code.delx.au - gnu-emacs/blob - lisp/international/mule-util.el

   1 ;;; mule-util.el --- utility functions for mulitilingual environment (mule)
   2
   3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
   4 ;; Licensed to the Free Software Foundation.
   5
   6 ;; Keywords: mule, multilingual
   7
   8 ;; This file is part of GNU Emacs.
   9
  10 ;; GNU Emacs is free software; you can redistribute it and/or modify
  11 ;; it under the terms of the GNU General Public License as published by
  12 ;; the Free Software Foundation; either version 2, or (at your option)
  13 ;; any later version.
  14
  15 ;; GNU Emacs is distributed in the hope that it will be useful,
  16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;; GNU General Public License for more details.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  22 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  23 ;; Boston, MA 02111-1307, USA.
  24
  25 ;;; Commentary:
  26
  27 ;;; Code:
  28
  29 ;;; String manipulations while paying attention to multibyte
  30 ;;; characters.
  31
  32 ;;;###autoload
  33 (defun string-to-sequence (string type)
  34   "Convert STRING to a sequence of TYPE which contains characters in STRING.
  35 TYPE should be `list' or `vector'."
  36 ;;;  (let ((len (length string))
  37 ;;;        (i 0)
  38 ;;;        val)
  39     (cond ((eq type 'list)
  40            ;; Applicable post-Emacs 20.2 and asymptotically ~10 times
  41            ;; faster than the code below:
  42            (append string nil))
  43 ;;;        (setq val (make-list len 0))
  44 ;;;        (let ((l val))
  45 ;;;          (while (< i len)
  46 ;;;            (setcar l (aref string i))
  47 ;;;            (setq l (cdr l) i (1+ i))))))
  48           ((eq type 'vector)
  49            ;; As above.
  50            (vconcat string))
  51 ;;;        (setq val (make-vector len 0))
  52 ;;;        (while (< i len)
  53 ;;;          (aset val i (aref string i))
  54 ;;;          (setq i (1+ i))))
  55           (t
  56            (error "Invalid type: %s" type)))
  57 ;;;    val)
  58 )
  59
  60 ;;;###autoload
  61 (make-obsolete 'string-to-sequence
  62                "use `string-to-list' or `string-to-vector'."
  63                "21.4")
  64
  65 ;;;###autoload
  66 (defsubst string-to-list (string)
  67   "Return a list of characters in STRING."
  68   (append string nil))
  69
  70 ;;;###autoload
  71 (defsubst string-to-vector (string)
  72   "Return a vector of characters in STRING."
  73   (vconcat string))
  74
  75 ;;;###autoload
  76 (defun store-substring (string idx obj)
  77   "Embed OBJ (string or character) at index IDX of STRING."
  78   (if (integerp obj)
  79       (aset string idx obj)
  80     (let ((len1 (length obj))
  81           (len2 (length string))
  82           (i 0))
  83       (while (< i len1)
  84         (aset string (+ idx i) (aref obj i))
  85         (setq i (1+ i)))))
  86   string)
  87
  88 ;;;###autoload
  89 (defun truncate-string-to-width (str end-column
  90                                      &optional start-column padding ellipsis)
  91   "Truncate string STR to end at column END-COLUMN.
  92 The optional 3rd arg START-COLUMN, if non-nil, specifies the starting
  93 column; that means to return the characters occupying columns
  94 START-COLUMN ... END-COLUMN of STR.  Both END-COLUMN and START-COLUMN
  95 are specified in terms of character display width in the current
  96 buffer; see also `char-width'.
  97
  98 The optional 4th arg PADDING, if non-nil, specifies a padding
  99 character (which should have a display width of 1) to add at the end
 100 of the result if STR doesn't reach column END-COLUMN, or if END-COLUMN
 101 comes in the middle of a character in STR.  PADDING is also added at
 102 the beginning of the result if column START-COLUMN appears in the
 103 middle of a character in STR.
 104
 105 If PADDING is nil, no padding is added in these cases, so
 106 the resulting string may be narrower than END-COLUMN.
 107
 108 If ELLIPSIS is non-nil, it should be a string which will replace the
 109 end of STR (including any padding) if it extends beyond END-COLUMN,
 110 unless the display width of STR is equal to or less than the display
 111 width of ELLIPSIS.  If it is non-nil and not a string, then ELLIPSIS
 112 defaults to \"...\"."
 113   (or start-column
 114       (setq start-column 0))
 115   (when (and ellipsis (not (stringp ellipsis)))
 116     (setq ellipsis "..."))
 117   (let ((str-len (length str))
 118         (str-width (string-width str))
 119         (ellipsis-len (if ellipsis (length ellipsis) 0))
 120         (ellipsis-width (if ellipsis (string-width ellipsis) 0))
 121         (idx 0)
 122         (column 0)
 123         (head-padding "") (tail-padding "")
 124         ch last-column last-idx from-idx)
 125     (condition-case nil
 126         (while (< column start-column)
 127           (setq ch (aref str idx)
 128                 column (+ column (char-width ch))
 129                 idx (1+ idx)))
 130       (args-out-of-range (setq idx str-len)))
 131     (if (< column start-column)
 132         (if padding (make-string end-column padding) "")
 133       (when (and padding (> column start-column))
 134         (setq head-padding (make-string (- column start-column) padding)))
 135       (setq from-idx idx)
 136       (when (>= end-column column)
 137         (if (and (< end-column str-width)
 138                  (> str-width ellipsis-width))
 139             (setq end-column (- end-column ellipsis-width))
 140           (setq ellipsis ""))
 141         (condition-case nil
 142             (while (< column end-column)
 143               (setq last-column column
 144                     last-idx idx
 145                     ch (aref str idx)
 146                     column (+ column (char-width ch))
 147                     idx (1+ idx)))
 148           (args-out-of-range (setq idx str-len)))
 149         (when (> column end-column)
 150           (setq column last-column
 151                 idx last-idx))
 152         (when (and padding (< column end-column))
 153           (setq tail-padding (make-string (- end-column column) padding))))
 154       (concat head-padding (substring str from-idx idx)
 155               tail-padding ellipsis))))
 156
 157 ;;; Test suite for truncate-string-to-width
 158 ;; (dolist (test '((("" 0) . "")
 159 ;;              (("x" 1) . "x")
 160 ;;              (("xy" 1) . "x")
 161 ;;              (("xy" 2 1) . "y")
 162 ;;              (("xy" 0) . "")
 163 ;;              (("xy" 3) . "xy")
 164 ;;              (("\e$AVP\e(B" 0) . "")
 165 ;;              (("\e$AVP\e(B" 1) . "")
 166 ;;              (("\e$AVP\e(B" 2) . "\e$AVP\e(B")
 167 ;;              (("\e$AVP\e(B" 1 nil ? ) . " ")
 168 ;;              (("\e$AVPND\e(B" 3 1 ? ) . "  ")
 169 ;;              (("x\e$AVP\e(Bx" 2) . "x")
 170 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 171 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 172 ;;              (("x\e$AVP\e(Bx" 4 1) . "\e$AVP\e(Bx")
 173 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 8 1 ? ) . "or\e$(CGQ\e(Be\e$(C1[\e(B")
 174 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 7 2 ? ) . "r\e$(CGQ\e(Be ")
 175 ;;              (("" 0 nil nil "...") . "")
 176 ;;              (("x" 3 nil nil "...") . "x")
 177 ;;              (("\e$AVP\e(B" 3 nil nil "...") . "\e$AVP\e(B")
 178 ;;              (("foo" 3 nil nil "...") . "foo")
 179 ;;              (("foo" 2 nil nil "...") . "fo") ;; XEmacs failure?
 180 ;;              (("foobar" 6 0 nil "...") . "foobar")
 181 ;;              (("foobarbaz" 6 nil nil "...") . "foo...")
 182 ;;              (("foobarbaz" 7 2 nil "...") . "ob...")
 183 ;;              (("foobarbaz" 9 3 nil "...") . "barbaz")
 184 ;;              (("\e$B$3\e(Bh\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo" 15 1 ?  t) . " h\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo")
 185 ;;              (("\e$B$3\e(Bh\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo" 14 1 ?  t) . " h\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(B...")
 186 ;;              (("x" 3 nil nil "\e$(0GnM$\e(B") . "x")
 187 ;;              (("\e$AVP\e(B" 2 nil nil "\e$(0GnM$\e(B") . "\e$AVP\e(B")
 188 ;;              (("\e$AVP\e(B" 1 nil ?x "\e$(0GnM$\e(B") . "x") ;; XEmacs error
 189 ;;              (("\e$AVPND\e(B" 3 nil ?  "\e$(0GnM$\e(B") . "\e$AVP\e(B ") ;; XEmacs error
 190 ;;              (("foobarbaz" 4 nil nil  "\e$(0GnM$\e(B") . "\e$(0GnM$\e(B")
 191 ;;              (("foobarbaz" 5 nil nil  "\e$(0GnM$\e(B") . "f\e$(0GnM$\e(B")
 192 ;;              (("foobarbaz" 6 nil nil  "\e$(0GnM$\e(B") . "fo\e$(0GnM$\e(B")
 193 ;;              (("foobarbaz" 8 3 nil "\e$(0GnM$\e(B") . "b\e$(0GnM$\e(B")
 194 ;;              (("\e$B$3\e(Bh\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo" 14 4 ?x "\e$BF|K\8l\e(B") . "xe\e$B$KF|K\8l\e(B")
 195 ;;              (("\e$B$3\e(Bh\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo" 13 4 ?x "\e$BF|K\8l\e(B") . "xex\e$BF|K\8l\e(B")
 196 ;;              ))
 197 ;;   (let (ret)
 198 ;;     (condition-case e
 199 ;;      (setq ret (apply #'truncate-string-to-width (car test)))
 200 ;;       (error (setq ret e)))
 201 ;;     (unless (equal ret (cdr test))
 202 ;;       (error "%s: expected %s, got %s"
 203 ;;           (prin1-to-string (cons 'truncate-string-to-width (car test)))
 204 ;;           (prin1-to-string (cdr test))
 205 ;;           (if (consp ret)
 206 ;;               (format "error: %s: %s" (car ret)
 207 ;;                       (prin1-to-string (cdr ret)))
 208 ;;             (prin1-to-string ret))))))
 209
 210 ;;; For backward compatibility ...
 211 ;;;###autoload
 212 (defalias 'truncate-string 'truncate-string-to-width)
 213
 214 ;;;###autoload
 215 (make-obsolete 'truncate-string 'truncate-string-to-width "20.1")
 216 \f
 217 ;;; Nested alist handler.  Nested alist is alist whose elements are
 218 ;;; also nested alist.
 219
 220 ;;;###autoload
 221 (defsubst nested-alist-p (obj)
 222   "Return t if OBJ is a nested alist.
 223
 224 Nested alist is a list of the form (ENTRY . BRANCHES), where ENTRY is
 225 any Lisp object, and BRANCHES is a list of cons cells of the form
 226 \(KEY-ELEMENT . NESTED-ALIST).
 227
 228 You can use a nested alist to store any Lisp object (ENTRY) for a key
 229 sequence KEYSEQ, where KEYSEQ is a sequence of KEY-ELEMENT.  KEYSEQ
 230 can be a string, a vector, or a list."
 231   (and obj (listp obj) (listp (cdr obj))))
 232
 233 ;;;###autoload
 234 (defun set-nested-alist (keyseq entry alist &optional len branches)
 235   "Set ENTRY for KEYSEQ in a nested alist ALIST.
 236 Optional 4th arg LEN non-nil means the first LEN elements in KEYSEQ
 237  is considered.
 238 Optional argument BRANCHES if non-nil is branches for a keyseq
 239 longer than KEYSEQ.
 240 See the documentation of `nested-alist-p' for more detail."
 241   (or (nested-alist-p alist)
 242       (error "Invalid argument %s" alist))
 243   (let ((islist (listp keyseq))
 244         (len (or len (length keyseq)))
 245         (i 0)
 246         key-elt slot)
 247     (while (< i len)
 248       (if (null (nested-alist-p alist))
 249           (error "Keyseq %s is too long for this nested alist" keyseq))
 250       (setq key-elt (if islist (nth i keyseq) (aref keyseq i)))
 251       (setq slot (assoc key-elt (cdr alist)))
 252       (if (null slot)
 253           (progn
 254             (setq slot (cons key-elt (list t)))
 255             (setcdr alist (cons slot (cdr alist)))))
 256       (setq alist (cdr slot))
 257       (setq i (1+ i)))
 258     (setcar alist entry)
 259     (if branches
 260         (setcdr (last alist) branches))))
 261
 262 ;;;###autoload
 263 (defun lookup-nested-alist (keyseq alist &optional len start nil-for-too-long)
 264   "Look up key sequence KEYSEQ in nested alist ALIST.  Return the definition.
 265 Optional 1st argument LEN specifies the length of KEYSEQ.
 266 Optional 2nd argument START specifies index of the starting key.
 267 The returned value is normally a nested alist of which
 268 car part is the entry for KEYSEQ.
 269 If ALIST is not deep enough for KEYSEQ, return number which is
 270  how many key elements at the front of KEYSEQ it takes
 271  to reach a leaf in ALIST.
 272 Optional 3rd argument NIL-FOR-TOO-LONG non-nil means return nil
 273  even if ALIST is not deep enough."
 274   (or (nested-alist-p alist)
 275       (error "Invalid argument %s" alist))
 276   (or len
 277       (setq len (length keyseq)))
 278   (let ((i (or start 0)))
 279     (if (catch 'lookup-nested-alist-tag
 280           (if (listp keyseq)
 281               (while (< i len)
 282                 (if (setq alist (cdr (assoc (nth i keyseq) (cdr alist))))
 283                     (setq i (1+ i))
 284                   (throw 'lookup-nested-alist-tag t))))
 285           (while (< i len)
 286             (if (setq alist (cdr (assoc (aref keyseq i) (cdr alist))))
 287                 (setq i (1+ i))
 288               (throw 'lookup-nested-alist-tag t))))
 289         ;; KEYSEQ is too long.
 290         (if nil-for-too-long nil i)
 291       alist)))
 292
 293 \f
 294 ;; Coding system related functions.
 295
 296 ;;;###autoload
 297 (defun coding-system-post-read-conversion (coding-system)
 298   "Return the value of CODING-SYSTEM's `post-read-conversion' property."
 299   (coding-system-get coding-system 'post-read-conversion))
 300
 301 ;;;###autoload
 302 (defun coding-system-pre-write-conversion (coding-system)
 303   "Return the value of CODING-SYSTEM's `pre-write-conversion' property."
 304   (coding-system-get coding-system 'pre-write-conversion))
 305
 306 ;;;###autoload
 307 (defun coding-system-translation-table-for-decode (coding-system)
 308   "Return the value of CODING-SYSTEM's `translation-table-for-decode' property."
 309   (coding-system-get coding-system 'translation-table-for-decode))
 310
 311 ;;;###autoload
 312 (defun coding-system-translation-table-for-encode (coding-system)
 313   "Return the value of CODING-SYSTEM's `translation-table-for-encode' property."
 314   (coding-system-get coding-system 'translation-table-for-encode))
 315
 316 ;;;###autoload
 317 (defun coding-system-equal (coding-system-1 coding-system-2)
 318   "Return t if and only if CODING-SYSTEM-1 and CODING-SYSTEM-2 are identical.
 319 Two coding systems are identical if two symbols are equal
 320 or one is an alias of the other."
 321   (or (eq coding-system-1 coding-system-2)
 322       (and (equal (coding-system-spec coding-system-1)
 323                   (coding-system-spec coding-system-2))
 324            (let ((eol-type-1 (coding-system-eol-type coding-system-1))
 325                  (eol-type-2 (coding-system-eol-type coding-system-2)))
 326              (or (eq eol-type-1 eol-type-2)
 327                  (and (vectorp eol-type-1) (vectorp eol-type-2)))))))
 328
 329 ;;;###autoload
 330 (defmacro detect-coding-with-priority (from to priority-list)
 331   "Detect a coding system of the text between FROM and TO with PRIORITY-LIST.
 332 PRIORITY-LIST is an alist of coding categories vs the corresponding
 333 coding systems ordered by priority."
 334   `(unwind-protect
 335        (let* ((prio-list ,priority-list)
 336               (coding-category-list coding-category-list)
 337               ,@(mapcar (function (lambda (x) (list x x)))
 338                         coding-category-list))
 339          (mapc (function (lambda (x) (set (car x) (cdr x))))
 340                prio-list)
 341          (set-coding-priority (mapcar #'car prio-list))
 342          (detect-coding-region ,from ,to))
 343      ;; We must restore the internal database.
 344      (set-coding-priority coding-category-list)
 345      (update-coding-systems-internal)))
 346
 347 ;;;###autoload
 348 (defun detect-coding-with-language-environment (from to lang-env)
 349   "Detect a coding system of the text between FROM and TO with LANG-ENV.
 350 The detection takes into account the coding system priorities for the
 351 language environment LANG-ENV."
 352   (let ((coding-priority (get-language-info lang-env 'coding-priority)))
 353     (if coding-priority
 354         (detect-coding-with-priority
 355          from to
 356          (mapcar (function (lambda (x)
 357                              (cons (coding-system-get x 'coding-category) x)))
 358                  coding-priority))
 359       (detect-coding-region from to))))
 360
 361 \f
 362 (provide 'mule-util)
 363
 364 ;; Local Variables:
 365 ;; coding: iso-2022-7bit
 366 ;; End:
 367
 368 ;;; mule-util.el ends here