1 ;;; cc-langs.el --- language specific settings for CC Mode -*- coding: utf-8 -*-
3 ;; Copyright (C) 1985, 1987, 1992-2016 Free Software Foundation, Inc.
5 ;; Authors: 2002- Alan Mackenzie
6 ;; 1998- Martin Stjernholm
7 ;; 1992-1999 Barry A. Warsaw
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
16 ;; This file is part of GNU Emacs.
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.
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.
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/>.
33 ;; HACKERS NOTE: There's heavy macro magic here. If you need to make
34 ;; changes in this or other files containing `c-lang-defconst' but
35 ;; don't want to read through the longer discussion below then read
38 ;; o A change in a `c-lang-defconst' or `c-lang-defvar' will not take
39 ;; effect if the file containing the mode init function (typically
40 ;; cc-mode.el) is byte compiled.
41 ;; o To make changes show in font locking you need to reevaluate the
42 ;; `*-font-lock-keywords-*' constants, which normally is easiest to
43 ;; do with M-x eval-buffer in cc-fonts.el.
44 ;; o In either case it's necessary to reinitialize the mode to make
45 ;; the changes show in an existing buffer.
47 ;;; Introduction to the language dependent variable system:
49 ;; This file contains all the language dependent variables, except
50 ;; those specific for font locking which reside in cc-fonts.el. As
51 ;; far as possible, all the differences between the languages that CC
52 ;; Mode supports are described with these variables only, so that the
53 ;; code can be shared.
55 ;; The language constant system (see cc-defs.el) is used to specify
56 ;; various language dependent info at a high level, such as lists of
57 ;; keywords, and then from them generate - at compile time - the
58 ;; various regexps and other low-level structures actually employed in
59 ;; the code at runtime.
61 ;; This system is also designed to make it easy for developers of
62 ;; derived modes to customize the source constants for new language
63 ;; variants, without having to keep up with the exact regexps etc that
64 ;; are used in each CC Mode version. It's possible from an external
65 ;; package to add a new language by inheriting an existing one, and
66 ;; then change specific constants as necessary for the new language.
67 ;; The old values for those constants (and the values of all the other
68 ;; high-level constants) may be used to build the new ones, and those
69 ;; new values will in turn be used by the low-level definitions here
70 ;; to build the runtime constants appropriately for the new language
71 ;; in the current version of CC Mode.
73 ;; Like elsewhere in CC Mode, the existence of a doc string signifies
74 ;; that a language constant is part of the external API, and that it
75 ;; therefore can be used with a high confidence that it will continue
76 ;; to work with future versions of CC Mode. Even so, it's not
77 ;; unlikely that such constants will change meaning slightly as this
78 ;; system is refined further; a certain degree of dependence on the CC
79 ;; Mode version is unavoidable when hooking in at this level. Also
80 ;; note that there's still work to be done to actually use these
81 ;; constants everywhere inside CC Mode; there are still hardcoded
82 ;; values in many places in the code.
84 ;; Separate packages will also benefit from the compile time
85 ;; evaluation; the byte compiled file(s) for them will contain the
86 ;; compiled runtime constants ready for use by (the byte compiled) CC
87 ;; Mode, and the source definitions in this file don't have to be
88 ;; loaded then. However, if a byte compiled package is loaded that
89 ;; has been compiled with a different version of CC Mode than the one
90 ;; currently loaded, then the compiled-in values will be discarded and
91 ;; new ones will be built when the mode is initialized. That will
92 ;; automatically trig a load of the file(s) containing the source
93 ;; definitions (i.e. this file and/or cc-fonts.el) if necessary.
95 ;; A small example of a derived mode is available at
96 ;; <http://cc-mode.sourceforge.net/derived-mode-ex.el>. It also
97 ;; contains some useful hints for derived mode developers.
99 ;;; Using language variables:
101 ;; The `c-lang-defvar' forms in this file comprise the language
102 ;; variables that CC Mode uses. It does not work to use
103 ;; `c-lang-defvar' anywhere else (which isn't much of a limitation
104 ;; since these variables sole purpose is to interface with the CC Mode
105 ;; core functions). The values in these `c-lang-defvar's are not
106 ;; evaluated right away but instead collected to a single large `setq'
107 ;; that can be inserted for a particular language with the
108 ;; `c-init-language-vars' macro.
110 ;; This file is only required at compile time, or when not running
111 ;; from byte compiled files, or when the source definitions for the
112 ;; language constants are requested.
118 (unless (fboundp 'declare-function) (defmacro declare-function (&rest r))))
122 (if (and (boundp 'byte-compile-dest-file)
123 (stringp byte-compile-dest-file))
124 (cons (file-name-directory byte-compile-dest-file) load-path)
126 (load "cc-bytecomp" nil t)))
128 (cc-require 'cc-defs)
129 (cc-require 'cc-vars)
132 ;; This file is not always loaded. See note above.
133 (cc-external-require (if (eq c--mapcan-status 'cl-mapcan) 'cl-lib 'cl))
136 ;;; Setup for the `c-lang-defvar' system.
139 ;; These are used to collect the init forms from the subsequent
140 ;; `c-lang-defvar' and `c-lang-setvar'. They are used to build the
141 ;; lambda in `c-make-init-lang-vars-fun' below, and to build `defvar's
142 ;; and `make-variable-buffer-local's in cc-engine and
143 ;; `make-local-variable's in `c-init-language-vars-for'.
144 (defvar c-lang-variable-inits nil)
145 (defvar c-lang-variable-inits-tail nil)
146 (setq c-lang-variable-inits (list nil)
147 c-lang-variable-inits-tail c-lang-variable-inits)
148 (defvar c-emacs-variable-inits nil)
149 (defvar c-emacs-variable-inits-tail nil)
150 (setq c-emacs-variable-inits (list nil)
151 c-emacs-variable-inits-tail c-emacs-variable-inits))
153 (defmacro c-lang-defvar (var val &optional doc)
154 "Declares the buffer local variable VAR to get the value VAL. VAL is
155 evaluated and assigned at mode initialization. More precisely, VAL is
156 evaluated and bound to VAR when the result from the macro
157 `c-init-language-vars' is evaluated.
159 `c-lang-const' is typically used in VAL to get the right value for the
160 language being initialized, and such calls will be macro expanded to
161 the evaluated constant value at compile time."
164 (eq (car-safe val) 'c-lang-const)
167 ;; Special case: If there's no docstring and the value is a
168 ;; simple (c-lang-const foo) where foo is the same name as VAR
169 ;; then take the docstring from the language constant foo.
170 (setq doc (get (intern (symbol-name (nth 1 val)) c-lang-constants)
171 'variable-documentation)))
175 (let ((elem (assq var (cdr c-lang-variable-inits))))
177 (setcdr elem (list val doc))
178 (setcdr c-lang-variable-inits-tail (list (list var val doc)))
179 (setq c-lang-variable-inits-tail (cdr c-lang-variable-inits-tail))))
181 ;; Return the symbol, like the other def* forms.
184 (defmacro c-lang-setvar (var val)
185 "Causes the variable VAR to be made buffer local and to get set to the
186 value VAL. VAL is evaluated and assigned at mode initialization. More
187 precisely, VAL is evaluated and bound to VAR when the result from the
188 macro `c-init-language-vars' is evaluated. VAR is typically a standard
189 Emacs variable like `comment-start'.
191 `c-lang-const' is typically used in VAL to get the right value for the
192 language being initialized, and such calls will be macro expanded to
193 the evaluated constant value at compile time."
194 (let ((elem (assq var (cdr c-emacs-variable-inits))))
196 (setcdr elem (list val)) ; Maybe remove "list", sometime. 2006-07-19
197 (setcdr c-emacs-variable-inits-tail (list (list var val)))
198 (setq c-emacs-variable-inits-tail (cdr c-emacs-variable-inits-tail))))
200 ;; Return the symbol, like the other def* forms.
203 (put 'c-lang-defvar 'lisp-indent-function 'defun)
204 ; (eval-after-load "edebug" ; 2006-07-09: def-edebug-spec is now in subr.el.
206 (def-edebug-spec c-lang-defvar
207 (&define name def-form &optional stringp)) ;)
209 ;; Suppress "might not be defined at runtime" warning.
210 ;; This file is only used when compiling other cc files.
211 ;; These are defined in cl as aliases to the cl- versions.
212 ;(declare-function delete-duplicates "cl-seq" (cl-seq &rest cl-keys) t)
213 ;(declare-function mapcan "cl-extra" (cl-func cl-seq &rest cl-rest) t)
216 ;; Some helper functions used when building the language constants.
218 (defun c-filter-ops (ops opgroup-filter op-filter &optional xlate)
219 ;; Extract a subset of the operators in the list OPS in a DWIM:ey
220 ;; way. The return value is a plain list of operators:
222 ;; OPS either has the structure of `c-operators', is a single
223 ;; group in `c-operators', or is a plain list of operators.
225 ;; OPGROUP-FILTER specifies how to select the operator groups. It
226 ;; can be t to choose all groups, a list of group type symbols
227 ;; (such as 'prefix) to accept, or a function which will be called
228 ;; with the group symbol for each group and should return non-nil
229 ;; if that group is to be included.
231 ;; OP-FILTER selects the operators. It is either t to select all
232 ;; operators, a string to select all operators for which `string-match'
233 ;; matches the operator with the string, or a function which will be
234 ;; called with the operator and should return non-nil when the operator
235 ;; is to be selected.
237 ;; If XLATE is given, it's a function which is called for each
238 ;; matching operator and its return value is collected instead.
239 ;; If it returns a list, the elements are spliced directly into
240 ;; the final result, which is returned as a list with duplicates
241 ;; removed using `equal'.
243 ;; `c-mode-syntax-table' for the current mode is in effect during
244 ;; the whole procedure.
245 (unless (listp (car-safe ops))
246 (setq ops (list ops)))
247 (cond ((eq opgroup-filter t)
248 (setq opgroup-filter (lambda (opgroup) t)))
249 ((not (functionp opgroup-filter))
250 (setq opgroup-filter `(lambda (opgroup)
251 (memq opgroup ',opgroup-filter)))))
252 (cond ((eq op-filter t)
253 (setq op-filter (lambda (op) t)))
255 (setq op-filter `(lambda (op)
256 (string-match ,op-filter op)))))
258 (setq xlate 'identity))
259 (c-with-syntax-table (c-lang-const c-mode-syntax-table)
260 (c--delete-duplicates
261 (c--mapcan (lambda (opgroup)
262 (when (if (symbolp (car opgroup))
263 (when (funcall opgroup-filter (car opgroup))
264 (setq opgroup (cdr opgroup))
267 (c--mapcan (lambda (op)
268 (when (funcall op-filter op)
269 (let ((res (funcall xlate op)))
270 (if (listp res) res (list res)))))
276 ;;; Various mode specific values that aren't language related.
278 (c-lang-defconst c-mode-menu
279 ;; The definition for the mode menu. The menu title is prepended to
280 ;; this before it's fed to `easy-menu-define'.
281 t `(["Comment Out Region" comment-region
282 (c-fn-region-is-active-p)]
283 ["Uncomment Region" (comment-region (region-beginning)
285 (c-fn-region-is-active-p)]
286 ["Indent Expression" c-indent-exp
287 (memq (char-after) '(?\( ?\[ ?\{))]
288 ["Indent Line or Region" c-indent-line-or-region t]
289 ["Fill Comment Paragraph" c-fill-paragraph t]
291 ["Backward Statement" c-beginning-of-statement t]
292 ["Forward Statement" c-end-of-statement t]
293 ,@(when (c-lang-const c-opt-cpp-prefix)
294 ;; Only applicable if there's a cpp preprocessor.
295 `(["Up Conditional" c-up-conditional t]
296 ["Backward Conditional" c-backward-conditional t]
297 ["Forward Conditional" c-forward-conditional t]
299 ["Macro Expand Region" c-macro-expand
300 (c-fn-region-is-active-p)]
301 ["Backslashify" c-backslash-region
302 (c-fn-region-is-active-p)]))
305 ["Set Style..." c-set-style t]
306 ["Show Current Style Name" (message
310 ["Guess Style from this Buffer" c-guess-buffer-no-install t]
311 ["Install the Last Guessed Style..." c-guess-install
312 (and c-guess-guessed-offsets-alist
313 c-guess-guessed-basic-offset) ]
314 ["View the Last Guessed Style" c-guess-view
315 (and c-guess-guessed-offsets-alist
316 c-guess-guessed-basic-offset) ])
319 ["Syntactic indentation" c-toggle-syntactic-indentation
320 :style toggle :selected c-syntactic-indentation]
321 ["Electric mode" c-toggle-electric-state
322 :style toggle :selected c-electric-flag]
323 ["Auto newline" c-toggle-auto-newline
324 :style toggle :selected c-auto-newline]
325 ["Hungry delete" c-toggle-hungry-state
326 :style toggle :selected c-hungry-delete-key]
327 ["Subword mode" c-subword-mode
328 :style toggle :selected (and (boundp 'c-subword-mode)
334 (defun c-populate-syntax-table (table)
335 "Populate the given syntax table as necessary for a C-like language.
336 This includes setting \\=' and \" as string delimiters, and setting up
337 the comment syntax to handle both line style \"//\" and block style
338 \"/*\" \"*/\" comments."
340 (modify-syntax-entry ?_ "_" table)
341 (modify-syntax-entry ?\\ "\\" table)
342 (modify-syntax-entry ?+ "." table)
343 (modify-syntax-entry ?- "." table)
344 (modify-syntax-entry ?= "." table)
345 (modify-syntax-entry ?% "." table)
346 (modify-syntax-entry ?< "." table)
347 (modify-syntax-entry ?> "." table)
348 (modify-syntax-entry ?& "." table)
349 (modify-syntax-entry ?| "." table)
350 (modify-syntax-entry ?\' "\"" table)
351 (modify-syntax-entry ?\240 "." table)
353 ;; Set up block and line oriented comments. The new C
354 ;; standard mandates both comment styles even in C, so since
355 ;; all languages now require dual comments, we make this the
359 ((memq '8-bit c-emacs-features)
360 (modify-syntax-entry ?/ ". 1456" table)
361 (modify-syntax-entry ?* ". 23" table))
363 ((memq '1-bit c-emacs-features)
364 (modify-syntax-entry ?/ ". 124b" table)
365 (modify-syntax-entry ?* ". 23" table))
367 (t (error "CC Mode is incompatible with this version of Emacs")))
369 (modify-syntax-entry ?\n "> b" table)
370 ;; Give CR the same syntax as newline, for selective-display
371 (modify-syntax-entry ?\^m "> b" table))
373 (c-lang-defconst c-make-mode-syntax-table
374 "Functions that generates the mode specific syntax tables.
375 The syntax tables aren't stored directly since they're quite large."
377 (let ((table (make-syntax-table)))
378 (c-populate-syntax-table table)
379 ;; Mode specific syntaxes.
380 ,(cond ((or (c-major-mode-is 'objc-mode) (c-major-mode-is 'java-mode))
381 ;; Let '@' be part of symbols in ObjC to cope with
382 ;; its compiler directives as single keyword tokens.
383 ;; This is then necessary since it's assumed that
384 ;; every keyword is a single symbol.
385 `(modify-syntax-entry ?@ "_" table))
386 ((c-major-mode-is 'pike-mode)
387 `(modify-syntax-entry ?@ "." table)))
390 (c-lang-defconst c-mode-syntax-table
391 ;; The syntax tables in evaluated form. Only used temporarily when
392 ;; the constants in this file are evaluated.
393 t (funcall (c-lang-const c-make-mode-syntax-table)))
395 (c-lang-defconst c++-make-template-syntax-table
396 ;; A variant of `c++-mode-syntax-table' that defines `<' and `>' as
397 ;; parenthesis characters. Used temporarily when template argument
398 ;; lists are parsed. Note that this encourages incorrect parsing of
399 ;; templates since they might contain normal operators that uses the
400 ;; '<' and '>' characters. Therefore this syntax table might go
401 ;; away when CC Mode handles templates correctly everywhere. WHILE
402 ;; THIS SYNTAX TABLE IS CURRENT, `c-parse-state' MUST _NOT_ BE
405 (java c++) `(lambda ()
406 (let ((table (funcall ,(c-lang-const c-make-mode-syntax-table))))
407 (modify-syntax-entry ?< "(>" table)
408 (modify-syntax-entry ?> ")<" table)
410 (c-lang-defvar c++-template-syntax-table
411 (and (c-lang-const c++-make-template-syntax-table)
412 ;; The next eval remove a superfluous ' from '(lambda. This
413 ;; gets rid of compilation warnings.
414 (funcall (eval (c-lang-const c++-make-template-syntax-table)))))
416 (c-lang-defconst c-make-no-parens-syntax-table
417 ;; A variant of the standard syntax table which is used to find matching
418 ;; "<"s and ">"s which have been marked as parens using syntax table
419 ;; properties. The other paren characters (e.g. "{", ")" "]") are given a
420 ;; non-paren syntax here. so that the list commands will work on "< ... >"
421 ;; even when there's unbalanced other parens inside them.
423 ;; This variable is nil for languages which don't have template stuff.
424 t (if (c-lang-const c-recognize-<>-arglists)
426 ;(if (c-lang-const c-recognize-<>-arglists)
427 (let ((table (funcall ,(c-lang-const c-make-mode-syntax-table))))
428 (modify-syntax-entry ?\( "." table)
429 (modify-syntax-entry ?\) "." table)
430 (modify-syntax-entry ?\[ "." table)
431 (modify-syntax-entry ?\] "." table)
432 (modify-syntax-entry ?\{ "." table)
433 (modify-syntax-entry ?\} "." table)
435 (c-lang-defvar c-no-parens-syntax-table
436 (and (c-lang-const c-make-no-parens-syntax-table)
437 ;; See comment in `c++template-syntax-table' about the next `eval'.
438 (funcall (eval (c-lang-const c-make-no-parens-syntax-table)))))
440 (c-lang-defconst c-identifier-syntax-modifications
441 "A list that describes the modifications that should be done to the
442 mode syntax table to get a syntax table that matches all identifiers
443 and keywords as words.
445 The list is just like the one used in `font-lock-defaults': Each
446 element is a cons where the car is the character to modify and the cdr
447 the new syntax, as accepted by `modify-syntax-entry'."
448 ;; The $ character is not allowed in most languages (one exception
449 ;; is Java which allows it for legacy reasons) but we still classify
450 ;; it as an identifier character since it's often used in various
451 ;; machine generated identifiers.
452 t '((?_ . "w") (?$ . "w"))
453 (objc java) (append '((?@ . "w"))
454 (c-lang-const c-identifier-syntax-modifications))
456 (c-lang-defvar c-identifier-syntax-modifications
457 (c-lang-const c-identifier-syntax-modifications))
459 (c-lang-defvar c-identifier-syntax-table
460 (let ((table (copy-syntax-table (c-mode-var "mode-syntax-table")))
461 (mods c-identifier-syntax-modifications)
466 (modify-syntax-entry (car mod) (cdr mod) table))
468 "Syntax table built on the mode syntax table but additionally
469 classifies symbol constituents like `_' and `$' as word constituents,
470 so that all identifiers are recognized as words.")
472 (c-lang-defconst c-get-state-before-change-functions
473 ;; For documentation see the following c-lang-defvar of the same name.
474 ;; The value here may be a list of functions or a single function.
476 c++ '(c-extend-region-for-CPP
477 c-before-change-check-<>-operators
478 c-invalidate-macro-cache)
479 (c objc) '(c-extend-region-for-CPP c-invalidate-macro-cache)
480 ;; java 'c-before-change-check-<>-operators
481 awk 'c-awk-record-region-clear-NL)
482 (c-lang-defvar c-get-state-before-change-functions
483 (let ((fs (c-lang-const c-get-state-before-change-functions)))
487 "If non-nil, a list of functions called from c-before-change-hook.
488 Typically these will record enough state to allow
489 `c-before-font-lock-function' to extend the region to fontify,
490 and may do such things as removing text-properties which must be
493 These functions will be run in the order given. Each of them
494 takes 2 parameters, the BEG and END supplied to every
495 before-change function; on entry, the buffer will have been
496 widened and match-data will have been saved; point is undefined
497 on both entry and exit; the return value is ignored.
499 The functions are called even when font locking isn't enabled.
501 When the mode is initialized, the functions are called with
502 parameters \(point-min) and \(point-max).")
504 (c-lang-defconst c-before-font-lock-functions
505 ;; For documentation see the following c-lang-defvar of the same name.
506 ;; The value here may be a list of functions or a single function.
507 t 'c-change-expand-fl-region
508 (c objc) '(c-neutralize-syntax-in-and-mark-CPP
509 c-change-expand-fl-region)
510 c++ '(c-neutralize-syntax-in-and-mark-CPP
511 c-restore-<>-properties
512 c-change-expand-fl-region)
513 java '(c-restore-<>-properties
514 c-change-expand-fl-region)
515 awk 'c-awk-extend-and-syntax-tablify-region)
516 (c-lang-defvar c-before-font-lock-functions
517 (let ((fs (c-lang-const c-before-font-lock-functions)))
521 "If non-nil, a list of functions called just before font locking.
522 Typically they will extend the region about to be fontified \(see
523 below) and will set `syntax-table' text properties on the region.
525 These functions will be run in the order given. Each of them
526 takes 3 parameters, the BEG, END, and OLD-LEN supplied to every
527 after-change function; point is undefined on both entry and exit;
528 on entry, the buffer will have been widened and match-data will
529 have been saved; the return value is ignored.
531 The functions may extend the region to be fontified by setting the
532 buffer local variables c-new-BEG and c-new-END.
534 The functions are called even when font locking is disabled.
536 When the mode is initialized, these functions are called with
537 parameters \(point-min), \(point-max) and <buffer size>.")
539 (c-lang-defconst c-before-context-fontification-functions
540 t 'c-context-expand-fl-region
542 ;; For documentation see the following c-lang-defvar of the same name.
543 ;; The value here may be a list of functions or a single function.
544 (c-lang-defvar c-before-context-fontification-functions
545 (let ((fs (c-lang-const c-before-context-fontification-functions)))
549 "If non-nil, a list of functions called just before context (or
550 other non-change) fontification is done. Typically they will
553 These functions will be run in the order given. Each of them
554 takes 2 parameters, the BEG and END of the region to be
555 fontified. Point is undefined on both entry and exit. On entry,
556 the buffer will have been widened and match-data will have been
557 saved; the return value is a cons of the adjusted
558 region, (NEW-BEG . NEW-END).")
561 ;;; Syntactic analysis ("virtual semicolons") for line-oriented languages (AWK).
562 (c-lang-defconst c-at-vsemi-p-fn
563 "Contains a function \"Is there a virtual semicolon at POS or point?\".
564 Such a function takes one optional parameter, a buffer position (defaults to
565 point), and returns nil or t. This variable contains nil for languages which
566 don't have EOL terminated statements. "
568 (c c++ objc) 'c-at-macro-vsemi-p
569 awk 'c-awk-at-vsemi-p)
570 (c-lang-defvar c-at-vsemi-p-fn (c-lang-const c-at-vsemi-p-fn))
572 (c-lang-defconst c-vsemi-status-unknown-p-fn
573 "Contains a function \"are we unsure whether there is a virtual semicolon on this line?\".
574 The (admittedly kludgy) purpose of such a function is to prevent an infinite
575 recursion in c-beginning-of-statement-1 when point starts at a `while' token.
576 The function MUST NOT UNDER ANY CIRCUMSTANCES call c-beginning-of-statement-1,
577 even indirectly. This variable contains nil for languages which don't have
578 EOL terminated statements."
580 (c c++ objc) 'c-macro-vsemi-status-unknown-p
581 awk 'c-awk-vsemi-status-unknown-p)
582 (c-lang-defvar c-vsemi-status-unknown-p-fn
583 (c-lang-const c-vsemi-status-unknown-p-fn))
586 ;;; Lexer-level syntax (identifiers, tokens etc).
588 (c-lang-defconst c-has-bitfields
589 "Whether the language has bitfield declarations."
592 (c-lang-defvar c-has-bitfields (c-lang-const c-has-bitfields))
594 (c-lang-defconst c-modified-constant
595 "Regexp that matches a “modified” constant literal such as \"L\\='a\\='\",
596 a “long character”. In particular, this recognizes forms of constant
597 which `c-backward-sexp' needs to be called twice to move backwards over."
599 (c c++ objc) "L'\\([^\\'\t\f\n\r]\\|\\\\.\\)'")
600 ;; FIXME!!! Extend this to cover strings, if needed. 2008-04-11
601 (c-lang-defvar c-modified-constant (c-lang-const c-modified-constant))
603 (c-lang-defconst c-symbol-start
604 "Regexp that matches the start of a symbol, i.e. any identifier or
605 keyword. It's unspecified how far it matches. Does not contain a \\|
606 operator at the top level."
607 t (concat "[" c-alpha "_]")
608 java (concat "[" c-alpha "_@]")
609 objc (concat "[" c-alpha "_@]")
610 pike (concat "[" c-alpha "_`]"))
611 (c-lang-defvar c-symbol-start (c-lang-const c-symbol-start))
613 (c-lang-defconst c-symbol-chars
614 "Set of characters that can be part of a symbol.
615 This is of the form that fits inside [ ] in a regexp."
616 ;; Pike note: With the backquote identifiers this would include most
617 ;; operator chars too, but they are handled with other means instead.
618 t (concat c-alnum "_$")
619 objc (concat c-alnum "_$@"))
620 (c-lang-defvar c-symbol-chars (c-lang-const c-symbol-chars))
622 (c-lang-defconst c-symbol-char-key
623 "Regexp matching a sequence of at least one identifier character."
624 t (concat "[" (c-lang-const c-symbol-chars) "]+"))
625 (c-lang-defvar c-symbol-char-key (c-lang-const c-symbol-char-key))
627 (c-lang-defconst c-symbol-key
628 "Regexp matching identifiers and keywords (with submatch 0). Assumed
629 to match if `c-symbol-start' matches on the same position."
630 t (concat (c-lang-const c-symbol-start)
631 "[" (c-lang-const c-symbol-chars) "]*")
633 ;; Use the value from C here since the operator backquote is
634 ;; covered by the other alternative.
635 (c-lang-const c-symbol-key c)
637 (c-make-keywords-re nil
638 (c-lang-const c-overloadable-operators))))
639 (c-lang-defvar c-symbol-key (c-lang-const c-symbol-key))
641 (c-lang-defconst c-symbol-key-depth
642 ;; Number of regexp grouping parens in `c-symbol-key'.
643 t (regexp-opt-depth (c-lang-const c-symbol-key)))
645 (c-lang-defconst c-nonsymbol-chars
646 "This is the set of chars that can't be part of a symbol, i.e. the
647 negation of `c-symbol-chars'."
648 t (concat "^" (c-lang-const c-symbol-chars)))
649 (c-lang-defvar c-nonsymbol-chars (c-lang-const c-nonsymbol-chars))
651 (c-lang-defconst c-nonsymbol-key
652 "Regexp that matches any character that can't be part of a symbol.
653 It's usually appended to other regexps to avoid matching a prefix.
654 It's assumed to not contain any submatchers."
655 ;; The same thing regarding Unicode identifiers applies here as to
657 t (concat "[" (c-lang-const c-nonsymbol-chars) "]"))
659 (c-lang-defconst c-identifier-ops
660 "The operators that make up fully qualified identifiers. nil in
661 languages that don't have such things. See `c-operators' for a
662 description of the format. Binary operators can concatenate symbols,
663 e.g. \"::\" in \"A::B::C\". Prefix operators can precede identifiers,
664 e.g. \"~\" in \"~A::B\". Other types of operators aren't supported.
666 This value is by default merged into `c-operators'."
668 c++ '((prefix "~" "??-" "compl")
671 ;; Java has "." to concatenate identifiers but it's also used for
672 ;; normal indexing. There's special code in the Java font lock
673 ;; rules to fontify qualified identifiers based on the standard
674 ;; naming conventions. We still define "." here to make
675 ;; `c-forward-name' move over as long names as possible which is
676 ;; necessary to e.g. handle throws clauses correctly.
677 java '((left-assoc "."))
678 idl '((left-assoc "::")
680 pike '((left-assoc "::")
684 (c-lang-defconst c-opt-identifier-concat-key
685 ;; Appendable adorned regexp matching the operators that join
686 ;; symbols to fully qualified identifiers, or nil in languages that
687 ;; don't have such things.
689 ;; This was a docstring constant in 5.30. It still works but is now
690 ;; considered internal - change `c-identifier-ops' instead.
691 t (let ((ops (c-filter-ops (c-lang-const c-identifier-ops)
692 '(left-assoc right-assoc)
695 (c-make-keywords-re 'appendable ops))))
696 (c-lang-defvar c-opt-identifier-concat-key
697 (c-lang-const c-opt-identifier-concat-key)
700 (c-lang-defconst c-opt-identifier-concat-key-depth
701 ;; Number of regexp grouping parens in `c-opt-identifier-concat-key'.
702 t (regexp-opt-depth (c-lang-const c-opt-identifier-concat-key)))
704 (c-lang-defconst c-opt-identifier-prefix-key
705 ;; Appendable adorned regexp matching operators that might precede
706 ;; an identifier and that are part of the identifier in that case.
707 ;; nil in languages without such things.
708 t (let ((ops (c-filter-ops (c-lang-const c-identifier-ops)
712 (c-make-keywords-re 'appendable ops))))
714 (c-lang-defconst c-after-id-concat-ops
715 "Operators that can occur after a binary operator on `c-identifier-ops'
716 in identifiers. nil in languages that don't have such things.
718 Operators here should also have appropriate entries in `c-operators' -
719 it's not taken care of by default."
721 ;; '~' for destructors in C++, '*' for member pointers.
723 ;; In Java we recognize '*' to deal with "foo.bar.*" that can occur
724 ;; in import declarations. (This will also match bogus things like
725 ;; "foo.*bar" but we don't bother.)
728 (c-lang-defconst c-opt-after-id-concat-key
729 ;; Regexp that must match the token after
730 ;; `c-opt-identifier-concat-key' for it to be considered an
731 ;; identifier concatenation operator (which e.g. causes the
732 ;; preceding identifier to be fontified as a reference). Assumed to
733 ;; be a string if `c-opt-identifier-concat-key' is.
735 ;; This was a docstring constant in 5.30. It still works but is now
736 ;; considered internal - change `c-after-id-concat-ops' instead.
737 t (concat (c-lang-const c-symbol-start)
738 (if (c-lang-const c-after-id-concat-ops)
739 (concat "\\|" (c-make-keywords-re 'appendable
740 (c-lang-const c-after-id-concat-ops)))
743 (c-lang-defconst c-identifier-start
744 "Regexp that matches the start of an (optionally qualified) identifier.
745 It should also match all keywords. It's unspecified how far it
747 t (concat (c-lang-const c-symbol-start)
748 (if (c-lang-const c-opt-identifier-prefix-key)
750 (c-lang-const c-opt-identifier-prefix-key))
752 (c-lang-defvar c-identifier-start (c-lang-const c-identifier-start))
754 (c-lang-defconst c-identifier-key
755 "Regexp matching a fully qualified identifier, like \"A::B::c\" in
756 C++. It does not recognize the full range of syntactic whitespace
757 between the tokens; `c-forward-name' has to be used for that. It
758 should also not match identifiers containing parenthesis groupings,
759 e.g. identifiers with template arguments such as \"A<X,Y>\" in C++."
760 ;; This regexp is more complex than strictly necessary to ensure
761 ;; that it can be matched with a minimum of backtracking.
762 t (concat (if (c-lang-const c-opt-identifier-prefix-key)
765 (c-lang-const c-opt-identifier-prefix-key)
766 (c-lang-const c-simple-ws) "*"
769 "\\(" (c-lang-const c-symbol-key) "\\)"
770 (if (c-lang-const c-opt-identifier-concat-key)
773 (c-lang-const c-simple-ws) "*"
774 (c-lang-const c-opt-identifier-concat-key)
775 (c-lang-const c-simple-ws) "*"
776 (if (c-lang-const c-after-id-concat-ops)
779 (c-make-keywords-re 'appendable
780 (c-lang-const c-after-id-concat-ops))
782 ;; For flexibility, consider the symbol match
783 ;; optional if we've hit a
784 ;; `c-after-id-concat-ops' operator. This is
785 ;; also necessary to handle the "*" that can
786 ;; end import declaration identifiers in Java.
788 (c-lang-const c-simple-ws) "*"
789 "\\(" (c-lang-const c-symbol-key) "\\)"
792 "\\(" (c-lang-const c-symbol-key) "\\)"
794 (concat "\\(" (c-lang-const c-symbol-key) "\\)"))
797 (c-lang-defvar c-identifier-key (c-lang-const c-identifier-key))
799 (c-lang-defconst c-identifier-last-sym-match
800 ;; This was a docstring constant in 5.30 but it's no longer used.
801 ;; It's only kept to avoid breaking third party code.
803 ;; Used to identify the submatch in `c-identifier-key' that
804 ;; surrounds the last symbol in the qualified identifier. It's a
805 ;; list of submatch numbers, of which the first that has a match is
806 ;; taken. It's assumed that at least one does when the regexp has
810 (c-lang-defconst c-string-escaped-newlines
811 "Set if the language support backslash escaped newlines inside string
815 (c-lang-defvar c-string-escaped-newlines
816 (c-lang-const c-string-escaped-newlines))
818 (c-lang-defconst c-multiline-string-start-char
819 "Set if the language supports multiline string literals without escaped
820 newlines. If t, all string literals are multiline. If a character,
821 only literals where the open quote is immediately preceded by that
822 literal are multiline."
825 (c-lang-defvar c-multiline-string-start-char
826 (c-lang-const c-multiline-string-start-char))
828 (c-lang-defconst c-opt-cpp-symbol
829 "The symbol which starts preprocessor constructs when in the margin."
832 (c-lang-defvar c-opt-cpp-symbol (c-lang-const c-opt-cpp-symbol))
834 (c-lang-defconst c-opt-cpp-prefix
835 "Regexp matching the prefix of a cpp directive in the languages that
836 normally use that macro preprocessor. Tested at bol or at boi.
837 Assumed to not contain any submatches or \\| operators."
838 ;; TODO (ACM, 2005-04-01). Amend the following to recognize escaped NLs;
839 ;; amend all uses of c-opt-cpp-prefix which count regexp-depth.
842 (c-lang-defvar c-opt-cpp-prefix (c-lang-const c-opt-cpp-prefix))
844 (c-lang-defconst c-anchored-cpp-prefix
845 "Regexp matching the prefix of a cpp directive anchored to BOL,
846 in the languages that have a macro preprocessor."
847 t "^\\s *\\(#\\)\\s *"
849 (c-lang-defvar c-anchored-cpp-prefix (c-lang-const c-anchored-cpp-prefix))
851 (c-lang-defconst c-opt-cpp-start
852 "Regexp matching the prefix of a cpp directive including the directive
853 name, or nil in languages without preprocessor support. The first
854 submatch surrounds the directive name."
855 t (if (c-lang-const c-opt-cpp-prefix)
856 (concat (c-lang-const c-opt-cpp-prefix)
857 "\\([" c-alnum "]+\\)"))
858 ;; Pike, being a scripting language, recognizes hash-bangs too.
859 pike (concat (c-lang-const c-opt-cpp-prefix)
860 "\\([" c-alnum "]+\\|!\\)"))
861 (c-lang-defvar c-opt-cpp-start (c-lang-const c-opt-cpp-start))
863 (c-lang-defconst c-cpp-message-directives
864 "List of cpp directives (without the prefix) that are followed by a
866 t (if (c-lang-const c-opt-cpp-prefix)
868 (c c++ objc pike) '("error" "warning"))
870 (c-lang-defconst c-cpp-include-directives
871 "List of cpp directives (without the prefix) that are followed by a
872 file name in angle brackets or quotes."
873 t (if (c-lang-const c-opt-cpp-prefix)
875 objc '("include" "import"))
877 (c-lang-defconst c-opt-cpp-macro-define
878 "Cpp directive (without the prefix) that is followed by a macro
879 definition, or nil if the language doesn't have any."
880 t (if (c-lang-const c-opt-cpp-prefix)
882 (c-lang-defvar c-opt-cpp-macro-define
883 (c-lang-const c-opt-cpp-macro-define))
885 (c-lang-defconst c-opt-cpp-macro-define-start
886 ;; Regexp matching everything up to the macro body of a cpp define, or the
887 ;; end of the logical line if there is none. Submatch 1 is the name of the
888 ;; macro. Set if c-opt-cpp-macro-define is.
889 t (if (c-lang-const c-opt-cpp-macro-define)
890 (concat (c-lang-const c-opt-cpp-prefix)
891 (c-lang-const c-opt-cpp-macro-define)
892 "[ \t]+\\(\\(\\sw\\|_\\)+\\)\\(([^)]*)\\)?"
894 "\\([ \t]\\|\\\\\n\\)*")))
895 (c-lang-defvar c-opt-cpp-macro-define-start
896 (c-lang-const c-opt-cpp-macro-define-start))
898 (c-lang-defconst c-opt-cpp-macro-define-id
899 ;; Regexp matching everything up to the end of the identifier defined
901 t (if (c-lang-const c-opt-cpp-macro-define)
902 (concat (c-lang-const c-opt-cpp-prefix) ; #
903 (c-lang-const c-opt-cpp-macro-define) ; define
904 "[ \t]+\\(\\sw\\|_\\)+")))
905 (c-lang-defvar c-opt-cpp-macro-define-id
906 (c-lang-const c-opt-cpp-macro-define-id))
908 (c-lang-defconst c-cpp-expr-directives
909 "List of cpp directives (without the prefix) that are followed by an
911 t (if (c-lang-const c-opt-cpp-prefix)
914 (c-lang-defconst c-cpp-expr-intro-re
915 "Regexp which matches the start of a CPP directive which contains an
916 expression, or nil if there aren't any in the language."
917 t (if (c-lang-const c-cpp-expr-directives)
919 (c-lang-const c-opt-cpp-prefix)
920 (c-make-keywords-re t (c-lang-const c-cpp-expr-directives)))))
921 (c-lang-defvar c-cpp-expr-intro-re
922 (c-lang-const c-cpp-expr-intro-re))
924 (c-lang-defconst c-cpp-expr-functions
925 "List of functions in cpp expressions."
926 t (if (c-lang-const c-opt-cpp-prefix)
928 pike '("defined" "efun" "constant"))
930 (c-lang-defconst c-assignment-operators
931 "List of all assignment operators."
932 t '("=" "*=" "/=" "%=" "+=" "-=" ">>=" "<<=" "&=" "^=" "|=")
933 java (append (c-lang-const c-assignment-operators)
935 c++ (append (c-lang-const c-assignment-operators)
936 '("and_eq" "or_eq" "xor_eq" "??!=" "??'="))
939 (c-lang-defconst c-operators
940 "List describing all operators, along with their precedence and
941 associativity. The order in the list corresponds to the precedence of
942 the operators: The operators in each element are a group with the same
943 precedence, and the group has higher precedence than the groups in all
944 following elements. The car of each element describes the type of the
945 operator group, and the cdr is a list of the operator tokens in it.
946 The operator group types are:
948 `prefix' Unary prefix operators.
949 `postfix' Unary postfix operators.
951 Unary postfix operators if and only if the chars have
953 `left-assoc' Binary left associative operators (i.e. a+b+c means (a+b)+c).
954 `right-assoc' Binary right associative operators (i.e. a=b=c means a=(b=c)).
955 `right-assoc-sequence'
956 Right associative operator that constitutes of a
957 sequence of tokens that separate expressions. All the
958 tokens in the group are in this case taken as
959 describing the sequence in one such operator, and the
960 order between them is therefore significant.
962 Operators containing a character with paren syntax are taken to match
963 with a corresponding open/close paren somewhere else. A postfix
964 operator with close paren syntax is taken to end a postfix expression
965 started somewhere earlier, rather than start a new one at point. Vice
966 versa for prefix operators with open paren syntax.
968 Note that operators like \".\" and \"->\" which in language references
969 often are described as postfix operators are considered binary here,
970 since CC Mode treats every identifier as an expression."
972 ;; There's currently no code in CC Mode that exploits all the info
973 ;; in this variable; precedence, associativity etc are present as a
974 ;; preparation for future work.
976 ;; FIXME!!! C++11's "auto" operators "=" and "->" need to go in here
977 ;; somewhere. 2012-03-24.
980 ,@(when (c-lang-const c-opt-cpp-prefix)
982 ,@(when (c-major-mode-is '(c-mode c++-mode))
985 ,@(when (c-major-mode-is '(c-mode c++-mode))
986 '("%:%:" "??=??=")))))
989 ,@(c-lang-const c-identifier-ops)
990 ,@(cond ((or (c-major-mode-is 'c++-mode) (c-major-mode-is 'java-mode))
991 `((postfix-if-paren "<" ">"))) ; Templates.
992 ((c-major-mode-is 'pike-mode)
993 `((prefix "global" "predef")))
994 ((c-major-mode-is 'java-mode)
995 `((prefix "super"))))
998 ,@(when (c-major-mode-is 'c++-mode)
999 ;; The following need special treatment.
1000 `((prefix "dynamic_cast" "static_cast"
1001 "reinterpret_cast" "const_cast" "typeid"
1004 ,@(unless (c-major-mode-is 'java-mode)
1006 (postfix "++" "--" "[" "]" "(" ")"
1007 ,@(when (c-major-mode-is '(c-mode c++-mode))
1008 '("<:" ":>" "??(" "??)")))
1011 (prefix "++" "--" "+" "-" "!" "~"
1012 ,@(when (c-major-mode-is 'c++-mode) '("not" "compl"))
1013 ,@(when (c-major-mode-is '(c-mode c++-mode))
1014 '("*" "&" "sizeof" "??-"))
1015 ,@(when (c-major-mode-is 'objc-mode)
1016 '("@selector" "@protocol" "@encode"))
1017 ;; The following need special treatment.
1018 ,@(cond ((c-major-mode-is 'c++-mode)
1020 ((c-major-mode-is 'java-mode)
1022 ((c-major-mode-is 'pike-mode)
1023 '("class" "lambda" "catch" "throw" "gauge")))
1025 ,@(when (c-major-mode-is 'pike-mode)
1026 '("[" "]"))) ; Type cast.
1028 ;; Member selection.
1029 ,@(when (c-major-mode-is 'c++-mode)
1030 `((left-assoc ".*" "->*")))
1033 (left-assoc "*" "/" "%")
1036 (left-assoc "+" "-")
1039 (left-assoc "<<" ">>"
1040 ,@(when (c-major-mode-is 'java-mode)
1044 (left-assoc "<" ">" "<=" ">="
1045 ,@(when (c-major-mode-is 'java-mode)
1049 (left-assoc "==" "!="
1050 ,@(when (c-major-mode-is 'c++-mode) '("not_eq")))
1054 ,@(when (c-major-mode-is 'c++-mode) '("bitand")))
1056 ;; Bitwise exclusive or.
1058 ,@(when (c-major-mode-is '(c-mode c++-mode))
1060 ,@(when (c-major-mode-is 'c++-mode) '("xor")))
1064 ,@(when (c-major-mode-is '(c-mode c++-mode))
1066 ,@(when (c-major-mode-is 'c++-mode) '("bitor")))
1070 ,@(when (c-major-mode-is 'c++-mode) '("and")))
1074 ,@(when (c-major-mode-is '(c-mode c++-mode))
1076 ,@(when (c-major-mode-is 'c++-mode) '("or")))
1079 (right-assoc-sequence "?" ":")
1082 (right-assoc ,@(c-lang-const c-assignment-operators))
1085 ,@(when (c-major-mode-is 'c++-mode)
1086 '((prefix "throw")))
1091 ;; IDL got its own definition since it has a much smaller operator
1092 ;; set than the other languages.
1093 idl `(;; Preprocessor.
1097 ,@(c-lang-const c-identifier-ops)
1099 (prefix "+" "-" "~")
1101 (left-assoc "*" "/" "%")
1103 (left-assoc "+" "-")
1105 (left-assoc "<<" ">>")
1113 (c-lang-defconst c-operator-list
1114 ;; The operators as a flat list (without duplicates).
1115 t (c-filter-ops (c-lang-const c-operators) t t))
1117 (c-lang-defconst c-overloadable-operators
1118 "List of the operators that are overloadable, in their \"identifier
1119 form\". See also `c-op-identifier-prefix'."
1121 c++ '("new" "delete" ;; Can be followed by "[]" but we ignore that.
1123 "^" "??'" "xor" "&" "bitand" "|" "??!" "bitor" "~" "??-" "compl"
1124 "!" "=" "<" ">" "+=" "-=" "*=" "/=" "%=" "^="
1125 "??'=" "xor_eq" "&=" "and_eq" "|=" "??!=" "or_eq"
1126 "<<" ">>" ">>=" "<<=" "==" "!=" "not_eq" "<=" ">="
1127 "&&" "and" "||" "??!??!" "or" "++" "--" "," "->*" "->"
1128 "()" "[]" "<::>" "??(??)")
1129 ;; These work like identifiers in Pike.
1130 pike '("`+" "`-" "`&" "`|" "`^" "`<<" "`>>" "`*" "`/" "`%" "`~"
1131 "`==" "`<" "`>" "`!" "`[]" "`[]=" "`->" "`->=" "`()" "``+"
1132 "``-" "``&" "``|" "``^" "``<<" "``>>" "``*" "``/" "``%"
1135 (c-lang-defconst c-overloadable-operators-regexp
1136 ;; Regexp tested after an "operator" token in C++.
1138 c++ (c-make-keywords-re nil (c-lang-const c-overloadable-operators)))
1139 (c-lang-defvar c-overloadable-operators-regexp
1140 (c-lang-const c-overloadable-operators-regexp))
1142 (c-lang-defconst c-opt-op-identifier-prefix
1143 "Regexp matching the token before the ones in
1144 `c-overloadable-operators' when operators are specified in their
1145 \"identifier form\". This typically matches \"operator\" in C++ where
1146 operator functions are specified as e.g. \"operator +\". It's nil in
1147 languages without operator functions or where the complete operator
1148 identifier is listed in `c-overloadable-operators'.
1150 This regexp is assumed to not match any non-operator identifier."
1152 c++ (c-make-keywords-re t '("operator")))
1153 (c-lang-defvar c-opt-op-identifier-prefix
1154 (c-lang-const c-opt-op-identifier-prefix))
1156 ;; Note: the following alias is an old name which was a mis-spelling. It has
1157 ;; been corrected above and throughout cc-engine.el. It will be removed at
1158 ;; some release very shortly in the future. ACM, 2006-04-14.
1159 (defvaralias 'c-opt-op-identitier-prefix 'c-opt-op-identifier-prefix)
1160 (make-obsolete-variable 'c-opt-op-identitier-prefix 'c-opt-op-identifier-prefix
1161 "CC Mode 5.31.4, 2006-04-14")
1163 (c-lang-defconst c-other-op-syntax-tokens
1164 "List of the tokens made up of characters in the punctuation or
1165 parenthesis syntax classes that have uses other than as expression
1167 t '("{" "}" "(" ")" "[" "]" ";" ":" "," "=" "/*" "*/" "//")
1168 (c c++ pike) (append '("#" "##" ; Used by cpp.
1170 (c-lang-const c-other-op-syntax-tokens))
1171 (c c++) (append '("*") (c-lang-const c-other-op-syntax-tokens))
1172 c++ (append '("&" "<%" "%>" "<:" ":>" "%:" "%:%:")
1173 (c-lang-const c-other-op-syntax-tokens))
1174 objc (append '("#" "##" ; Used by cpp.
1176 (c-lang-const c-other-op-syntax-tokens))
1177 idl (append '("#" "##") ; Used by cpp.
1178 (c-lang-const c-other-op-syntax-tokens))
1179 pike (append '("..")
1180 (c-lang-const c-other-op-syntax-tokens)
1181 (c-lang-const c-overloadable-operators))
1182 awk '("{" "}" "(" ")" "[" "]" ";" "," "=" "/"))
1184 (c-lang-defconst c-all-op-syntax-tokens
1185 ;; List of all tokens in the punctuation and parenthesis syntax
1187 t (c--delete-duplicates (append (c-lang-const c-other-op-syntax-tokens)
1188 (c-lang-const c-operator-list))
1189 :test 'string-equal))
1191 (c-lang-defconst c-nonsymbol-token-char-list
1192 ;; List containing all chars not in the word, symbol or
1193 ;; syntactically irrelevant syntax classes, i.e. all punctuation,
1194 ;; parenthesis and string delimiter chars.
1195 t (c-with-syntax-table (c-lang-const c-mode-syntax-table)
1196 ;; Only go through the chars in the printable ASCII range. No
1197 ;; language so far has 8-bit or widestring operators.
1198 (let (list (char 32))
1200 (or (memq (char-syntax char) '(?w ?_ ?< ?> ?\ ))
1201 (setq list (cons (c-int-to-char char) list)))
1202 (setq char (1+ char)))
1205 (c-lang-defconst c-nonsymbol-token-regexp
1206 ;; Regexp matching all tokens in the punctuation and parenthesis
1207 ;; syntax classes. Note that this also matches ".", which can start
1209 t (c-make-keywords-re nil
1210 (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1212 "\\`\\(\\s.\\)+\\'")))
1213 (c-lang-defvar c-nonsymbol-token-regexp
1214 (c-lang-const c-nonsymbol-token-regexp))
1216 (c-lang-defconst c-assignment-op-regexp
1217 ;; Regexp matching all assignment operators and only them. The
1218 ;; beginning of the first submatch is used to detect the end of the
1219 ;; token, along with the end of the whole match.
1220 t (if (c-lang-const c-assignment-operators)
1222 ;; Need special case for "=" since it's a prefix of "==".
1225 (c-make-keywords-re nil
1226 (c--set-difference (c-lang-const c-assignment-operators)
1228 :test 'string-equal)))
1230 (c-lang-defvar c-assignment-op-regexp
1231 (c-lang-const c-assignment-op-regexp))
1233 (c-lang-defconst c-:$-multichar-token-regexp
1234 ;; Regexp matching all tokens ending in ":" which are longer than one char.
1235 ;; Currently (2016-01-07) only used in C++ Mode.
1236 t (c-make-keywords-re nil
1237 (c-filter-ops (c-lang-const c-operators) t ".+:$")))
1238 (c-lang-defvar c-:$-multichar-token-regexp
1239 (c-lang-const c-:$-multichar-token-regexp))
1241 (c-lang-defconst c-<>-multichar-token-regexp
1242 ;; Regexp matching all tokens containing "<" or ">" which are longer
1244 t (c-make-keywords-re nil
1245 (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1248 (c-lang-defvar c-<>-multichar-token-regexp
1249 (c-lang-const c-<>-multichar-token-regexp))
1251 (c-lang-defconst c-<-op-cont-regexp
1252 ;; Regexp matching the second and subsequent characters of all
1253 ;; multicharacter tokens that begin with "<".
1254 t (c-make-keywords-re nil
1255 (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1258 (lambda (op) (substring op 1)))))
1259 (c-lang-defvar c-<-op-cont-regexp (c-lang-const c-<-op-cont-regexp))
1261 (c-lang-defconst c->-op-cont-tokens
1262 ;; A list of second and subsequent characters of all multicharacter tokens
1263 ;; that begin with ">".
1264 t (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1267 (lambda (op) (substring op 1)))
1268 java (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1270 "\\`>[^>]\\|\\`>>[^>]"
1271 (lambda (op) (substring op 1))))
1273 (c-lang-defconst c->-op-cont-regexp
1274 ;; Regexp matching the second and subsequent characters of all
1275 ;; multicharacter tokens that begin with ">".
1276 t (c-make-keywords-re nil (c-lang-const c->-op-cont-tokens)))
1277 (c-lang-defvar c->-op-cont-regexp (c-lang-const c->-op-cont-regexp))
1279 (c-lang-defconst c->-op-without->-cont-regexp
1280 ;; Regexp matching the second and subsequent characters of all
1281 ;; multicharacter tokens that begin with ">" except for those beginning with
1283 t (c-make-keywords-re nil
1285 (c-lang-const c->-op-cont-tokens)
1286 (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1289 (lambda (op) (substring op 1)))
1290 :test 'string-equal)))
1291 (c-lang-defvar c->-op-without->-cont-regexp
1292 (c-lang-const c->-op-without->-cont-regexp))
1294 (c-lang-defconst c-multichar->-op-not->>-regexp
1295 ;; Regexp matching multichar tokens containing ">", except ">>"
1296 t (c-make-keywords-re nil
1298 (c-filter-ops (c-lang-const c-all-op-syntax-tokens)
1301 (c-lang-defvar c-multichar->-op-not->>-regexp
1302 (c-lang-const c-multichar->-op-not->>-regexp))
1304 (c-lang-defconst c-stmt-delim-chars
1305 ;; The characters that should be considered to bound statements. To
1306 ;; optimize `c-crosses-statement-barrier-p' somewhat, it's assumed to
1307 ;; begin with "^" to negate the set. If ? : operators should be
1308 ;; detected then the string must end with "?:".
1310 (c-lang-defvar c-stmt-delim-chars (c-lang-const c-stmt-delim-chars))
1312 (c-lang-defconst c-stmt-delim-chars-with-comma
1313 ;; Variant of `c-stmt-delim-chars' that additionally contains ','.
1315 (c-lang-defvar c-stmt-delim-chars-with-comma
1316 (c-lang-const c-stmt-delim-chars-with-comma))
1318 (c-lang-defconst c-auto-ops
1319 ;; Ops which signal C++11's new auto uses.
1322 (c-lang-defconst c-auto-ops-re
1323 t (c-make-keywords-re nil (c-lang-const c-auto-ops)))
1324 (c-lang-defvar c-auto-ops-re (c-lang-const c-auto-ops-re))
1326 (c-lang-defconst c-haskell-op
1327 ;; Op used in the new C++11 auto function definition, indicating type.
1330 (c-lang-defconst c-haskell-op-re
1331 t (c-make-keywords-re nil (c-lang-const c-haskell-op)))
1332 (c-lang-defvar c-haskell-op-re (c-lang-const c-haskell-op-re))
1334 ;;; Syntactic whitespace.
1336 (c-lang-defconst c-simple-ws
1337 "Regexp matching an ordinary whitespace character.
1338 Does not contain a \\| operator at the top level."
1339 ;; "\\s " is not enough since it doesn't match line breaks.
1340 t "\\(\\s \\|[\n\r]\\)")
1342 (c-lang-defconst c-simple-ws-depth
1343 ;; Number of regexp grouping parens in `c-simple-ws'.
1344 t (regexp-opt-depth (c-lang-const c-simple-ws)))
1346 (c-lang-defconst c-line-comment-starter
1347 "String that starts line comments, or nil if such don't exist.
1348 Line comments are always terminated by newlines. At least one of
1349 `c-block-comment-starter' and this one is assumed to be set.
1351 Note that it's currently not enough to set this to support a new
1352 comment style. Other stuff like the syntax table must also be set up
1356 (c-lang-defvar c-line-comment-starter (c-lang-const c-line-comment-starter))
1358 (c-lang-defconst c-block-comment-starter
1359 "String that starts block comments, or nil if such don't exist.
1360 Block comments are ended by `c-block-comment-ender', which is assumed
1361 to be set if this is. At least one of `c-line-comment-starter' and
1362 this one is assumed to be set.
1364 Note that it's currently not enough to set this to support a new
1365 comment style. Other stuff like the syntax table must also be set up
1370 (c-lang-defconst c-block-comment-ender
1371 "String that ends block comments, or nil if such don't exist.
1373 Note that it's currently not enough to set this to support a new
1374 comment style. Other stuff like the syntax table must also be set up
1379 (c-lang-defconst c-comment-start-regexp
1380 ;; Regexp to match the start of any type of comment.
1381 t (let ((re (c-make-keywords-re nil
1382 (list (c-lang-const c-line-comment-starter)
1383 (c-lang-const c-block-comment-starter)))))
1384 (if (memq 'gen-comment-delim c-emacs-features)
1385 (concat re "\\|\\s!")
1387 (c-lang-defvar c-comment-start-regexp (c-lang-const c-comment-start-regexp))
1389 (c-lang-defconst c-block-comment-start-regexp
1390 ;; Regexp which matches the start of a block comment (if such exists in the
1392 t (if (c-lang-const c-block-comment-starter)
1393 (regexp-quote (c-lang-const c-block-comment-starter))
1395 (c-lang-defvar c-block-comment-start-regexp
1396 (c-lang-const c-block-comment-start-regexp))
1398 (c-lang-defconst c-line-comment-start-regexp
1399 ;; Regexp which matches the start of a line comment (if such exists in the
1400 ;; language; it does in all 7 CC Mode languages).
1401 t (if (c-lang-const c-line-comment-starter)
1402 (regexp-quote (c-lang-const c-line-comment-starter))
1404 (c-lang-defvar c-line-comment-start-regexp
1405 (c-lang-const c-line-comment-start-regexp))
1407 (c-lang-defconst c-literal-start-regexp
1408 ;; Regexp to match the start of comments and string literals.
1409 t (concat (c-lang-const c-comment-start-regexp)
1411 (if (memq 'gen-string-delim c-emacs-features)
1414 (c-lang-defvar c-literal-start-regexp (c-lang-const c-literal-start-regexp))
1416 (c-lang-defconst c-doc-comment-start-regexp
1417 "Regexp to match the start of documentation comments."
1419 ;; From font-lock.el: `doxygen' uses /*! while others use /**.
1420 (c c++ objc) "/\\*[*!]"
1423 (c-lang-defvar c-doc-comment-start-regexp
1424 (c-lang-const c-doc-comment-start-regexp))
1426 (c-lang-defconst comment-start
1427 "String that starts comments inserted with M-; etc.
1428 `comment-start' is initialized from this."
1429 ;; Default: Prefer line comments to block comments, and pad with a space.
1430 t (concat (or (c-lang-const c-line-comment-starter)
1431 (c-lang-const c-block-comment-starter))
1433 ;; In C we still default to the block comment style since line
1434 ;; comments aren't entirely portable.
1436 (c-lang-setvar comment-start (c-lang-const comment-start))
1438 (c-lang-defconst comment-end
1439 "String that ends comments inserted with M-; etc.
1440 `comment-end' is initialized from this."
1441 ;; Default: Use block comment style if comment-start uses block
1442 ;; comments, and pad with a space in that case.
1443 t (if (string-match (concat "\\`\\("
1444 (c-lang-const c-block-comment-start-regexp)
1446 (c-lang-const comment-start))
1447 (concat " " (c-lang-const c-block-comment-ender))
1449 (c-lang-setvar comment-end (c-lang-const comment-end))
1451 (c-lang-defconst comment-start-skip
1452 "Regexp to match the start of a comment plus everything up to its body.
1453 `comment-start-skip' is initialized from this."
1454 ;; Default: Allow the last char of the comment starter(s) to be
1455 ;; repeated, then allow any amount of horizontal whitespace.
1458 (mapcar (lambda (cs)
1460 (concat (regexp-quote cs) "+")))
1461 (list (c-lang-const c-line-comment-starter)
1462 (c-lang-const c-block-comment-starter)))
1465 (c-lang-setvar comment-start-skip (c-lang-const comment-start-skip))
1467 (c-lang-defconst comment-end-can-be-escaped
1468 "When non-nil, escaped EOLs inside comments are valid.
1469 This works in Emacs >= 25.1."
1472 (c-lang-setvar comment-end-can-be-escaped
1473 (c-lang-const comment-end-can-be-escaped))
1475 (c-lang-defconst c-syntactic-ws-start
1476 ;; Regexp matching any sequence that can start syntactic whitespace.
1477 ;; The only uncertain case is '#' when there are cpp directives.
1479 (c-make-keywords-re nil
1480 (append (list (c-lang-const c-line-comment-starter)
1481 (c-lang-const c-block-comment-starter)
1482 (when (c-lang-const c-opt-cpp-prefix)
1486 (when (memq 'gen-comment-delim c-emacs-features)
1488 (c-lang-defvar c-syntactic-ws-start (c-lang-const c-syntactic-ws-start))
1490 (c-lang-defconst c-syntactic-ws-end
1491 ;; Regexp matching any single character that might end syntactic whitespace.
1493 (c-make-keywords-re nil
1494 (append (when (c-lang-const c-block-comment-ender)
1497 (elt (c-lang-const c-block-comment-ender)
1499 (c-lang-const c-block-comment-ender)))))))
1501 (when (memq 'gen-comment-delim c-emacs-features)
1503 (c-lang-defvar c-syntactic-ws-end (c-lang-const c-syntactic-ws-end))
1505 (c-lang-defconst c-unterminated-block-comment-regexp
1506 ;; Regexp matching an unterminated block comment that doesn't
1507 ;; contain line breaks, or nil in languages without block comments.
1508 ;; Does not contain a \| operator at the top level.
1509 t (when (c-lang-const c-block-comment-starter)
1511 (regexp-quote (c-lang-const c-block-comment-starter))
1512 ;; It's messy to cook together a regexp that matches anything
1513 ;; but c-block-comment-ender.
1514 (let ((end (c-lang-const c-block-comment-ender)))
1515 (cond ((= (length end) 1)
1516 (concat "[^" end "\n\r]*"))
1518 (concat "[^" (substring end 0 1) "\n\r]*"
1520 (regexp-quote (substring end 0 1)) "+"
1522 ;; The quoting rules inside char classes are silly. :P
1523 (cond ((= (elt end 0) (elt end 1))
1524 (concat (substring end 0 1) "\n\r"))
1525 ((= (elt end 1) ?\])
1526 (concat (substring end 1 2) "\n\r"
1527 (substring end 0 1)))
1529 (concat (substring end 0 1) "\n\r"
1530 (substring end 1 2))))
1532 "[^" (substring end 0 1) "\n\r]*"
1535 (error "Can't handle a block comment ender of length %s"
1538 (c-lang-defconst c-block-comment-regexp
1539 ;; Regexp matching a block comment that doesn't contain line breaks,
1540 ;; or nil in languages without block comments. The reason we don't
1541 ;; allow line breaks is to avoid going very far and risk running out
1542 ;; of regexp stack; this regexp is intended to handle only short
1543 ;; comments that might be put in the middle of limited constructs
1544 ;; like declarations. Does not contain a \| operator at the top
1546 t (when (c-lang-const c-unterminated-block-comment-regexp)
1548 (c-lang-const c-unterminated-block-comment-regexp)
1549 (let ((end (c-lang-const c-block-comment-ender)))
1550 (cond ((= (length end) 1)
1553 (concat (regexp-quote (substring end 0 1)) "+"
1554 (regexp-quote (substring end 1 2))))
1556 (error "Can't handle a block comment ender of length %s"
1559 (c-lang-defconst c-nonwhite-syntactic-ws
1560 ;; Regexp matching a piece of syntactic whitespace that isn't a
1561 ;; sequence of simple whitespace characters. As opposed to
1562 ;; `c-(forward|backward)-syntactic-ws', this doesn't regard cpp
1563 ;; directives as syntactic whitespace.
1564 t (c-concat-separated
1565 (list (when (c-lang-const c-line-comment-starter)
1566 (concat (regexp-quote (c-lang-const c-line-comment-starter))
1568 (c-lang-const c-block-comment-regexp)
1570 (when (memq 'gen-comment-delim c-emacs-features)
1574 (c-lang-defconst c-syntactic-ws
1575 ;; Regexp matching syntactic whitespace, including possibly the
1576 ;; empty string. As opposed to `c-(forward|backward)-syntactic-ws',
1577 ;; this doesn't regard cpp directives as syntactic whitespace. Does
1578 ;; not contain a \| operator at the top level.
1579 t (concat (c-lang-const c-simple-ws) "*"
1581 (concat "\\(" (c-lang-const c-nonwhite-syntactic-ws) "\\)"
1582 (c-lang-const c-simple-ws) "*")
1585 (c-lang-defconst c-syntactic-ws-depth
1586 ;; Number of regexp grouping parens in `c-syntactic-ws'.
1587 t (regexp-opt-depth (c-lang-const c-syntactic-ws)))
1589 (c-lang-defconst c-nonempty-syntactic-ws
1590 ;; Regexp matching syntactic whitespace, which is at least one
1591 ;; character long. As opposed to `c-(forward|backward)-syntactic-ws',
1592 ;; this doesn't regard cpp directives as syntactic whitespace. Does
1593 ;; not contain a \| operator at the top level.
1595 (c-lang-const c-simple-ws)
1597 (c-lang-const c-nonwhite-syntactic-ws)
1600 (c-lang-defconst c-nonempty-syntactic-ws-depth
1601 ;; Number of regexp grouping parens in `c-nonempty-syntactic-ws'.
1602 t (regexp-opt-depth (c-lang-const c-nonempty-syntactic-ws)))
1604 (c-lang-defconst c-single-line-syntactic-ws
1605 ;; Regexp matching syntactic whitespace without any line breaks. As
1606 ;; opposed to `c-(forward|backward)-syntactic-ws', this doesn't
1607 ;; regard cpp directives as syntactic whitespace. Does not contain
1608 ;; a \| operator at the top level.
1609 t (if (c-lang-const c-block-comment-regexp)
1611 (c-lang-const c-block-comment-regexp)
1615 (c-lang-defconst c-single-line-syntactic-ws-depth
1616 ;; Number of regexp grouping parens in `c-single-line-syntactic-ws'.
1617 t (regexp-opt-depth (c-lang-const c-single-line-syntactic-ws)))
1619 (c-lang-defconst c-syntactic-eol
1620 ;; Regexp that matches when there is no syntactically significant
1621 ;; text before eol. Macros are regarded as syntactically
1622 ;; significant text here.
1623 t (concat (c-lang-const c-single-line-syntactic-ws)
1624 ;; Match eol (possibly inside a block comment or preceded
1625 ;; by a line continuation backslash), or the beginning of a
1626 ;; line comment. Note: This has to be modified for awk
1627 ;; where line comments start with '#'.
1630 (list (when (c-lang-const c-line-comment-starter)
1631 (regexp-quote (c-lang-const c-line-comment-starter)))
1632 (when (c-lang-const c-unterminated-block-comment-regexp)
1633 (concat (c-lang-const c-unterminated-block-comment-regexp)
1639 (c-lang-defvar c-syntactic-eol (c-lang-const c-syntactic-eol))
1644 ;; The Emacs variables beginning-of-defun-function and end-of-defun-function
1645 ;; will be set so that commands like `mark-defun' and `narrow-to-defun' work
1646 ;; right. In older Emacsen, the key sequences C-M-a and C-M-e are, however,
1647 ;; bound directly to the CC Mode functions, allowing optimization for large n.
1648 ;; From Emacs 23, this isn't necessary any more, since n is passed to the two
1650 (c-lang-defconst beginning-of-defun-function
1651 "Function to which beginning-of-defun-function will be set."
1652 t 'c-beginning-of-defun
1653 awk 'c-awk-beginning-of-defun)
1654 (c-lang-setvar beginning-of-defun-function
1655 (c-lang-const beginning-of-defun-function))
1657 (c-lang-defconst end-of-defun-function
1658 "Function to which end-of-defun-function will be set."
1660 awk 'c-awk-end-of-defun)
1661 (c-lang-setvar end-of-defun-function (c-lang-const end-of-defun-function))
1663 ;;; In-comment text handling.
1665 (c-lang-defconst c-paragraph-start
1666 "Regexp to append to `paragraph-start'."
1668 java "\\(@[a-zA-Z]+\\>\\|$\\)" ; For Javadoc.
1669 pike "\\(@[a-zA-Z_-]+\\>\\([^{]\\|$\\)\\|$\\)") ; For Pike refdoc.
1670 (c-lang-defvar c-paragraph-start (c-lang-const c-paragraph-start))
1672 (c-lang-defconst c-paragraph-separate
1673 "Regexp to append to `paragraph-separate'."
1675 pike (c-lang-const c-paragraph-start))
1676 (c-lang-defvar c-paragraph-separate (c-lang-const c-paragraph-separate))
1681 ;; Note: All and only all language constants containing keyword lists
1682 ;; should end with "-kwds"; they're automatically collected into the
1683 ;; `c-kwds-lang-consts' list below and used to build `c-keywords' etc.
1685 (c-lang-defconst c-primitive-type-kwds
1686 "Primitive type keywords. As opposed to the other keyword lists, the
1687 keywords listed here are fontified with the type face instead of the
1690 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1691 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1692 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1695 Do not try to modify this list for end user customizations; the
1696 `*-font-lock-extra-types' variable, where `*' is the mode prefix, is
1697 the appropriate place for that."
1698 t '("char" "double" "float" "int" "long" "short" "signed"
1701 '("_Bool" "_Complex" "_Imaginary") ; Conditionally defined in C99.
1702 (c-lang-const c-primitive-type-kwds))
1704 '("bool" "wchar_t" "char16_t" "char32_t")
1705 (c-lang-const c-primitive-type-kwds))
1706 ;; Objective-C extends C, but probably not the new stuff in C99.
1708 '("id" "Class" "SEL" "IMP" "BOOL")
1709 (c-lang-const c-primitive-type-kwds))
1710 java '("boolean" "byte" "char" "double" "float" "int" "long" "short" "void")
1711 idl '("Object" "ValueBase" "any" "boolean" "char" "double" "fixed" "float"
1712 "long" "octet" "sequence" "short" "string" "void" "wchar" "wstring"
1715 ;; The following can't really end a type, but we have to specify them
1716 ;; here due to the assumption in `c-primitive-type-prefix-kwds'. It
1717 ;; doesn't matter that much.
1718 "unsigned" "strong")
1719 pike '(;; this_program isn't really a keyword, but it's practically
1720 ;; used as a builtin type.
1721 "array" "float" "function" "int" "mapping" "mixed" "multiset"
1722 "object" "program" "string" "this_program" "void"))
1724 (c-lang-defconst c-primitive-type-key
1725 ;; An adorned regexp that matches `c-primitive-type-kwds'.
1726 t (c-make-keywords-re t (c-lang-const c-primitive-type-kwds)))
1727 (c-lang-defvar c-primitive-type-key (c-lang-const c-primitive-type-key))
1729 (c-lang-defconst c-primitive-type-prefix-kwds
1730 "Keywords that might act as prefixes for primitive types. Assumed to
1731 be a subset of `c-primitive-type-kwds'."
1733 (c c++) '("long" "short" "signed" "unsigned")
1734 idl '("long" "unsigned"
1738 (c-lang-defconst c-typedef-kwds
1739 "Prefix keyword(s) like \"typedef\" which make a type declaration out
1740 of a variable declaration."
1744 (c-lang-defconst c-typedef-key
1745 ;; Adorned regexp matching `c-typedef-kwds'.
1746 t (c-make-keywords-re t (c-lang-const c-typedef-kwds)))
1747 (c-lang-defvar c-typedef-key (c-lang-const c-typedef-key))
1749 (c-lang-defconst c-typeof-kwds
1750 "Keywords followed by a parenthesized expression, which stands for
1751 the type of that expression."
1753 c '("typeof") ; longstanding GNU C(++) extension.
1754 c++ '("decltype" "typeof"))
1756 (c-lang-defconst c-typeof-key
1757 ;; Adorned regexp matching `c-typeof-kwds'.
1758 t (c-make-keywords-re t (c-lang-const c-typeof-kwds)))
1759 (c-lang-defvar c-typeof-key (c-lang-const c-typeof-key))
1761 (c-lang-defconst c-type-prefix-kwds
1762 "Keywords where the following name - if any - is a type name, and
1763 where the keyword together with the symbol works as a type in
1766 Note that an alternative if the second part doesn't hold is
1767 `c-type-list-kwds'. Keywords on this list are typically also present
1768 on one of the `*-decl-kwds' lists."
1770 c '("struct" "union" "enum")
1771 c++ (append '("class" "typename")
1772 (c-lang-const c-type-prefix-kwds c)))
1774 (c-lang-defconst c-type-prefix-key
1775 ;; Adorned regexp matching `c-type-prefix-kwds'.
1776 t (c-make-keywords-re t (c-lang-const c-type-prefix-kwds)))
1777 (c-lang-defvar c-type-prefix-key (c-lang-const c-type-prefix-key))
1779 (c-lang-defconst c-type-modifier-kwds
1780 "Type modifier keywords. These can occur almost anywhere in types
1781 but they don't build a type of themselves. Unlike the keywords on
1782 `c-primitive-type-kwds', they are fontified with the keyword face and
1785 c '("const" "restrict" "volatile")
1786 c++ '("const" "constexpr" "noexcept" "volatile" "throw" "final" "override")
1787 objc '("const" "volatile"))
1789 (c-lang-defconst c-opt-type-modifier-key
1790 ;; Adorned regexp matching `c-type-modifier-kwds', or nil in
1791 ;; languages without such keywords.
1792 t (and (c-lang-const c-type-modifier-kwds)
1793 (c-make-keywords-re t (c-lang-const c-type-modifier-kwds))))
1794 (c-lang-defvar c-opt-type-modifier-key (c-lang-const c-opt-type-modifier-key))
1796 (c-lang-defconst c-opt-type-component-key
1797 ;; An adorned regexp that matches `c-primitive-type-prefix-kwds' and
1798 ;; `c-type-modifier-kwds', or nil in languages without any of them.
1799 t (and (or (c-lang-const c-primitive-type-prefix-kwds)
1800 (c-lang-const c-type-modifier-kwds))
1801 (c-make-keywords-re t
1802 (append (c-lang-const c-primitive-type-prefix-kwds)
1803 (c-lang-const c-type-modifier-kwds)))))
1804 (c-lang-defvar c-opt-type-component-key
1805 (c-lang-const c-opt-type-component-key))
1807 (c-lang-defconst c-type-start-kwds
1808 ;; All keywords that can start a type (i.e. are either a type prefix
1809 ;; or a complete type).
1810 t (c--delete-duplicates (append (c-lang-const c-primitive-type-kwds)
1811 (c-lang-const c-type-prefix-kwds)
1812 (c-lang-const c-type-modifier-kwds))
1813 :test 'string-equal))
1815 (c-lang-defconst c-class-decl-kwds
1816 "Keywords introducing declarations where the following block (if any)
1817 contains another declaration level that should be considered a class.
1819 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1820 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1821 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1824 Note that presence on this list does not automatically treat the
1825 following identifier as a type; the keyword must also be present on
1826 `c-type-prefix-kwds' or `c-type-list-kwds' to accomplish that."
1828 c '("struct" "union")
1829 c++ '("class" "struct" "union")
1830 objc '("struct" "union"
1831 "@interface" "@implementation" "@protocol")
1832 java '("class" "@interface" "interface")
1833 idl '("component" "eventtype" "exception" "home" "interface" "struct"
1836 "storagehome" "storagetype"
1838 "catalog" "executor" "manages" "segment")
1841 (c-lang-defconst c-class-key
1842 ;; Regexp matching the start of a class.
1843 t (c-make-keywords-re t (c-lang-const c-class-decl-kwds)))
1844 (c-lang-defvar c-class-key (c-lang-const c-class-key))
1846 (c-lang-defconst c-brace-list-decl-kwds
1847 "Keywords introducing declarations where the following block (if
1848 any) is a brace list.
1850 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1851 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1852 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1857 (c-lang-defconst c-brace-list-key
1858 ;; Regexp matching the start of declarations where the following
1859 ;; block is a brace list.
1860 t (c-make-keywords-re t (c-lang-const c-brace-list-decl-kwds)))
1861 (c-lang-defvar c-brace-list-key (c-lang-const c-brace-list-key))
1863 (c-lang-defconst c-after-brace-list-decl-kwds
1864 "Keywords that might follow keywords in `c-brace-list-decl-kwds'
1865 and precede the opening brace."
1867 c++ '("class" "struct"))
1869 (c-lang-defconst c-after-brace-list-key
1870 ;; Regexp matching keywords that can fall between a brace-list
1871 ;; keyword and the associated brace list.
1872 t (c-make-keywords-re t (c-lang-const c-after-brace-list-decl-kwds)))
1873 (c-lang-defvar c-after-brace-list-key (c-lang-const c-after-brace-list-key))
1875 (c-lang-defconst c-recognize-post-brace-list-type-p
1876 "Set to t when we recognize a colon and then a type after an enum,
1877 e.g., enum foo : int { A, B, C };"
1880 (c-lang-defvar c-recognize-post-brace-list-type-p
1881 (c-lang-const c-recognize-post-brace-list-type-p))
1883 (c-lang-defconst c-other-block-decl-kwds
1884 "Keywords where the following block (if any) contains another
1885 declaration level that should not be considered a class. For every
1886 keyword here, CC Mode will add a set of special syntactic symbols for
1887 those blocks. E.g. if the keyword is \"foo\" then there will be
1888 `foo-open', `foo-close', and `infoo' symbols.
1890 The intention is that this category should be used for block
1891 constructs that aren't related to object orientation concepts like
1892 classes (which thus also include e.g. interfaces, templates,
1893 contracts, structs, etc). The more pragmatic distinction is that
1894 while most want some indentation inside classes, it's fairly common
1895 that they don't want it in some of these constructs, so it should be
1896 simple to configure that differently from classes. See also
1897 `c-class-decl-kwds'.
1899 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1900 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1901 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1904 (c objc) '("extern")
1905 c++ '("namespace" "extern")
1910 (c-lang-defconst c-other-decl-block-key
1911 ;; Regexp matching the start of blocks besides classes that contain
1912 ;; another declaration level.
1913 t (c-make-keywords-re t (c-lang-const c-other-block-decl-kwds)))
1914 (c-lang-defvar c-other-decl-block-key (c-lang-const c-other-decl-block-key))
1916 (c-lang-defvar c-other-decl-block-key-in-symbols-alist
1920 (if (string= elt "extern")
1922 (intern (concat "in" elt)))))
1923 (c-lang-const c-other-block-decl-kwds))
1924 "Alist associating keywords in c-other-decl-block-decl-kwds with
1925 their matching \"in\" syntactic symbols.")
1927 (c-lang-defconst c-typedef-decl-kwds
1928 "Keywords introducing declarations where the identifier(s) being
1931 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1932 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1933 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1935 ;; Default to `c-class-decl-kwds' and `c-brace-list-decl-kwds'
1936 ;; (since e.g. "Foo" is a type that's being defined in "class Foo
1938 t (append (c-lang-const c-class-decl-kwds)
1939 (c-lang-const c-brace-list-decl-kwds))
1940 ;; Languages that have a "typedef" construct.
1941 (c c++ objc idl pike) (append (c-lang-const c-typedef-decl-kwds)
1943 ;; Unlike most other languages, exception names are not handled as
1944 ;; types in IDL since they only can occur in "raises" specs.
1945 idl (delete "exception" (append (c-lang-const c-typedef-decl-kwds) nil)))
1947 (c-lang-defconst c-typedef-decl-key
1948 t (c-make-keywords-re t (c-lang-const c-typedef-decl-kwds)))
1949 (c-lang-defvar c-typedef-decl-key (c-lang-const c-typedef-decl-key))
1951 (c-lang-defconst c-typeless-decl-kwds
1952 "Keywords introducing declarations where the \(first) identifier
1953 \(declarator) follows directly after the keyword, without any type.
1955 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1956 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1957 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1959 ;; Default to `c-class-decl-kwds' and `c-brace-list-decl-kwds'
1960 ;; (since e.g. "Foo" is the identifier being defined in "class Foo
1962 t (append (c-lang-const c-class-decl-kwds)
1963 (c-lang-const c-brace-list-decl-kwds))
1964 c++ (append (c-lang-const c-typeless-decl-kwds) '("auto")) ; C++11.
1965 ;; Note: "manages" for CORBA CIDL clashes with its presence on
1966 ;; `c-type-list-kwds' for IDL.
1967 idl (append (c-lang-const c-typeless-decl-kwds)
1968 '("factory" "finder" "native"
1973 pike (append (c-lang-const c-class-decl-kwds)
1976 (c-lang-defconst c-modifier-kwds
1977 "Keywords that can prefix normal declarations of identifiers
1978 \(and typically act as flags). Things like argument declarations
1979 inside function headers are also considered declarations in this
1982 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
1983 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
1984 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
1987 (c c++) '("auto" "extern" "inline" "register" "static")
1988 c++ (append '("explicit" "friend" "mutable" "template" "thread_local"
1990 (c-lang-const c-modifier-kwds))
1991 objc '("auto" "bycopy" "byref" "extern" "in" "inout" "oneway" "out" "static")
1992 ;; FIXME: Some of those below ought to be on `c-other-decl-kwds' instead.
1993 idl '("abstract" "attribute" "const" "consumes" "custom" "emits" "import"
1994 "in" "inout" "local" "multiple" "oneway" "out" "private" "provides"
1995 "public" "publishes" "readonly" "typeid" "typeprefix" "uses"
1999 "bindsTo" "delegatesTo" "implements" "proxy" "storedOn")
2000 ;; Note: "const" is not used in Java, but it's still a reserved keyword.
2001 java '("abstract" "const" "default" "final" "native" "private" "protected"
2002 "public" "static" "strictfp" "synchronized" "transient" "volatile")
2003 pike '("final" "inline" "local" "nomask" "optional" "private" "protected"
2004 "public" "static" "variant"))
2006 (c-lang-defconst c-other-decl-kwds
2007 "Keywords that can start or prefix any declaration level construct,
2008 besides those on `c-class-decl-kwds', `c-brace-list-decl-kwds',
2009 `c-other-block-decl-kwds', `c-typedef-decl-kwds',
2010 `c-typeless-decl-kwds' and `c-modifier-kwds'.
2012 If any of these also are on `c-type-list-kwds', `c-ref-list-kwds',
2013 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
2014 `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses
2017 objc '("@class" "@end" "@defs")
2018 java '("import" "package")
2019 pike '("import" "inherit"))
2021 (c-lang-defconst c-decl-start-kwds
2022 "Keywords that always start declarations, wherever they occur.
2023 This can be used for declarations that aren't recognized by the normal
2024 combination of `c-decl-prefix-re' and `c-decl-start-re'."
2026 ;; Classes can be declared anywhere in a Pike expression.
2029 (c-lang-defconst c-decl-hangon-kwds
2030 "Keywords that can occur anywhere in a declaration level construct.
2031 This is used for self-contained things that can be tacked on anywhere
2032 on a declaration and that should be ignored to be able to recognize it
2033 correctly. Typical cases are compiler extensions like
2034 \"__attribute__\" or \"__declspec\":
2036 __declspec(noreturn) void foo();
2037 class __declspec(dllexport) classname {...};
2038 void foo() __attribute__((noreturn));
2040 Note that unrecognized plain symbols are skipped anyway if they occur
2041 before the type, so such things are not necessary to mention here.
2042 Mentioning them here is necessary only if they can occur in other
2043 places, or if they are followed by a construct that must be skipped
2044 over \(like the parens in the \"__attribute__\" and \"__declspec\"
2045 examples above). In the last case, they alse need to be present on
2046 one of `c-type-list-kwds', `c-ref-list-kwds',
2047 `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds',
2048 `c-<>-type-kwds', or `c-<>-arglist-kwds'."
2049 ;; NB: These are currently not recognized in all parts of a
2050 ;; declaration. Specifically, they aren't recognized in the middle
2051 ;; of multi-token types, inside declarators, and between the
2052 ;; identifier and the arglist paren of a function declaration.
2054 ;; FIXME: This ought to be user customizable since compiler stuff
2055 ;; like this usually is wrapped in project specific macros. (It'd
2056 ;; of course be even better if we could cope without knowing this.)
2058 (c c++) '(;; GCC extension.
2063 (c-lang-defconst c-decl-hangon-key
2064 ;; Adorned regexp matching `c-decl-hangon-kwds'.
2065 t (c-make-keywords-re t (c-lang-const c-decl-hangon-kwds)))
2066 (c-lang-defvar c-decl-hangon-key (c-lang-const c-decl-hangon-key))
2068 (c-lang-defconst c-prefix-spec-kwds
2069 ;; All keywords that can occur in the preamble of a declaration.
2070 ;; They typically occur before the type, but they are also matched
2071 ;; after presumptive types since we often can't be sure that
2072 ;; something is a type or just some sort of macro in front of the
2073 ;; declaration. They might be ambiguous with types or type
2075 t (c--delete-duplicates (append (c-lang-const c-class-decl-kwds)
2076 (c-lang-const c-brace-list-decl-kwds)
2077 (c-lang-const c-other-block-decl-kwds)
2078 (c-lang-const c-typedef-decl-kwds)
2079 (c-lang-const c-typeless-decl-kwds)
2080 (c-lang-const c-modifier-kwds)
2081 (c-lang-const c-other-decl-kwds)
2082 (c-lang-const c-decl-start-kwds)
2083 (c-lang-const c-decl-hangon-kwds))
2084 :test 'string-equal))
2086 (c-lang-defconst c-prefix-spec-kwds-re
2087 ;; Adorned regexp of `c-prefix-spec-kwds'.
2088 t (c-make-keywords-re t (c-lang-const c-prefix-spec-kwds)))
2090 (c-lang-defvar c-prefix-spec-kwds-re (c-lang-const c-prefix-spec-kwds-re))
2092 (c-lang-defconst c-specifier-key
2093 ;; Adorned regexp of the keywords in `c-prefix-spec-kwds' that aren't
2094 ;; ambiguous with types or type prefixes. These are the keywords (like
2095 ;; extern, namespace, but NOT template) that can modify a declaration.
2096 t (c-make-keywords-re t
2097 (c--set-difference (c-lang-const c-prefix-spec-kwds)
2098 (append (c-lang-const c-type-start-kwds)
2099 (c-lang-const c-<>-arglist-kwds))
2100 :test 'string-equal)))
2101 (c-lang-defvar c-specifier-key (c-lang-const c-specifier-key))
2103 (c-lang-defconst c-postfix-spec-kwds
2104 ;; Keywords that can occur after argument list of a function header
2105 ;; declaration, i.e. in the "K&R region".
2106 t (append (c-lang-const c-postfix-decl-spec-kwds)
2107 (c-lang-const c-decl-hangon-kwds)))
2109 (c-lang-defconst c-not-decl-init-keywords
2110 ;; Adorned regexp matching all keywords that can't appear at the
2111 ;; start of a declaration.
2112 t (c-make-keywords-re t
2113 (c--set-difference (c-lang-const c-keywords)
2114 (append (c-lang-const c-type-start-kwds)
2115 (c-lang-const c-prefix-spec-kwds)
2116 (c-lang-const c-typeof-kwds))
2117 :test 'string-equal)))
2118 (c-lang-defvar c-not-decl-init-keywords
2119 (c-lang-const c-not-decl-init-keywords))
2121 (c-lang-defconst c-not-primitive-type-keywords
2122 "List of all keywords apart from primitive types (like \"int\")."
2123 t (c--set-difference (c-lang-const c-keywords)
2124 (c-lang-const c-primitive-type-kwds)
2125 :test 'string-equal)
2126 ;; The "more" for C++ is the QT keyword (as in "more slots:").
2127 ;; This variable is intended for use in c-beginning-of-statement-1.
2128 c++ (append (c-lang-const c-not-primitive-type-keywords) '("more")))
2130 (c-lang-defconst c-not-primitive-type-keywords-regexp
2131 t (c-make-keywords-re t
2132 (c-lang-const c-not-primitive-type-keywords)))
2133 (c-lang-defvar c-not-primitive-type-keywords-regexp
2134 (c-lang-const c-not-primitive-type-keywords-regexp))
2136 (c-lang-defconst c-protection-kwds
2137 "Access protection label keywords in classes."
2139 c++ '("private" "protected" "public")
2140 objc '("@private" "@protected" "@public"))
2142 (c-lang-defconst c-block-decls-with-vars
2143 "Keywords introducing declarations that can contain a block which
2144 might be followed by variable declarations, e.g. like \"foo\" in
2145 \"class Foo { ... } foo;\". So if there is a block in a declaration
2146 like that, it ends with the following `;' and not right away.
2148 The keywords on list are assumed to also be present on one of the
2149 `*-decl-kwds' lists."
2151 (c objc) '("struct" "union" "enum" "typedef")
2152 c++ '("class" "struct" "union" "enum" "typedef"))
2154 (c-lang-defconst c-opt-block-decls-with-vars-key
2155 ;; Regexp matching the `c-block-decls-with-vars' keywords, or nil in
2156 ;; languages without such constructs.
2157 t (and (c-lang-const c-block-decls-with-vars)
2158 (c-make-keywords-re t (c-lang-const c-block-decls-with-vars))))
2159 (c-lang-defvar c-opt-block-decls-with-vars-key
2160 (c-lang-const c-opt-block-decls-with-vars-key))
2162 (c-lang-defconst c-postfix-decl-spec-kwds
2163 "Keywords introducing extra declaration specifiers in the region
2164 between the header and the body \(i.e. the \"K&R-region\") in
2167 java '("extends" "implements" "throws")
2168 idl '("context" "getraises" "manages" "primarykey" "raises" "setraises"
2171 "as" "const" "implements" "of" "ref"))
2173 (c-lang-defconst c-postfix-decl-spec-key
2174 ;; Regexp matching the keywords in `c-postfix-decl-spec-kwds'.
2175 t (c-make-keywords-re t (c-lang-const c-postfix-decl-spec-kwds)))
2176 (c-lang-defvar c-postfix-decl-spec-key
2177 (c-lang-const c-postfix-decl-spec-key))
2179 (c-lang-defconst c-nonsymbol-sexp-kwds
2180 "Keywords that may be followed by a nonsymbol sexp before whatever
2181 construct it's part of continues."
2183 (c c++ objc) '("extern"))
2185 (c-lang-defconst c-type-list-kwds
2186 "Keywords that may be followed by a comma separated list of type
2187 identifiers, where each optionally can be prefixed by keywords. (Can
2188 also be used for the special case when the list can contain only one
2191 Assumed to be mutually exclusive with `c-ref-list-kwds'. There's no
2192 reason to put keywords on this list if they are on `c-type-prefix-kwds'.
2193 There's also no reason to add keywords that prefixes a normal
2194 declaration consisting of a type followed by a declarator (list), so
2195 the keywords on `c-modifier-kwds' should normally not be listed here
2198 Note: Use `c-typeless-decl-kwds' for keywords followed by a function
2199 or variable identifier (that's being defined)."
2203 java '("import" "new" "extends" "super" "implements" "throws")
2204 idl '("manages" "native" "primarykey" "supports"
2206 "as" "implements" "of" "scope")
2209 (c-lang-defconst c-ref-list-kwds
2210 "Keywords that may be followed by a comma separated list of
2211 reference (i.e. namespace/scope/module) identifiers, where each
2212 optionally can be prefixed by keywords. (Can also be used for the
2213 special case when the list can contain only one element.) Assumed to
2214 be mutually exclusive with `c-type-list-kwds'.
2216 Note: Use `c-typeless-decl-kwds' for keywords followed by a function
2217 or variable identifier (that's being defined)."
2221 idl '("import" "module"
2226 (c-lang-defconst c-colon-type-list-kwds
2227 "Keywords that may be followed (not necessarily directly) by a colon
2228 and then a comma separated list of type identifiers, where each
2229 optionally can be prefixed by keywords. (Can also be used for the
2230 special case when the list can contain only one element.)"
2232 c++ '("class" "struct")
2233 idl '("component" "eventtype" "home" "interface" "valuetype"
2235 "storagehome" "storagetype"))
2237 (c-lang-defconst c-colon-type-list-re
2238 "Regexp matched after the keywords in `c-colon-type-list-kwds' to skip
2239 forward to the colon. The end of the match is assumed to be directly
2240 after the colon, so the regexp should end with \":\". Must be a
2241 regexp if `c-colon-type-list-kwds' isn't nil."
2242 t (if (c-lang-const c-colon-type-list-kwds)
2243 ;; Disallow various common punctuation chars that can't come
2244 ;; before the ":" that starts the inherit list after "class"
2245 ;; or "struct" in C++. (Also used as default for other
2247 "[^][{}();,/#=:]*:"))
2248 (c-lang-defvar c-colon-type-list-re (c-lang-const c-colon-type-list-re))
2250 (c-lang-defconst c-paren-nontype-kwds
2251 "Keywords that may be followed by a parenthesis expression that doesn't
2252 contain type identifiers."
2254 (c c++) '(;; GCC extension.
2259 (c-lang-defconst c-paren-type-kwds
2260 "Keywords that may be followed by a parenthesis expression containing
2261 type identifiers separated by arbitrary tokens."
2266 pike '("array" "function" "int" "mapping" "multiset" "object" "program"))
2268 (c-lang-defconst c-paren-any-kwds
2269 t (c--delete-duplicates (append (c-lang-const c-paren-nontype-kwds)
2270 (c-lang-const c-paren-type-kwds))
2271 :test 'string-equal))
2273 (c-lang-defconst c-<>-type-kwds
2274 "Keywords that may be followed by an angle bracket expression
2275 containing type identifiers separated by \",\". The difference from
2276 `c-<>-arglist-kwds' is that unknown names are taken to be types and
2277 not other identifiers. `c-recognize-<>-arglists' is assumed to be set
2285 (c-lang-defconst c-<>-arglist-kwds
2286 "Keywords that can be followed by a C++ style template arglist; see
2287 `c-recognize-<>-arglists' for details. That language constant is
2288 assumed to be set if this isn't nil."
2291 idl '("fixed" "string" "wstring"))
2293 (c-lang-defconst c-<>-sexp-kwds
2294 ;; All keywords that can be followed by an angle bracket sexp.
2295 t (c--delete-duplicates (append (c-lang-const c-<>-type-kwds)
2296 (c-lang-const c-<>-arglist-kwds))
2297 :test 'string-equal))
2299 (c-lang-defconst c-opt-<>-sexp-key
2300 ;; Adorned regexp matching keywords that can be followed by an angle
2301 ;; bracket sexp. Always set when `c-recognize-<>-arglists' is.
2302 t (c-make-keywords-re t (c-lang-const c-<>-sexp-kwds)))
2303 (c-lang-defvar c-opt-<>-sexp-key (c-lang-const c-opt-<>-sexp-key))
2305 (c-lang-defconst c-brace-id-list-kwds
2306 "Keywords that may be followed by a brace block containing a comma
2307 separated list of identifier definitions, i.e. like the list of
2308 identifiers that follows the type in a normal declaration."
2309 t (c-lang-const c-brace-list-decl-kwds))
2311 (c-lang-defconst c-block-stmt-1-kwds
2312 "Statement keywords followed directly by a substatement."
2314 c++ '("do" "else" "try")
2315 objc '("do" "else" "@finally" "@try")
2316 java '("do" "else" "finally" "try")
2319 (c-lang-defconst c-block-stmt-1-key
2320 ;; Regexp matching the start of any statement followed directly by a
2321 ;; substatement (doesn't match a bare block, however).
2322 t (c-make-keywords-re t (c-lang-const c-block-stmt-1-kwds)))
2323 (c-lang-defvar c-block-stmt-1-key (c-lang-const c-block-stmt-1-key))
2325 (c-lang-defconst c-block-stmt-1-2-kwds
2326 "Statement keywords optionally followed by a paren sexp.
2327 Keywords here should also be in `c-block-stmt-1-kwds'."
2331 (c-lang-defconst c-block-stmt-1-2-key
2332 ;; Regexp matching the start of a statement which may be followed by a
2333 ;; paren sexp and will then be followed by a substatement.
2334 t (c-make-keywords-re t (c-lang-const c-block-stmt-1-2-kwds)))
2335 (c-lang-defvar c-block-stmt-1-2-key (c-lang-const c-block-stmt-1-2-key))
2337 (c-lang-defconst c-block-stmt-2-kwds
2338 "Statement keywords followed by a paren sexp and then by a substatement."
2339 t '("for" "if" "switch" "while")
2340 c++ '("for" "if" "switch" "while" "catch")
2341 objc '("for" "if" "switch" "while" "@catch" "@synchronized")
2342 java '("for" "if" "switch" "while" "catch" "synchronized")
2344 pike '("for" "if" "switch" "while" "foreach")
2345 awk '("for" "if" "while"))
2347 (c-lang-defconst c-block-stmt-2-key
2348 ;; Regexp matching the start of any statement followed by a paren sexp
2349 ;; and then by a substatement.
2350 t (c-make-keywords-re t (c-lang-const c-block-stmt-2-kwds)))
2351 (c-lang-defvar c-block-stmt-2-key (c-lang-const c-block-stmt-2-key))
2353 (c-lang-defconst c-block-stmt-kwds
2354 ;; Union of `c-block-stmt-1-kwds' and `c-block-stmt-2-kwds'.
2355 t (c--delete-duplicates (append (c-lang-const c-block-stmt-1-kwds)
2356 (c-lang-const c-block-stmt-2-kwds))
2357 :test 'string-equal))
2359 (c-lang-defconst c-opt-block-stmt-key
2360 ;; Regexp matching the start of any statement that has a
2361 ;; substatement (except a bare block). Nil in languages that
2362 ;; don't have such constructs.
2363 t (if (or (c-lang-const c-block-stmt-1-kwds)
2364 (c-lang-const c-block-stmt-2-kwds))
2365 (c-make-keywords-re t
2366 (append (c-lang-const c-block-stmt-1-kwds)
2367 (c-lang-const c-block-stmt-2-kwds)))))
2368 (c-lang-defvar c-opt-block-stmt-key (c-lang-const c-opt-block-stmt-key))
2370 (c-lang-defconst c-simple-stmt-kwds
2371 "Statement keywords followed by an expression or nothing."
2372 t '("break" "continue" "goto" "return")
2373 objc '("break" "continue" "goto" "return" "@throw")
2374 ;; Note: `goto' is not valid in Java, but the keyword is still reserved.
2375 java '("break" "continue" "goto" "return" "throw")
2377 pike '("break" "continue" "return")
2378 awk '(;; Not sure about "delete", "exit", "getline", etc. ; ACM 2002/5/30
2379 "break" "continue" "return" "delete" "exit" "getline" "next"
2380 "nextfile" "print" "printf"))
2382 (c-lang-defconst c-simple-stmt-key
2383 ;; Adorned regexp matching `c-simple-stmt-kwds'.
2384 t (c-make-keywords-re t (c-lang-const c-simple-stmt-kwds)))
2385 (c-lang-defvar c-simple-stmt-key (c-lang-const c-simple-stmt-key))
2387 (c-lang-defconst c-paren-stmt-kwds
2388 "Statement keywords followed by a parenthesis expression that
2389 nevertheless contains a list separated with `;' and not `,'."
2393 (c-lang-defconst c-paren-stmt-key
2394 ;; Adorned regexp matching `c-paren-stmt-kwds'.
2395 t (c-make-keywords-re t (c-lang-const c-paren-stmt-kwds)))
2396 (c-lang-defvar c-paren-stmt-key (c-lang-const c-paren-stmt-key))
2398 (c-lang-defconst c-asm-stmt-kwds
2399 "Statement keywords followed by an assembler expression."
2401 (c c++) '("asm" "__asm__")) ;; Not standard, but common.
2403 (c-lang-defconst c-opt-asm-stmt-key
2404 ;; Regexp matching the start of an assembler statement. Nil in
2405 ;; languages that don't support that.
2406 t (if (c-lang-const c-asm-stmt-kwds)
2407 (c-make-keywords-re t (c-lang-const c-asm-stmt-kwds))))
2408 (c-lang-defvar c-opt-asm-stmt-key (c-lang-const c-opt-asm-stmt-key))
2410 (c-lang-defconst c-case-kwds
2411 "The keyword(s) which introduce a \"case\" like construct.
2412 This construct is \"<keyword> <expression> :\"."
2416 (c-lang-defconst c-case-kwds-regexp
2417 ;; Adorned regexp matching any "case"-like keyword.
2418 t (c-make-keywords-re t (c-lang-const c-case-kwds)))
2419 (c-lang-defvar c-case-kwds-regexp (c-lang-const c-case-kwds-regexp))
2421 (c-lang-defconst c-label-kwds
2422 "Keywords introducing colon terminated labels in blocks."
2423 t '("case" "default"))
2425 (c-lang-defconst c-label-kwds-regexp
2426 ;; Adorned regexp matching any keyword that introduces a label.
2427 t (c-make-keywords-re t (c-lang-const c-label-kwds)))
2428 (c-lang-defvar c-label-kwds-regexp (c-lang-const c-label-kwds-regexp))
2430 (c-lang-defconst c-before-label-kwds
2431 "Keywords that might be followed by a label identifier."
2433 (java pike) (append '("break" "continue")
2434 (c-lang-const c-before-label-kwds))
2438 (c-lang-defconst c-constant-kwds
2439 "Keywords for constants."
2441 c '("NULL" ;; Not a keyword, but practically works as one.
2442 "false" "true") ; Defined in C99.
2445 (c-lang-const c-constant-kwds c))
2446 objc '("nil" "Nil" "YES" "NO" "NS_DURING" "NS_HANDLER" "NS_ENDHANDLER")
2447 idl '("TRUE" "FALSE")
2448 java '("true" "false" "null") ; technically "literals", not keywords
2449 pike '("UNDEFINED")) ;; Not a keyword, but practically works as one.
2451 (c-lang-defconst c-primary-expr-kwds
2452 "Keywords besides constants and operators that start primary expressions."
2454 c++ '("operator" "this")
2455 objc '("super" "self")
2457 pike '("this")) ;; Not really a keyword, but practically works as one.
2459 (c-lang-defconst c-expr-kwds
2460 ;; Keywords that can occur anywhere in expressions. Built from
2461 ;; `c-primary-expr-kwds' and all keyword operators in `c-operators'.
2462 t (c--delete-duplicates
2463 (append (c-lang-const c-primary-expr-kwds)
2464 (c-filter-ops (c-lang-const c-operator-list)
2466 "\\`\\(\\w\\|\\s_\\)+\\'"))
2467 :test 'string-equal))
2469 (c-lang-defconst c-lambda-kwds
2470 "Keywords that start lambda constructs, i.e. function definitions in
2475 (c-lang-defconst c-inexpr-block-kwds
2476 "Keywords that start constructs followed by statement blocks which can
2477 be used in expressions \(the gcc extension for this in C and C++ is
2478 handled separately by `c-recognize-paren-inexpr-blocks')."
2480 pike '("catch" "gauge"))
2482 (c-lang-defconst c-inexpr-class-kwds
2483 "Keywords that can start classes inside expressions."
2488 (c-lang-defconst c-inexpr-brace-list-kwds
2489 "Keywords that can start brace list blocks inside expressions.
2490 Note that Java specific rules are currently applied to tell this from
2491 `c-inexpr-class-kwds'."
2495 (c-lang-defconst c-opt-inexpr-brace-list-key
2496 ;; Regexp matching the start of a brace list in an expression, or
2497 ;; nil in languages that don't have such things. This should not
2498 ;; match brace lists recognized through `c-special-brace-lists'.
2499 t (and (c-lang-const c-inexpr-brace-list-kwds)
2500 (c-make-keywords-re t (c-lang-const c-inexpr-brace-list-kwds))))
2501 (c-lang-defvar c-opt-inexpr-brace-list-key
2502 (c-lang-const c-opt-inexpr-brace-list-key))
2504 (c-lang-defconst c-decl-block-key
2505 ;; Regexp matching keywords in any construct that contain another
2506 ;; declaration level, i.e. that isn't followed by a function block
2507 ;; or brace list. When the first submatch matches, it's an
2508 ;; unambiguous construct, otherwise it's an ambiguous match that
2509 ;; might also be the return type of a function declaration.
2510 t (let* ((decl-kwds (append (c-lang-const c-class-decl-kwds)
2511 (c-lang-const c-other-block-decl-kwds)
2512 (c-lang-const c-inexpr-class-kwds)))
2513 (unambiguous (c--set-difference decl-kwds
2514 (c-lang-const c-type-start-kwds)
2515 :test 'string-equal))
2516 (ambiguous (c--intersection decl-kwds
2517 (c-lang-const c-type-start-kwds)
2518 :test 'string-equal)))
2520 (concat (c-make-keywords-re t unambiguous)
2522 (c-make-keywords-re t ambiguous))
2523 (c-make-keywords-re t unambiguous))))
2524 (c-lang-defvar c-decl-block-key (c-lang-const c-decl-block-key))
2526 (c-lang-defconst c-bitfield-kwds
2527 "Keywords that can introduce bitfields."
2529 (c c++ objc) '("char" "int" "long" "signed" "unsigned"))
2531 (c-lang-defconst c-opt-bitfield-key
2532 ;; Regexp matching the start of a bitfield (not uniquely), or nil in
2533 ;; languages without bitfield support.
2535 (c c++) (c-make-keywords-re t (c-lang-const c-bitfield-kwds)))
2536 (c-lang-defvar c-opt-bitfield-key (c-lang-const c-opt-bitfield-key))
2538 (c-lang-defconst c-other-kwds
2539 "Keywords not accounted for by any other `*-kwds' language constant."
2542 ;; In CORBA CIDL: (These are declaration keywords that never
2543 ;; can start a declaration.)
2544 "entity" "process" "service" "session" "storage"))
2547 ;;; Constants built from keywords.
2549 ;; Note: No `*-kwds' language constants may be defined below this point.
2552 (defconst c-kwds-lang-consts
2553 ;; List of all the language constants that contain keyword lists.
2555 (mapatoms (lambda (sym)
2556 (when (and (boundp sym)
2557 (string-match "-kwds\\'" (symbol-name sym)))
2558 ;; Make the list of globally interned symbols
2559 ;; instead of ones interned in `c-lang-constants'.
2560 (setq list (cons (intern (symbol-name sym)) list))))
2564 (c-lang-defconst c-keywords
2565 ;; All keywords as a list.
2566 t (c--delete-duplicates
2567 (c-lang-defconst-eval-immediately
2568 `(append ,@(mapcar (lambda (kwds-lang-const)
2569 `(c-lang-const ,kwds-lang-const))
2572 :test 'string-equal))
2574 (c-lang-defconst c-keywords-regexp
2575 ;; All keywords as an adorned regexp.
2576 t (c-make-keywords-re t (c-lang-const c-keywords)))
2577 (c-lang-defvar c-keywords-regexp (c-lang-const c-keywords-regexp))
2579 (c-lang-defconst c-keyword-member-alist
2580 ;; An alist with all the keywords in the cars. The cdr for each
2581 ;; keyword is a list of the symbols for the `*-kwds' lists that
2583 t (let ((kwd-list-alist
2584 (c-lang-defconst-eval-immediately
2585 `(list ,@(mapcar (lambda (kwds-lang-const)
2586 `(cons ',kwds-lang-const
2587 (c-lang-const ,kwds-lang-const)))
2588 c-kwds-lang-consts))))
2589 lang-const kwd-list kwd
2591 (while kwd-list-alist
2592 (setq lang-const (caar kwd-list-alist)
2593 kwd-list (cdar kwd-list-alist)
2594 kwd-list-alist (cdr kwd-list-alist))
2596 (setq kwd (car kwd-list)
2597 kwd-list (cdr kwd-list))
2598 (unless (setq elem (assoc kwd result-alist))
2599 (setq result-alist (cons (setq elem (list kwd)) result-alist)))
2600 (unless (memq lang-const (cdr elem))
2601 (setcdr elem (cons lang-const (cdr elem))))))
2604 (c-lang-defvar c-keywords-obarray
2605 ;; An obarray containing all keywords as symbols. The property list
2606 ;; of each symbol has a non-nil entry for the specific `*-kwds'
2607 ;; lists it's a member of.
2609 ;; E.g. to see whether the string str contains a keyword on
2610 ;; `c-class-decl-kwds', one can do like this:
2611 ;; (get (intern-soft str c-keyword-obarray) 'c-class-decl-kwds)
2612 ;; Which preferably is written using the associated functions in
2614 ;; (c-keyword-member (c-keyword-sym str) 'c-class-decl-kwds)
2616 ;; The obarray is not stored directly as a language constant since
2617 ;; the printed representation for obarrays used in .elc files isn't
2620 (let* ((alist (c-lang-const c-keyword-member-alist))
2622 (obarray (make-vector (* (length alist) 2) 0)))
2624 (setq kwd (caar alist)
2625 lang-const-list (cdar alist)
2627 (setplist (intern kwd obarray)
2628 ;; Emacs has an odd bug that causes `mapcan' to fail
2629 ;; with unintelligible errors. (XEmacs works.)
2630 ;; (2015-06-24): This bug has not yet been fixed.
2631 ;;(mapcan (lambda (lang-const)
2632 ;; (list lang-const t))
2634 (apply 'nconc (mapcar (lambda (lang-const)
2635 (list lang-const t))
2639 (c-lang-defconst c-regular-keywords-regexp
2640 ;; Adorned regexp matching all keywords that should be fontified
2641 ;; with the keywords face. I.e. that aren't types or constants.
2642 t (c-make-keywords-re t
2643 (c--set-difference (c-lang-const c-keywords)
2644 (append (c-lang-const c-primitive-type-kwds)
2645 (c-lang-const c-constant-kwds))
2646 :test 'string-equal)))
2647 (c-lang-defvar c-regular-keywords-regexp
2648 (c-lang-const c-regular-keywords-regexp))
2650 (c-lang-defconst c-primary-expr-regexp
2651 ;; Regexp matching the start of any primary expression, i.e. any
2652 ;; literal, symbol, prefix operator, and '('. It doesn't need to
2653 ;; exclude keywords; they are excluded afterwards unless the second
2654 ;; submatch matches. If the first but not the second submatch
2655 ;; matches then it is an ambiguous primary expression; it could also
2656 ;; be a match of e.g. an infix operator. (The case with ambiguous
2657 ;; keyword operators isn't handled.)
2659 t (let* ((prefix-ops
2660 (c-filter-ops (c-lang-const c-operators)
2663 ;; Filter out the special case prefix
2664 ;; operators that are close parens.
2665 (not (string-match "\\s)" op)))))
2667 (nonkeyword-prefix-ops
2668 (c-filter-ops prefix-ops
2670 "\\`\\(\\s.\\|\\s(\\|\\s)\\)+\\'"))
2673 (c-filter-ops (c-lang-const c-operators)
2678 right-assoc-sequence)
2681 (unambiguous-prefix-ops (c--set-difference nonkeyword-prefix-ops
2683 :test 'string-equal))
2684 (ambiguous-prefix-ops (c--intersection nonkeyword-prefix-ops
2686 :test 'string-equal)))
2690 ;; Take out all symbol class operators from `prefix-ops' and make the
2691 ;; first submatch from them together with `c-primary-expr-kwds'.
2692 (c-make-keywords-re t
2693 (append (c-lang-const c-primary-expr-kwds)
2694 (c--set-difference prefix-ops nonkeyword-prefix-ops
2695 :test 'string-equal)))
2698 ;; Match all ambiguous operators.
2699 (c-make-keywords-re nil
2700 (c--intersection nonkeyword-prefix-ops in-or-postfix-ops
2701 :test 'string-equal))
2705 ;; Now match all other symbols.
2706 (c-lang-const c-symbol-start)
2709 ;; The chars that can start integer and floating point
2714 ;; The unambiguous operators from `prefix-ops'.
2715 (c-make-keywords-re nil
2716 (c--set-difference nonkeyword-prefix-ops in-or-postfix-ops
2717 :test 'string-equal))
2720 ;; Match string and character literals.
2722 (if (memq 'gen-string-delim c-emacs-features)
2725 (c-lang-defvar c-primary-expr-regexp (c-lang-const c-primary-expr-regexp))
2728 ;;; Additional constants for parser-level constructs.
2730 (c-lang-defconst c-decl-start-colon-kwd-re
2731 "Regexp matching a keyword that is followed by a colon, where
2732 the whole construct can precede a declaration.
2733 E.g. \"public:\" in C++."
2735 c++ (c-make-keywords-re t (c-lang-const c-protection-kwds)))
2736 (c-lang-defvar c-decl-start-colon-kwd-re
2737 (c-lang-const c-decl-start-colon-kwd-re))
2739 (c-lang-defconst c-decl-prefix-re
2740 "Regexp matching something that might precede a declaration, cast or
2741 label, such as the last token of a preceding statement or declaration.
2742 This is used in the common situation where a declaration or cast
2743 doesn't start with any specific token that can be searched for.
2745 The regexp should not match bob; that is done implicitly. It can't
2746 require a match longer than one token. The end of the token is taken
2747 to be at the end of the first submatch, which is assumed to always
2748 match. It's undefined whether identifier syntax (see
2749 `c-identifier-syntax-table') is in effect or not. This regexp is
2750 assumed to be a superset of `c-label-prefix-re' if
2751 `c-recognize-colon-labels' is set.
2753 Besides this, `c-decl-start-kwds' is used to find declarations.
2755 Note: This variable together with `c-decl-start-re' and
2756 `c-decl-start-kwds' is only used to detect \"likely\"
2757 declaration/cast/label starts. I.e. they might produce more matches
2758 but should not miss anything (or else it's necessary to use text
2759 properties - see the next note). Wherever they match, the following
2760 construct is analyzed to see if it indeed is a declaration, cast or
2761 label. That analysis is not cheap, so it's important that not too
2762 many false matches are triggered.
2764 Note: If a declaration/cast/label start can't be detected with this
2765 variable, it's necessary to use the `c-type' text property with the
2766 value `c-decl-end' on the last char of the last token preceding the
2767 declaration. See the comment blurb at the start of cc-engine.el for
2770 ;; We match a sequence of characters to skip over things like \"};\"
2771 ;; more quickly. We match ")" in C for K&R region declarations, and
2772 ;; in all languages except Java for when a cpp macro definition
2773 ;; begins with a declaration.
2775 java "\\([{}(;,<]+\\)"
2776 ;; Match "<" in C++ to get the first argument in a template arglist.
2777 ;; In that case there's an additional check in `c-find-decl-spots'
2778 ;; that it got open paren syntax. Match ":" to aid in picking up
2779 ;; "public:", etc. This involves additional checks in
2780 ;; `c-find-decl-prefix-search' to prevent a match of identifiers
2782 c++ "\\([{}();:,<]+\\)"
2783 ;; Additionally match the protection directives in Objective-C.
2784 ;; Note that this doesn't cope with the longer directives, which we
2785 ;; would have to match from start to end since they don't end with
2786 ;; any easily recognized characters.
2787 objc (concat "\\([{}();,]+\\|"
2788 (c-make-keywords-re nil (c-lang-const c-protection-kwds))
2790 ;; Pike is like C but we also match "[" for multiple value
2791 ;; assignments and type casts.
2792 pike "\\([{}()[;,]+\\)")
2793 (c-lang-defvar c-decl-prefix-re (c-lang-const c-decl-prefix-re)
2796 (c-lang-defconst c-decl-start-re
2797 "Regexp matching the start of any declaration, cast or label.
2798 It's used on the token after the one `c-decl-prefix-re' matched. This
2799 regexp should not try to match those constructs accurately as it's
2800 only used as a sieve to avoid spending more time checking other
2802 t (c-lang-const c-identifier-start))
2803 (c-lang-defvar c-decl-start-re (c-lang-const c-decl-start-re))
2805 (c-lang-defconst c-decl-prefix-or-start-re
2806 ;; Regexp matching something that might precede or start a
2807 ;; declaration, cast or label.
2809 ;; If the first submatch matches, it's taken to match the end of a
2810 ;; token that might precede such a construct, e.g. ';', '}' or '{'.
2811 ;; It's built from `c-decl-prefix-re'.
2813 ;; If the first submatch did not match, the match of the whole
2814 ;; regexp is taken to be at the first token in the declaration.
2815 ;; `c-decl-start-re' is not checked in this case.
2817 ;; Design note: The reason the same regexp is used to match both
2818 ;; tokens that precede declarations and start them is to avoid an
2819 ;; extra regexp search from the previous declaration spot in
2820 ;; `c-find-decl-spots'. Users of `c-find-decl-spots' also count on
2821 ;; that it finds all declaration/cast/label starts in approximately
2822 ;; linear order, so we can't do the searches in two separate passes.
2823 t (if (c-lang-const c-decl-start-kwds)
2824 (concat (c-lang-const c-decl-prefix-re)
2826 (c-make-keywords-re t (c-lang-const c-decl-start-kwds)))
2827 (c-lang-const c-decl-prefix-re)))
2828 (c-lang-defvar c-decl-prefix-or-start-re
2829 (c-lang-const c-decl-prefix-or-start-re)
2832 (c-lang-defconst c-cast-parens
2833 ;; List containing the paren characters that can open a cast, or nil in
2834 ;; languages without casts.
2835 t (c-filter-ops (c-lang-const c-operators)
2838 (lambda (op) (elt op 0))))
2839 (c-lang-defvar c-cast-parens (c-lang-const c-cast-parens))
2841 (c-lang-defconst c-block-prefix-disallowed-chars
2842 "List of syntactically relevant characters that never can occur before
2843 the open brace in any construct that contains a brace block, e.g. in
2844 the \"class Foo: public Bar\" part of:
2846 class Foo: public Bar {int x();} a, *b;
2848 If parens can occur, the chars inside those aren't filtered with this
2851 `<' and `>' should be disallowed even if angle bracket arglists can
2852 occur. That since the search function needs to stop at them anyway to
2853 ensure they are given paren syntax.
2855 This is used to skip backward from the open brace to find the region
2856 in which to look for a construct like \"class\", \"enum\",
2857 \"namespace\" or whatever. That skipping should be as tight as
2858 possible for good performance."
2860 ;; Default to all chars that only occurs in nonsymbol tokens outside
2862 t (c--set-difference
2863 (c-lang-const c-nonsymbol-token-char-list)
2864 (c-filter-ops (append (c-lang-const c-identifier-ops)
2866 (c-lang-const c-after-id-concat-ops))))
2871 (while (string-match "\\(\\s.\\|\\s(\\|\\s)\\)"
2873 (setq res (cons (aref op (match-beginning 1)) res)
2877 ;; Allow cpp operations (where applicable).
2878 t (if (c-lang-const c-opt-cpp-prefix)
2879 (c--set-difference (c-lang-const c-block-prefix-disallowed-chars)
2881 (c-lang-const c-block-prefix-disallowed-chars))
2883 ;; Allow ':' for inherit list starters.
2884 (c++ objc idl) (c--set-difference (c-lang-const c-block-prefix-disallowed-chars)
2887 ;; Allow ',' for multiple inherits.
2888 (c++ java) (c--set-difference (c-lang-const c-block-prefix-disallowed-chars)
2891 ;; Allow parentheses for anonymous inner classes in Java and class
2892 ;; initializer lists in Pike.
2893 (java pike) (c--set-difference (c-lang-const c-block-prefix-disallowed-chars)
2896 ;; Allow '"' for extern clauses (e.g. extern "C" {...}).
2897 (c c++ objc) (c--set-difference (c-lang-const c-block-prefix-disallowed-chars)
2900 (c-lang-defconst c-block-prefix-charset
2901 ;; `c-block-prefix-disallowed-chars' as an inverted charset suitable
2902 ;; for `c-syntactic-skip-backward'.
2903 t (c-make-bare-char-alt (c-lang-const c-block-prefix-disallowed-chars) t))
2904 (c-lang-defvar c-block-prefix-charset (c-lang-const c-block-prefix-charset))
2906 (c-lang-defconst c-type-decl-prefix-key
2907 "Regexp matching any declarator operator that might precede the
2908 identifier in a declaration, e.g. the \"*\" in \"char *argv\". This
2909 regexp should match \"(\" if parentheses are valid in declarators.
2910 The end of the first submatch is taken as the end of the operator.
2911 Identifier syntax is in effect when this is matched \(see
2912 `c-identifier-syntax-table')."
2913 t (if (c-lang-const c-type-modifier-kwds)
2914 (concat (regexp-opt (c-lang-const c-type-modifier-kwds) t) "\\>")
2915 ;; Default to a regexp that never matches.
2917 ;; Check that there's no "=" afterwards to avoid matching tokens
2919 (c objc) (concat "\\("
2922 (c-lang-const c-type-decl-prefix-key)
2928 (c-lang-const c-type-decl-prefix-key)
2931 ;; If this matches there's special treatment in
2932 ;; `c-font-lock-declarators' and
2933 ;; `c-font-lock-declarations' that check for a
2934 ;; complete name followed by ":: *".
2935 (c-lang-const c-identifier-start)
2939 pike "\\(\\*\\)\\([^=]\\|$\\)")
2940 (c-lang-defvar c-type-decl-prefix-key (c-lang-const c-type-decl-prefix-key)
2943 (c-lang-defconst c-type-decl-suffix-key
2944 "Regexp matching the declarator operators that might follow after the
2945 identifier in a declaration, e.g. the \"[\" in \"char argv[]\". This
2946 regexp should match \")\" if parentheses are valid in declarators. If
2947 it matches an open paren of some kind, the type declaration check
2948 continues at the corresponding close paren, otherwise the end of the
2949 first submatch is taken as the end of the operator. Identifier syntax
2950 is in effect when this is matched (see `c-identifier-syntax-table')."
2951 ;; Default to a regexp that matches `c-type-modifier-kwds' and a
2952 ;; function argument list parenthesis.
2953 t (if (c-lang-const c-type-modifier-kwds)
2955 (regexp-opt (c-lang-const c-type-modifier-kwds) t) "\\>"
2958 (c c++ objc) (concat
2961 (if (c-lang-const c-type-modifier-kwds)
2964 ;; "throw" in `c-type-modifier-kwds' is followed
2965 ;; by a parenthesis list, but no extra measures
2966 ;; are necessary to handle that.
2967 (regexp-opt (c-lang-const c-type-modifier-kwds) t)
2973 (c-lang-defvar c-type-decl-suffix-key (c-lang-const c-type-decl-suffix-key)
2976 (c-lang-defconst c-after-suffixed-type-decl-key
2977 "This regexp is matched after a declarator expression where
2978 `c-type-decl-suffix-key' has matched. If it matches then the
2979 construct is taken as a declaration. It's typically used to match the
2980 beginning of a function body or whatever might occur after the
2981 function header in a function declaration or definition. It's
2982 undefined whether identifier syntax (see `c-identifier-syntax-table')
2983 is in effect or not.
2985 Note that it's used in cases like after \"foo (bar)\" so it should
2986 only match when it's certain that it's a declaration, e.g., \"{\" but
2987 not \",\" or \";\"."
2989 ;; If K&R style declarations should be recognized then one could
2990 ;; consider to match the start of any symbol since we want to match
2991 ;; the start of the first declaration in the "K&R region". That
2992 ;; could however produce false matches on code like "FOO(bar) x"
2993 ;; where FOO is a cpp macro, so it's better to leave it out and rely
2994 ;; on the other heuristics in that case.
2995 t (if (c-lang-const c-postfix-spec-kwds)
2996 ;; Add on the keywords in `c-postfix-spec-kwds'.
2997 (concat (c-lang-const c-after-suffixed-type-decl-key)
2999 (c-make-keywords-re t (c-lang-const c-postfix-spec-kwds)))
3000 (c-lang-const c-after-suffixed-type-decl-key))
3001 ;; Also match the colon that starts a base class initializer list in
3002 ;; C++. That can be confused with a function call before the colon
3003 ;; in a ? : operator, but we count on that `c-decl-prefix-re' won't
3004 ;; match before such a thing (as a declaration-level construct;
3005 ;; matches inside arglist contexts are already excluded).
3007 (c-lang-defvar c-after-suffixed-type-decl-key
3008 (c-lang-const c-after-suffixed-type-decl-key)
3011 (c-lang-defconst c-after-suffixed-type-maybe-decl-key
3012 ;; Regexp that in addition to `c-after-suffixed-type-decl-key'
3013 ;; matches ";" and ",".
3014 t (concat "\\(" (c-lang-const c-after-suffixed-type-decl-key) "\\)"
3016 (c-lang-defvar c-after-suffixed-type-maybe-decl-key
3017 (c-lang-const c-after-suffixed-type-maybe-decl-key))
3019 (c-lang-defconst c-opt-type-concat-key
3020 "Regexp matching operators that concatenate types, e.g. the \"|\" in
3021 \"int|string\" in Pike. The end of the first submatch is taken as the
3022 end of the operator. nil in languages without such operators. It's
3023 undefined whether identifier syntax (see `c-identifier-syntax-table')
3024 is in effect or not."
3026 pike "\\([|.&]\\)\\($\\|[^|.&]\\)")
3027 (c-lang-defvar c-opt-type-concat-key (c-lang-const c-opt-type-concat-key)
3030 (c-lang-defconst c-opt-type-suffix-key
3031 "Regexp matching operators that might follow after a type, or nil in
3032 languages that don't have such operators. The end of the first
3033 submatch is taken as the end of the operator. This should not match
3034 things like C++ template arglists if `c-recognize-<>-arglists' is set.
3035 It's undefined whether identifier syntax (see `c-identifier-syntax-table')
3036 is in effect or not."
3038 (c c++ objc pike) "\\(\\.\\.\\.\\)"
3039 java (concat "\\(\\[" (c-lang-const c-simple-ws) "*\\]\\|\\.\\.\\.\\)"))
3040 (c-lang-defvar c-opt-type-suffix-key (c-lang-const c-opt-type-suffix-key))
3042 (c-lang-defvar c-known-type-key
3043 ;; Regexp matching the known type identifiers. This is initialized
3044 ;; from the type keywords and `*-font-lock-extra-types'. The first
3045 ;; submatch is the one that matches the type. Note that this regexp
3046 ;; assumes that symbol constituents like '_' and '$' have word
3049 (when (boundp (c-mode-symbol "font-lock-extra-types"))
3050 (c-mode-var "font-lock-extra-types")))
3052 (delq nil (mapcar (lambda (re)
3053 (when (string-match "[][.*+?^$\\]" re)
3057 (delq nil (mapcar (lambda (re)
3058 (unless (string-match "[][.*+?^$\\]" re)
3063 (append (list (c-make-keywords-re nil
3064 (append (c-lang-const c-primitive-type-kwds)
3070 (c-lang-defconst c-special-brace-lists
3071 "List of open- and close-chars that makes up a pike-style brace list,
3072 i.e. for a ([ ]) list there should be a cons (?\\[ . ?\\]) in this
3075 pike '((?{ . ?}) (?\[ . ?\]) (?< . ?>)))
3076 (c-lang-defvar c-special-brace-lists (c-lang-const c-special-brace-lists))
3078 (c-lang-defconst c-recognize-knr-p
3079 "Non-nil means K&R style argument declarations are valid."
3082 (c-lang-defvar c-recognize-knr-p (c-lang-const c-recognize-knr-p))
3084 (c-lang-defconst c-recognize-typeless-decls
3085 "Non-nil means function declarations without return type should be
3086 recognized. That can introduce an ambiguity with parenthesized macro
3087 calls before a brace block. This setting does not affect declarations
3088 that are preceded by a declaration starting keyword, so
3089 e.g. `c-typeless-decl-kwds' may still be used when it's set to nil."
3091 (c c++ objc java) t)
3092 (c-lang-defvar c-recognize-typeless-decls
3093 (c-lang-const c-recognize-typeless-decls))
3095 (c-lang-defconst c-recognize-<>-arglists
3096 "Non-nil means C++ style template arglists should be handled. More
3097 specifically, this means a comma separated list of types or
3098 expressions surrounded by \"<\" and \">\". It's always preceded by an
3099 identifier or one of the keywords on `c-<>-type-kwds' or
3100 `c-<>-arglist-kwds'. If there's an identifier before then the whole
3101 expression is considered to be a type."
3102 t (or (consp (c-lang-const c-<>-type-kwds))
3103 (consp (c-lang-const c-<>-arglist-kwds)))
3104 java t) ; 2008-10-19. This is crude. The syntax for java
3105 ; generics is not yet coded in CC Mode.
3106 (c-lang-defvar c-recognize-<>-arglists (c-lang-const c-recognize-<>-arglists))
3108 (c-lang-defconst c-<>-notable-chars-re
3109 "A regexp matching any single character notable inside a <...> construct.
3110 This must include \"<\" and \">\", and should include \",\", and
3111 any character which cannot be valid inside such a construct.
3112 This is used in `c-forward-<>-arglist-recur' to try to detect
3113 sequences of tokens which cannot be a template/generic construct.
3114 When \"(\" is present, that defun will attempt to parse a
3115 parenthesized expression inside the template. When \")\" is
3116 present it will treat an unbalanced closing paren as a sign of
3117 the invalidity of the putative template construct."
3120 (c-lang-defvar c-<>-notable-chars-re (c-lang-const c-<>-notable-chars-re))
3122 (c-lang-defconst c-enums-contain-decls
3123 "Non-nil means that an enum structure can contain declarations."
3126 (c-lang-defvar c-enums-contain-decls (c-lang-const c-enums-contain-decls))
3128 (c-lang-defconst c-recognize-paren-inits
3129 "Non-nil means that parenthesis style initializers exist,
3130 i.e. constructs like
3134 in addition to the more classic
3139 (c-lang-defvar c-recognize-paren-inits (c-lang-const c-recognize-paren-inits))
3141 (c-lang-defconst c-recognize-paren-inexpr-blocks
3142 "Non-nil to recognize gcc style in-expression blocks,
3143 i.e. compound statements surrounded by parentheses inside expressions."
3146 (c-lang-defvar c-recognize-paren-inexpr-blocks
3147 (c-lang-const c-recognize-paren-inexpr-blocks))
3149 (c-lang-defconst c-opt-<>-arglist-start
3150 ;; Regexp matching the start of angle bracket arglists in languages
3151 ;; where `c-recognize-<>-arglists' is set. Does not exclude
3152 ;; keywords outside `c-<>-arglist-kwds'. The first submatch is
3153 ;; assumed to surround the preceding symbol. The whole match is
3154 ;; assumed to end directly after the opening "<".
3155 t (if (c-lang-const c-recognize-<>-arglists)
3157 (c-lang-const c-symbol-key)
3159 (c-lang-const c-syntactic-ws)
3161 (c-lang-defvar c-opt-<>-arglist-start (c-lang-const c-opt-<>-arglist-start))
3163 (c-lang-defconst c-opt-<>-arglist-start-in-paren
3164 ;; Regexp that in addition to `c-opt-<>-arglist-start' matches close
3165 ;; parens. The first submatch is assumed to surround
3166 ;; `c-opt-<>-arglist-start'.
3167 t (if (c-lang-const c-opt-<>-arglist-start)
3169 (c-lang-const c-opt-<>-arglist-start)
3171 (c-lang-defvar c-opt-<>-arglist-start-in-paren
3172 (c-lang-const c-opt-<>-arglist-start-in-paren))
3174 (c-lang-defconst c-opt-postfix-decl-spec-key
3175 ;; Regexp matching the beginning of a declaration specifier in the
3176 ;; region between the header and the body of a declaration.
3178 ;; TODO: This is currently not used uniformly; c++-mode and
3179 ;; java-mode each have their own ways of using it.
3182 (c-lang-const c-simple-ws) "*"
3183 "\\(virtual" (c-lang-const c-simple-ws) "+\\)?\\("
3184 (c-make-keywords-re nil (c-lang-const c-protection-kwds))
3185 "\\)" (c-lang-const c-simple-ws) "+"
3186 "\\(" (c-lang-const c-symbol-key) "\\)")
3187 java (c-make-keywords-re t (c-lang-const c-postfix-spec-kwds)))
3188 (c-lang-defvar c-opt-postfix-decl-spec-key
3189 (c-lang-const c-opt-postfix-decl-spec-key))
3191 (c-lang-defconst c-recognize-colon-labels
3192 "Non-nil if generic labels ending with \":\" should be recognized.
3193 That includes labels in code and access keys in classes. This does
3194 not apply to labels recognized by `c-label-kwds' and
3195 `c-opt-extra-label-key'."
3197 (c c++ objc java pike) t)
3198 (c-lang-defvar c-recognize-colon-labels
3199 (c-lang-const c-recognize-colon-labels))
3201 (c-lang-defconst c-label-prefix-re
3202 "Regexp like `c-decl-prefix-re' that matches any token that can precede
3203 a generic colon label. Not used if `c-recognize-colon-labels' is
3206 (c-lang-defvar c-label-prefix-re
3207 (c-lang-const c-label-prefix-re))
3209 (c-lang-defconst c-nonlabel-token-key
3210 "Regexp matching things that can't occur in generic colon labels,
3211 neither in a statement nor in a declaration context. The regexp is
3212 tested at the beginning of every sexp in a suspected label,
3213 i.e. before \":\". Only used if `c-recognize-colon-labels' is set."
3215 ;; All keywords except `c-label-kwds' and `c-protection-kwds'.
3216 (c-make-keywords-re t
3217 (c--set-difference (c-lang-const c-keywords)
3218 (append (c-lang-const c-label-kwds)
3219 (c-lang-const c-protection-kwds))
3220 :test 'string-equal)))
3221 ;; Don't allow string literals, except in AWK. Character constants are OK.
3222 (c objc java pike idl) (concat "\"\\|"
3223 (c-lang-const c-nonlabel-token-key))
3224 ;; Also check for open parens in C++, to catch member init lists in
3225 ;; constructors. We normally allow it so that macros with arguments
3227 c++ (concat "\\s(\\|\"\\|" (c-lang-const c-nonlabel-token-key)))
3228 (c-lang-defvar c-nonlabel-token-key (c-lang-const c-nonlabel-token-key))
3230 (c-lang-defconst c-nonlabel-token-2-key
3231 "Regexp matching things that can't occur two symbols before a colon in
3232 a label construct. This catches C++'s inheritance construct \"class foo
3233 : bar\". Only used if `c-recognize-colon-labels' is set."
3234 t "\\<\\>" ; matches nothing
3235 c++ (c-make-keywords-re t '("class")))
3236 (c-lang-defvar c-nonlabel-token-2-key (c-lang-const c-nonlabel-token-2-key))
3238 (c-lang-defconst c-opt-extra-label-key
3239 "Optional regexp matching labels.
3240 Normally, labels are detected according to `c-nonlabel-token-key',
3241 `c-decl-prefix-re' and `c-nonlabel-decl-prefix-re'. This regexp can
3242 be used if there are additional labels that aren't recognized that
3245 objc (c-make-keywords-re t (c-lang-const c-protection-kwds)))
3246 (c-lang-defvar c-opt-extra-label-key (c-lang-const c-opt-extra-label-key))
3248 (c-lang-defconst c-opt-friend-key
3249 ;; Regexp describing friend declarations classes, or nil in
3250 ;; languages that don't have such things.
3252 ;; TODO: Ought to use `c-prefix-spec-kwds-re' or similar, and the
3253 ;; template skipping isn't done properly. This will disappear soon.
3255 c++ (concat "friend" (c-lang-const c-simple-ws) "+"
3258 (c-lang-const c-simple-ws) "*"
3260 (c-lang-const c-simple-ws) "*"
3262 (c-lang-const c-simple-ws) "+")))
3263 (c-lang-defvar c-opt-friend-key (c-lang-const c-opt-friend-key))
3265 (c-lang-defconst c-opt-method-key
3266 ;; Special regexp to match the start of Objective-C methods. The
3267 ;; first submatch is assumed to end after the + or - key.
3270 ;; TODO: Ought to use a better method than anchoring on bol.
3273 (c-lang-const c-simple-ws) "*"
3274 (concat "\\(" ; Return type.
3276 (c-lang-const c-simple-ws) "*"
3278 "\\(" (c-lang-const c-symbol-key) "\\)"))
3279 (c-lang-defvar c-opt-method-key (c-lang-const c-opt-method-key))
3281 (c-lang-defconst c-type-decl-end-used
3282 ;; Must be set in buffers where the `c-type' text property might be
3283 ;; used with the value `c-decl-end'.
3285 ;; `c-decl-end' is used to mark the ends of labels and access keys
3286 ;; to make interactive refontification work better.
3287 t (or (c-lang-const c-recognize-colon-labels)
3288 (and (c-lang-const c-label-kwds) t))
3289 ;; `c-decl-end' is used to mark the end of the @-style directives in
3292 (c-lang-defvar c-type-decl-end-used (c-lang-const c-type-decl-end-used))
3294 (c-lang-defconst c-maybe-decl-faces
3295 "List of faces that might be put at the start of a type when
3296 `c-font-lock-declarations' runs. This must be evaluated (with `eval') at
3297 runtime to get the actual list of faces. This ensures that face name
3298 aliases in Emacs are resolved."
3301 c-reference-face-name
3302 font-lock-keyword-face)
3303 java (append (c-lang-const c-maybe-decl-faces)
3304 '(font-lock-preprocessor-face)))
3305 (c-lang-defvar c-maybe-decl-faces (c-lang-const c-maybe-decl-faces))
3308 ;;; Wrap up the `c-lang-defvar' system.
3310 ;; Compile in the list of language variables that has been collected
3311 ;; with the `c-lang-defvar' and `c-lang-setvar' macros. Note that the
3312 ;; first element of each is nil.
3313 (defconst c-lang-variable-inits (cc-eval-when-compile c-lang-variable-inits))
3314 (defconst c-emacs-variable-inits (cc-eval-when-compile c-emacs-variable-inits))
3316 ;; Make the `c-lang-setvar' variables buffer local in the current buffer.
3317 ;; These are typically standard emacs variables such as `comment-start'.
3318 (defmacro c-make-emacs-variables-local ()
3320 ,@(mapcar (lambda (init)
3321 `(make-local-variable ',(car init)))
3322 (cdr c-emacs-variable-inits))))
3324 (defun c-make-init-lang-vars-fun (mode)
3325 "Create a function that initializes all the language dependent variables
3328 This function should be evaluated at compile time, so that the
3329 function it returns is byte compiled with all the evaluated results
3330 from the language constants. Use the `c-init-language-vars' macro to
3331 accomplish that conveniently."
3333 (if (cc-bytecomp-is-compiling)
3334 ;; No need to byte compile this lambda since the byte compiler is
3335 ;; smart enough to detect the `funcall' construct in the
3336 ;; `c-init-language-vars' macro below and compile it all straight
3337 ;; into the function that contains `c-init-language-vars'.
3340 ;; This let sets up the context for `c-mode-var' and similar
3341 ;; that could be in the result from `c--macroexpand-all'.
3342 (let ((c-buffer-is-cc-mode ',mode)
3343 current-var source-eval)
3344 (c-make-emacs-variables-local)
3347 (if (eq c-version-sym ',c-version-sym)
3348 (setq ,@(let ((c-buffer-is-cc-mode mode)
3349 (c-lang-const-expansion 'immediate))
3350 ;; `c-lang-const' will expand to the evaluated
3351 ;; constant immediately in `c--macroexpand-all'
3355 `(current-var ',(car init)
3356 ,(car init) ,(c--macroexpand-all
3358 ;; Note: The following `append' copies the
3359 ;; first argument. That list is small, so
3360 ;; this doesn't matter too much.
3361 (append (cdr c-emacs-variable-inits)
3362 (cdr c-lang-variable-inits)))))
3364 ;; This diagnostic message isn't useful for end
3365 ;; users, so it's disabled.
3366 ;;(unless (get ',mode 'c-has-warned-lang-consts)
3367 ;; (message ,(concat "%s compiled with CC Mode %s "
3368 ;; "but loaded with %s - evaluating "
3369 ;; "language constants from source")
3370 ;; ',mode ,c-version c-version)
3371 ;; (put ',mode 'c-has-warned-lang-consts t))
3373 (setq source-eval t)
3374 (let ((init ',(append (cdr c-emacs-variable-inits)
3375 (cdr c-lang-variable-inits))))
3376 (dolist (var-init init)
3377 (setq current-var (car var-init))
3378 (set (car var-init) (eval (cadr var-init))))))
3382 (message "Eval error in the `c-lang-defvar' or `c-lang-setvar' for `%s'%s: %S"
3386 (fallback source eval - %s compiled with CC Mode %s but loaded with %s)"
3387 ',mode ,c-version c-version)
3390 (signal (car err) (cdr err)))))))
3392 ;; Being evaluated from source. Always use the dynamic method to
3393 ;; work well when `c-lang-defvar's in this file are reevaluated
3397 (let ((c-buffer-is-cc-mode ',mode)
3398 (init (append (cdr c-emacs-variable-inits)
3399 (cdr c-lang-variable-inits)))
3401 (c-make-emacs-variables-local)
3404 (dolist (var-init init)
3405 (setq current-var (car var-init))
3406 (set (car var-init) (eval (cadr var-init))))
3411 "Eval error in the `c-lang-defvar' or `c-lang-setver' for `%s' (source eval): %S"
3413 (signal (car err) (cdr err)))))))
3416 (defmacro c-init-language-vars (mode)
3417 "Initialize all the language dependent variables for the given mode.
3418 This macro is expanded at compile time to a form tailored for the mode
3419 in question, so MODE must be a constant. Therefore MODE is not
3420 evaluated and should not be quoted."
3421 `(funcall ,(c-make-init-lang-vars-fun mode)))
3424 (cc-provide 'cc-langs)
3427 ;; indent-tabs-mode: t
3430 ;;; cc-langs.el ends here