lisp/international/mule-util.el

   1 ;;; mule-util.el --- utility functions for mulitilingual environment (mule)
   2
   3 ;; Copyright (C) 1997, 1998, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005, 2006, 2007  Free Software Foundation, Inc.
   5 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   6 ;;   2005, 2006, 2007
   7 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   8 ;;   Registration Number H14PRO021
   9
  10 ;; Keywords: mule, multilingual
  11
  12 ;; This file is part of GNU Emacs.
  13
  14 ;; GNU Emacs is free software; you can redistribute it and/or modify
  15 ;; it under the terms of the GNU General Public License as published by
  16 ;; the Free Software Foundation; either version 2, or (at your option)
  17 ;; any later version.
  18
  19 ;; GNU Emacs is distributed in the hope that it will be useful,
  20 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  21 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22 ;; GNU General Public License for more details.
  23
  24 ;; You should have received a copy of the GNU General Public License
  25 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  26 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  27 ;; Boston, MA 02110-1301, USA.
  28
  29 ;;; Commentary:
  30
  31 ;;; Code:
  32
  33 ;;; String manipulations while paying attention to multibyte
  34 ;;; characters.
  35
  36 ;;;###autoload
  37 (defun string-to-sequence (string type)
  38   "Convert STRING to a sequence of TYPE which contains characters in STRING.
  39 TYPE should be `list' or `vector'."
  40 ;;;  (let ((len (length string))
  41 ;;;        (i 0)
  42 ;;;        val)
  43     (cond ((eq type 'list)
  44            ;; Applicable post-Emacs 20.2 and asymptotically ~10 times
  45            ;; faster than the code below:
  46            (append string nil))
  47 ;;;        (setq val (make-list len 0))
  48 ;;;        (let ((l val))
  49 ;;;          (while (< i len)
  50 ;;;            (setcar l (aref string i))
  51 ;;;            (setq l (cdr l) i (1+ i))))))
  52           ((eq type 'vector)
  53            ;; As above.
  54            (vconcat string))
  55 ;;;        (setq val (make-vector len 0))
  56 ;;;        (while (< i len)
  57 ;;;          (aset val i (aref string i))
  58 ;;;          (setq i (1+ i))))
  59           (t
  60            (error "Invalid type: %s" type)))
  61 ;;;    val)
  62 )
  63
  64 ;;;###autoload
  65 (make-obsolete 'string-to-sequence
  66                "use `string-to-list' or `string-to-vector'."
  67                "22.1")
  68
  69 ;;;###autoload
  70 (defsubst string-to-list (string)
  71   "Return a list of characters in STRING."
  72   (append string nil))
  73
  74 ;;;###autoload
  75 (defsubst string-to-vector (string)
  76   "Return a vector of characters in STRING."
  77   (vconcat string))
  78
  79 ;;;###autoload
  80 (defun store-substring (string idx obj)
  81   "Embed OBJ (string or character) at index IDX of STRING."
  82   (if (integerp obj)
  83       (aset string idx obj)
  84     (let ((len1 (length obj))
  85           (len2 (length string))
  86           (i 0))
  87       (while (< i len1)
  88         (aset string (+ idx i) (aref obj i))
  89         (setq i (1+ i)))))
  90   string)
  91
  92 ;;;###autoload
  93 (defun truncate-string-to-width (str end-column
  94                                      &optional start-column padding ellipsis)
  95   "Truncate string STR to end at column END-COLUMN.
  96 The optional 3rd arg START-COLUMN, if non-nil, specifies the starting
  97 column; that means to return the characters occupying columns
  98 START-COLUMN ... END-COLUMN of STR.  Both END-COLUMN and START-COLUMN
  99 are specified in terms of character display width in the current
 100 buffer; see also `char-width'.
 101
 102 The optional 4th arg PADDING, if non-nil, specifies a padding
 103 character (which should have a display width of 1) to add at the end
 104 of the result if STR doesn't reach column END-COLUMN, or if END-COLUMN
 105 comes in the middle of a character in STR.  PADDING is also added at
 106 the beginning of the result if column START-COLUMN appears in the
 107 middle of a character in STR.
 108
 109 If PADDING is nil, no padding is added in these cases, so
 110 the resulting string may be narrower than END-COLUMN.
 111
 112 If ELLIPSIS is non-nil, it should be a string which will replace the
 113 end of STR (including any padding) if it extends beyond END-COLUMN,
 114 unless the display width of STR is equal to or less than the display
 115 width of ELLIPSIS.  If it is non-nil and not a string, then ELLIPSIS
 116 defaults to \"...\"."
 117   (or start-column
 118       (setq start-column 0))
 119   (when (and ellipsis (not (stringp ellipsis)))
 120     (setq ellipsis "..."))
 121   (let ((str-len (length str))
 122         (str-width (string-width str))
 123         (ellipsis-len (if ellipsis (length ellipsis) 0))
 124         (ellipsis-width (if ellipsis (string-width ellipsis) 0))
 125         (idx 0)
 126         (column 0)
 127         (head-padding "") (tail-padding "")
 128         ch last-column last-idx from-idx)
 129     (condition-case nil
 130         (while (< column start-column)
 131           (setq ch (aref str idx)
 132                 column (+ column (char-width ch))
 133                 idx (1+ idx)))
 134       (args-out-of-range (setq idx str-len)))
 135     (if (< column start-column)
 136         (if padding (make-string end-column padding) "")
 137       (when (and padding (> column start-column))
 138         (setq head-padding (make-string (- column start-column) padding)))
 139       (setq from-idx idx)
 140       (when (>= end-column column)
 141         (if (and (< end-column str-width)
 142                  (> str-width ellipsis-width))
 143             (setq end-column (- end-column ellipsis-width))
 144           (setq ellipsis ""))
 145         (condition-case nil
 146             (while (< column end-column)
 147               (setq last-column column
 148                     last-idx idx
 149                     ch (aref str idx)
 150                     column (+ column (char-width ch))
 151                     idx (1+ idx)))
 152           (args-out-of-range (setq idx str-len)))
 153         (when (> column end-column)
 154           (setq column last-column
 155                 idx last-idx))
 156         (when (and padding (< column end-column))
 157           (setq tail-padding (make-string (- end-column column) padding))))
 158       (concat head-padding (substring str from-idx idx)
 159               tail-padding ellipsis))))
 160
 161 ;;; Test suite for truncate-string-to-width
 162 ;; (dolist (test '((("" 0) . "")
 163 ;;              (("x" 1) . "x")
 164 ;;              (("xy" 1) . "x")
 165 ;;              (("xy" 2 1) . "y")
 166 ;;              (("xy" 0) . "")
 167 ;;              (("xy" 3) . "xy")
 168 ;;              (("\e$AVP\e(B" 0) . "")
 169 ;;              (("\e$AVP\e(B" 1) . "")
 170 ;;              (("\e$AVP\e(B" 2) . "\e$AVP\e(B")
 171 ;;              (("\e$AVP\e(B" 1 nil ? ) . " ")
 172 ;;              (("\e$AVPND\e(B" 3 1 ? ) . "  ")
 173 ;;              (("x\e$AVP\e(Bx" 2) . "x")
 174 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 175 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 176 ;;              (("x\e$AVP\e(Bx" 4 1) . "\e$AVP\e(Bx")
 177 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 8 1 ? ) . "or\e$(CGQ\e(Be\e$(C1[\e(B")
 178 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 7 2 ? ) . "r\e$(CGQ\e(Be ")
 179 ;;              (("" 0 nil nil "...") . "")
 180 ;;              (("x" 3 nil nil "...") . "x")
 181 ;;              (("\e$AVP\e(B" 3 nil nil "...") . "\e$AVP\e(B")
 182 ;;              (("foo" 3 nil nil "...") . "foo")
 183 ;;              (("foo" 2 nil nil "...") . "fo") ;; XEmacs failure?
 184 ;;              (("foobar" 6 0 nil "...") . "foobar")
 185 ;;              (("foobarbaz" 6 nil nil "...") . "foo...")
 186 ;;              (("foobarbaz" 7 2 nil "...") . "ob...")
 187 ;;              (("foobarbaz" 9 3 nil "...") . "barbaz")
 188 ;;              (("\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$O\e(Bo")
 189 ;;              (("\e$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$A\e(B...")
 190 ;;              (("x" 3 nil nil "\e$(0GnM$\e(B") . "x")
 191 ;;              (("\e$AVP\e(B" 2 nil nil "\e$(0GnM$\e(B") . "\e$AVP\e(B")
 192 ;;              (("\e$AVP\e(B" 1 nil ?x "\e$(0GnM$\e(B") . "x") ;; XEmacs error
 193 ;;              (("\e$AVPND\e(B" 3 nil ?  "\e$(0GnM$\e(B") . "\e$AVP\e(B ") ;; XEmacs error
 194 ;;              (("foobarbaz" 4 nil nil  "\e$(0GnM$\e(B") . "\e$(0GnM$\e(B")
 195 ;;              (("foobarbaz" 5 nil nil  "\e$(0GnM$\e(B") . "f\e$(0GnM$\e(B")
 196 ;;              (("foobarbaz" 6 nil nil  "\e$(0GnM$\e(B") . "fo\e$(0GnM$\e(B")
 197 ;;              (("foobarbaz" 8 3 nil "\e$(0GnM$\e(B") . "b\e$(0GnM$\e(B")
 198 ;;              (("\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")
 199 ;;              (("\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")
 200 ;;              ))
 201 ;;   (let (ret)
 202 ;;     (condition-case e
 203 ;;      (setq ret (apply #'truncate-string-to-width (car test)))
 204 ;;       (error (setq ret e)))
 205 ;;     (unless (equal ret (cdr test))
 206 ;;       (error "%s: expected %s, got %s"
 207 ;;           (prin1-to-string (cons 'truncate-string-to-width (car test)))
 208 ;;           (prin1-to-string (cdr test))
 209 ;;           (if (consp ret)
 210 ;;               (format "error: %s: %s" (car ret)
 211 ;;                       (prin1-to-string (cdr ret)))
 212 ;;             (prin1-to-string ret))))))
 213
 214 \f
 215 ;;; Nested alist handler.  Nested alist is alist whose elements are
 216 ;;; also nested alist.
 217
 218 ;;;###autoload
 219 (defsubst nested-alist-p (obj)
 220   "Return t if OBJ is a nested alist.
 221
 222 Nested alist is a list of the form (ENTRY . BRANCHES), where ENTRY is
 223 any Lisp object, and BRANCHES is a list of cons cells of the form
 224 \(KEY-ELEMENT . NESTED-ALIST).
 225
 226 You can use a nested alist to store any Lisp object (ENTRY) for a key
 227 sequence KEYSEQ, where KEYSEQ is a sequence of KEY-ELEMENT.  KEYSEQ
 228 can be a string, a vector, or a list."
 229   (and obj (listp obj) (listp (cdr obj))))
 230
 231 ;;;###autoload
 232 (defun set-nested-alist (keyseq entry alist &optional len branches)
 233   "Set ENTRY for KEYSEQ in a nested alist ALIST.
 234 Optional 4th arg LEN non-nil means the first LEN elements in KEYSEQ
 235  is considered.
 236 Optional argument BRANCHES if non-nil is branches for a keyseq
 237 longer than KEYSEQ.
 238 See the documentation of `nested-alist-p' for more detail."
 239   (or (nested-alist-p alist)
 240       (error "Invalid argument %s" alist))
 241   (let ((islist (listp keyseq))
 242         (len (or len (length keyseq)))
 243         (i 0)
 244         key-elt slot)
 245     (while (< i len)
 246       (if (null (nested-alist-p alist))
 247           (error "Keyseq %s is too long for this nested alist" keyseq))
 248       (setq key-elt (if islist (nth i keyseq) (aref keyseq i)))
 249       (setq slot (assoc key-elt (cdr alist)))
 250       (if (null slot)
 251           (progn
 252             (setq slot (cons key-elt (list t)))
 253             (setcdr alist (cons slot (cdr alist)))))
 254       (setq alist (cdr slot))
 255       (setq i (1+ i)))
 256     (setcar alist entry)
 257     (if branches
 258         (setcdr (last alist) branches))))
 259
 260 ;;;###autoload
 261 (defun lookup-nested-alist (keyseq alist &optional len start nil-for-too-long)
 262   "Look up key sequence KEYSEQ in nested alist ALIST.  Return the definition.
 263 Optional 1st argument LEN specifies the length of KEYSEQ.
 264 Optional 2nd argument START specifies index of the starting key.
 265 The returned value is normally a nested alist of which
 266 car part is the entry for KEYSEQ.
 267 If ALIST is not deep enough for KEYSEQ, return number which is
 268  how many key elements at the front of KEYSEQ it takes
 269  to reach a leaf in ALIST.
 270 Optional 3rd argument NIL-FOR-TOO-LONG non-nil means return nil
 271  even if ALIST is not deep enough."
 272   (or (nested-alist-p alist)
 273       (error "Invalid argument %s" alist))
 274   (or len
 275       (setq len (length keyseq)))
 276   (let ((i (or start 0)))
 277     (if (catch 'lookup-nested-alist-tag
 278           (if (listp keyseq)
 279               (while (< i len)
 280                 (if (setq alist (cdr (assoc (nth i keyseq) (cdr alist))))
 281                     (setq i (1+ i))
 282                   (throw 'lookup-nested-alist-tag t))))
 283           (while (< i len)
 284             (if (setq alist (cdr (assoc (aref keyseq i) (cdr alist))))
 285                 (setq i (1+ i))
 286               (throw 'lookup-nested-alist-tag t))))
 287         ;; KEYSEQ is too long.
 288         (if nil-for-too-long nil i)
 289       alist)))
 290
 291 \f
 292 ;; Coding system related functions.
 293
 294 ;;;###autoload
 295 (defun coding-system-post-read-conversion (coding-system)
 296   "Return the value of CODING-SYSTEM's `post-read-conversion' property."
 297   (coding-system-get coding-system 'post-read-conversion))
 298
 299 ;;;###autoload
 300 (defun coding-system-pre-write-conversion (coding-system)
 301   "Return the value of CODING-SYSTEM's `pre-write-conversion' property."
 302   (coding-system-get coding-system 'pre-write-conversion))
 303
 304 ;;;###autoload
 305 (defun coding-system-translation-table-for-decode (coding-system)
 306   "Return the value of CODING-SYSTEM's `translation-table-for-decode' property."
 307   (coding-system-get coding-system 'translation-table-for-decode))
 308
 309 ;;;###autoload
 310 (defun coding-system-translation-table-for-encode (coding-system)
 311   "Return the value of CODING-SYSTEM's `translation-table-for-encode' property."
 312   (coding-system-get coding-system 'translation-table-for-encode))
 313
 314 ;;;###autoload
 315 (defmacro detect-coding-with-priority (from to priority-list)
 316   "Detect a coding system of the text between FROM and TO with PRIORITY-LIST.
 317 PRIORITY-LIST is an alist of coding categories vs the corresponding
 318 coding systems ordered by priority."
 319   `(unwind-protect
 320        (let* ((prio-list ,priority-list)
 321               (coding-category-list coding-category-list)
 322               ,@(mapcar (function (lambda (x) (list x x)))
 323                         coding-category-list))
 324          (mapc (function (lambda (x) (set (car x) (cdr x))))
 325                prio-list)
 326          (set-coding-priority (mapcar #'car prio-list))
 327          ;; Changing the binding of a coding category requires this call.
 328          (update-coding-systems-internal)
 329          (detect-coding-region ,from ,to))
 330      ;; We must restore the internal database.
 331      (set-coding-priority coding-category-list)
 332      (update-coding-systems-internal)))
 333
 334 ;;;###autoload
 335 (defun detect-coding-with-language-environment (from to lang-env)
 336   "Detect a coding system of the text between FROM and TO with LANG-ENV.
 337 The detection takes into account the coding system priorities for the
 338 language environment LANG-ENV."
 339   (let ((coding-priority (get-language-info lang-env 'coding-priority)))
 340     (if coding-priority
 341         (detect-coding-with-priority
 342          from to
 343          (mapcar (function (lambda (x)
 344                              (cons (coding-system-get x 'coding-category) x)))
 345                  coding-priority))
 346       (detect-coding-region from to))))
 347
 348 ;;;###autoload
 349 (defun char-displayable-p (char)
 350   "Return non-nil if we should be able to display CHAR.
 351 On a multi-font display, the test is only whether there is an
 352 appropriate font from the selected frame's fontset to display CHAR's
 353 charset in general.  Since fonts may be specified on a per-character
 354 basis, this may not be accurate."
 355   (cond ((< char 256)
 356          ;; Single byte characters are always displayable.
 357          t)
 358         ((not enable-multibyte-characters)
 359          ;; Maybe there's a font for it, but we can't put it in the buffer.
 360          nil)
 361         ((display-multi-font-p)
 362          ;; On a window system, a character is displayable if we have
 363          ;; a font for that character in the default face of the
 364          ;; currently selected frame.
 365          (car (internal-char-font nil char)))
 366         (t
 367          (let ((coding (terminal-coding-system)))
 368            (if coding
 369                (let ((safe-chars (coding-system-get coding 'safe-chars))
 370                      (safe-charsets (coding-system-get coding 'safe-charsets)))
 371                  (or (and safe-chars
 372                           (aref safe-chars char))
 373                      (and safe-charsets
 374                           (memq (char-charset char) safe-charsets)))))))))
 375 \f
 376 (provide 'mule-util)
 377
 378 ;; Local Variables:
 379 ;; coding: iso-2022-7bit
 380 ;; End:
 381
 382 ;; arch-tag: 5bdb52b6-a3a5-4529-b7a0-37d01b0e570b
 383 ;;; mule-util.el ends here