lisp/international/mule-util.el

   1 ;;; mule-util.el --- utility functions for multilingual environment (mule)
   2
   3 ;; Copyright (C) 1997-1998, 2000-2011  Free Software Foundation, Inc.
   4 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5 ;;   2005, 2006, 2007, 2008, 2009, 2010, 2011
   6 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   7 ;;   Registration Number H14PRO021
   8 ;; Copyright (C) 2003
   9 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
  10 ;;   Registration Number H13PRO009
  11
  12 ;; Keywords: mule, multilingual
  13
  14 ;; This file is part of GNU Emacs.
  15
  16 ;; GNU Emacs is free software: you can redistribute it and/or modify
  17 ;; it under the terms of the GNU General Public License as published by
  18 ;; the Free Software Foundation, either version 3 of the License, or
  19 ;; (at your option) any later version.
  20
  21 ;; GNU Emacs is distributed in the hope that it will be useful,
  22 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  23 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24 ;; GNU General Public License for more details.
  25
  26 ;; You should have received a copy of the GNU General Public License
  27 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  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$A$3\e(Bh\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(Bl\e$A$O\e(Bo" 15 1 ?  t) . " h\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(Bl\e$A$O\e(Bo")
 189 ;;              (("\e$A$3\e(Bh\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(Bl\e$A$O\e(Bo" 14 1 ?  t) . " h\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(B...")
 190 ;;              (("x" 3 nil nil "\e$(Gemk#\e(B") . "x")
 191 ;;              (("\e$AVP\e(B" 2 nil nil "\e$(Gemk#\e(B") . "\e$AVP\e(B")
 192 ;;              (("\e$AVP\e(B" 1 nil ?x "\e$(Gemk#\e(B") . "x") ;; XEmacs error
 193 ;;              (("\e$AVPND\e(B" 3 nil ?  "\e$(Gemk#\e(B") . "\e$AVP\e(B ") ;; XEmacs error
 194 ;;              (("foobarbaz" 4 nil nil  "\e$(Gemk#\e(B") . "\e$(Gemk#\e(B")
 195 ;;              (("foobarbaz" 5 nil nil  "\e$(Gemk#\e(B") . "f\e$(Gemk#\e(B")
 196 ;;              (("foobarbaz" 6 nil nil  "\e$(Gemk#\e(B") . "fo\e$(Gemk#\e(B")
 197 ;;              (("foobarbaz" 8 3 nil "\e$(Gemk#\e(B") . "b\e$(Gemk#\e(B")
 198 ;;              (("\e$A$3\e(Bh\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(Bl\e$A$O\e(Bo" 14 4 ?x "\e$AHU1>\e$(Gk#\e(B") . "xe\e$A$KHU1>\e$(Gk#\e(B")
 199 ;;              (("\e$A$3\e(Bh\e$A$s\e(Be\e$A$K\e(Bl\e$A$A\e(Bl\e$A$O\e(Bo" 13 4 ?x "\e$AHU1>\e$(Gk#\e(B") . "xex\e$AHU1>\e$(Gk#\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  are considered.
 236 Optional 5th 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       (unless slot
 251         (setq slot (cons key-elt (list t)))
 252         (setcdr alist (cons slot (cdr alist))))
 253       (setq alist (cdr slot))
 254       (setq i (1+ i)))
 255     (setcar alist entry)
 256     (if branches
 257         (setcdr (last alist) branches))))
 258
 259 ;;;###autoload
 260 (defun lookup-nested-alist (keyseq alist &optional len start nil-for-too-long)
 261   "Look up key sequence KEYSEQ in nested alist ALIST.  Return the definition.
 262 Optional 3rd argument LEN specifies the length of KEYSEQ.
 263 Optional 4th argument START specifies index of the starting key.
 264 The returned value is normally a nested alist of which
 265 car part is the entry for KEYSEQ.
 266 If ALIST is not deep enough for KEYSEQ, return number which is
 267  how many key elements at the front of KEYSEQ it takes
 268  to reach a leaf in ALIST.
 269 Optional 5th argument NIL-FOR-TOO-LONG non-nil means return nil
 270  even if ALIST is not deep enough."
 271   (or (nested-alist-p alist)
 272       (error "Invalid argument %s" alist))
 273   (or len
 274       (setq len (length keyseq)))
 275   (let ((i (or start 0)))
 276     (if (catch 'lookup-nested-alist-tag
 277           (if (listp keyseq)
 278               (while (< i len)
 279                 (if (setq alist (cdr (assoc (nth i keyseq) (cdr alist))))
 280                     (setq i (1+ i))
 281                   (throw 'lookup-nested-alist-tag t))))
 282           (while (< i len)
 283             (if (setq alist (cdr (assoc (aref keyseq i) (cdr alist))))
 284                 (setq i (1+ i))
 285               (throw 'lookup-nested-alist-tag t))))
 286         ;; KEYSEQ is too long.
 287         (if nil-for-too-long nil i)
 288       alist)))
 289
 290 \f
 291 ;; Coding system related functions.
 292
 293 ;;;###autoload
 294 (defun coding-system-post-read-conversion (coding-system)
 295   "Return the value of CODING-SYSTEM's `post-read-conversion' property."
 296   (coding-system-get coding-system :post-read-conversion))
 297
 298 ;;;###autoload
 299 (defun coding-system-pre-write-conversion (coding-system)
 300   "Return the value of CODING-SYSTEM's `pre-write-conversion' property."
 301   (coding-system-get coding-system :pre-write-conversion))
 302
 303 ;;;###autoload
 304 (defun coding-system-translation-table-for-decode (coding-system)
 305   "Return the value of CODING-SYSTEM's `decode-translation-table' property."
 306   (coding-system-get coding-system :decode-translation-table))
 307
 308 ;;;###autoload
 309 (defun coding-system-translation-table-for-encode (coding-system)
 310   "Return the value of CODING-SYSTEM's `encode-translation-table' property."
 311   (coding-system-get coding-system :encode-translation-table))
 312
 313 ;;;###autoload
 314 (defmacro with-coding-priority (coding-systems &rest body)
 315   "Execute BODY like `progn' with CODING-SYSTEMS at the front of priority list.
 316 CODING-SYSTEMS is a list of coding systems.  See `set-coding-system-priority'.
 317 This affects the implicit sorting of lists of coding systems returned by
 318 operations such as `find-coding-systems-region'."
 319   (let ((current (make-symbol "current")))
 320   `(let ((,current (coding-system-priority-list)))
 321      (apply #'set-coding-system-priority ,coding-systems)
 322      (unwind-protect
 323          (progn ,@body)
 324        (apply #'set-coding-system-priority ,current)))))
 325 ;;;###autoload(put 'with-coding-priority 'lisp-indent-function 1)
 326 (put 'with-coding-priority 'edebug-form-spec t)
 327
 328 ;;;###autoload
 329 (defmacro detect-coding-with-priority (from to priority-list)
 330   "Detect a coding system of the text between FROM and TO with PRIORITY-LIST.
 331 PRIORITY-LIST is an alist of coding categories vs the corresponding
 332 coding systems ordered by priority."
 333   `(with-coding-priority (mapcar #'cdr ,priority-list)
 334      (detect-coding-region ,from ,to)))
 335 (make-obsolete 'detect-coding-with-priority
 336                "use `with-coding-priority' and `detect-coding-region'." "23.1")
 337
 338 ;;;###autoload
 339 (defun detect-coding-with-language-environment (from to lang-env)
 340   "Detect a coding system for the text between FROM and TO with LANG-ENV.
 341 The detection takes into account the coding system priorities for the
 342 language environment LANG-ENV."
 343   (let ((coding-priority (get-language-info lang-env 'coding-priority)))
 344     (if coding-priority
 345         (with-coding-priority coding-priority
 346           (detect-coding-region from to)))))
 347
 348 (declare-function internal-char-font "fontset.c" (position &optional ch))
 349
 350 ;;;###autoload
 351 (defun char-displayable-p (char)
 352   "Return non-nil if we should be able to display CHAR.
 353 On a multi-font display, the test is only whether there is an
 354 appropriate font from the selected frame's fontset to display
 355 CHAR's charset in general.  Since fonts may be specified on a
 356 per-character basis, this may not be accurate."
 357   (cond ((< char 128)
 358          ;; ASCII characters are always displayable.
 359          t)
 360         ((not enable-multibyte-characters)
 361          ;; Maybe there's a font for it, but we can't put it in the buffer.
 362          nil)
 363         ((display-multi-font-p)
 364          ;; On a window system, a character is displayable if we have
 365          ;; a font for that character in the default face of the
 366          ;; currently selected frame.
 367          (car (internal-char-font nil char)))
 368         (t
 369          ;; On a terminal, a character is displayable if the coding
 370          ;; system for the terminal can encode it.
 371          (let ((coding (terminal-coding-system)))
 372            (when coding
 373              (let ((cs-list (coding-system-get coding :charset-list)))
 374                (cond
 375                  ((listp cs-list)
 376                   (catch 'tag
 377                     (mapc #'(lambda (charset)
 378                               (if (encode-char char charset)
 379                                   (throw 'tag charset)))
 380                           cs-list)
 381                     nil))
 382                  ((eq cs-list 'iso-2022)
 383                   (catch 'tag2
 384                     (mapc #'(lambda (charset)
 385                               (if (and (plist-get (charset-plist charset)
 386                                                   :iso-final-char)
 387                                        (encode-char char charset))
 388                                   (throw 'tag2 charset)))
 389                           charset-list)
 390                     nil))
 391                  ((eq cs-list 'emacs-mule)
 392                   (catch 'tag3
 393                     (mapc #'(lambda (charset)
 394                               (if (and (plist-get (charset-plist charset)
 395                                                   :emacs-mule-id)
 396                                        (encode-char char charset))
 397                                   (throw 'tag3 charset)))
 398                           charset-list)
 399                     nil)))))))))
 400 \f
 401 (provide 'mule-util)
 402
 403 ;; Local Variables:
 404 ;; coding: iso-2022-7bit
 405 ;; End:
 406
 407 ;;; mule-util.el ends here