))
11246 (c-calc-offset langelem)))
11247
11248 (defun c-get-syntactic-indentation (langelems)
11249 ;; Calculate the syntactic indentation from a syntactic description
11250 ;; as returned by `c-guess-syntax'.
11251 ;;
11252 ;; Note that topmost-intro always has an anchor position at bol, for
11253 ;; historical reasons. It's often used together with other symbols
11254 ;; that has more sane positions. Since we always use the first
11255 ;; found anchor position, we rely on that these other symbols always
11256 ;; precede topmost-intro in the LANGELEMS list.
11257 ;;
11258 ;; This function might do hidden buffer changes.
11259 (let ((indent 0) anchor)
11260
11261 (while langelems
11262 (let* ((c-syntactic-element (car langelems))
11263 (res (c-calc-offset c-syntactic-element)))
11264
11265 (if (vectorp res)
11266 ;; Got an absolute column that overrides any indentation
11267 ;; we've collected so far, but not the relative
11268 ;; indentation we might get for the nested structures
11269 ;; further down the langelems list.
11270 (setq indent (elt res 0)
11271 anchor (point-min)) ; A position at column 0.
11272
11273 ;; Got a relative change of the current calculated
11274 ;; indentation.
11275 (setq indent (+ indent res))
11276
11277 ;; Use the anchor position from the first syntactic
11278 ;; element with one.
11279 (unless anchor
11280 (setq anchor (c-langelem-pos (car langelems)))))
11281
11282 (setq langelems (cdr langelems))))
11283
11284 (if anchor
11285 (+ indent (save-excursion
11286 (goto-char anchor)
11287 (current-column)))
11288 indent)))
11289
11290 \f
11291 (cc-provide 'cc-engine)
11292
11293 ;; Local Variables:
11294 ;; indent-tabs-mode: t
11295 ;; tab-width: 8
11296 ;; End:
11297 ;;; cc-engine.el ends here