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 ;; Copyright (C) 2002 Free Software Foundation, Inc.
   6 ;; Copyright (C) 2003
   7 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   8 ;;   Registration Number H13PRO009
   9
  10
  11 ;; Keywords: mule, multilingual
  12
  13 ;; This file is part of GNU Emacs.
  14
  15 ;; GNU Emacs is free software; you can redistribute it and/or modify
  16 ;; it under the terms of the GNU General Public License as published by
  17 ;; the Free Software Foundation; either version 2, or (at your option)
  18 ;; any later version.
  19
  20 ;; GNU Emacs is distributed in the hope that it will be useful,
  21 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  22 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  23 ;; GNU General Public License for more details.
  24
  25 ;; You should have received a copy of the GNU General Public License
  26 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  27 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  28 ;; Boston, MA 02111-1307, USA.
  29
  30 ;;; Commentary:
  31
  32 ;;; Code:
  33
  34 ;;; String manipulations while paying attention to multibyte
  35 ;;; characters.
  36
  37 ;;;###autoload
  38 (defun string-to-sequence (string type)
  39   "Convert STRING to a sequence of TYPE which contains characters in STRING.
  40 TYPE should be `list' or `vector'."
  41 ;;;  (let ((len (length string))
  42 ;;;        (i 0)
  43 ;;;        val)
  44     (cond ((eq type 'list)
  45            ;; Applicable post-Emacs 20.2 and asymptotically ~10 times
  46            ;; faster than the code below:
  47            (append string nil))
  48 ;;;        (setq val (make-list len 0))
  49 ;;;        (let ((l val))
  50 ;;;          (while (< i len)
  51 ;;;            (setcar l (aref string i))
  52 ;;;            (setq l (cdr l) i (1+ i))))))
  53           ((eq type 'vector)
  54            ;; As above.
  55            (vconcat string))
  56 ;;;        (setq val (make-vector len 0))
  57 ;;;        (while (< i len)
  58 ;;;          (aset val i (aref string i))
  59 ;;;          (setq i (1+ i))))
  60           (t
  61            (error "Invalid type: %s" type)))
  62 ;;;    val)
  63 )
  64
  65 ;;;###autoload
  66 (make-obsolete 'string-to-sequence
  67                "use `string-to-list' or `string-to-vector'."
  68                "21.4")
  69
  70 ;;;###autoload
  71 (defsubst string-to-list (string)
  72   "Return a list of characters in STRING."
  73   (append string nil))
  74
  75 ;;;###autoload
  76 (defsubst string-to-vector (string)
  77   "Return a vector of characters in STRING."
  78   (vconcat string))
  79
  80 ;;;###autoload
  81 (defun store-substring (string idx obj)
  82   "Embed OBJ (string or character) at index IDX of STRING."
  83   (if (integerp obj)
  84       (aset string idx obj)
  85     (let ((len1 (length obj))
  86           (len2 (length string))
  87           (i 0))
  88       (while (< i len1)
  89         (aset string (+ idx i) (aref obj i))
  90         (setq i (1+ i)))))
  91   string)
  92
  93 ;;;###autoload
  94 (defun truncate-string-to-width (str end-column
  95                                      &optional start-column padding ellipsis)
  96   "Truncate string STR to end at column END-COLUMN.
  97 The optional 3rd arg START-COLUMN, if non-nil, specifies the starting
  98 column; that means to return the characters occupying columns
  99 START-COLUMN ... END-COLUMN of STR.  Both END-COLUMN and START-COLUMN
 100 are specified in terms of character display width in the current
 101 buffer; see also `char-width'.
 102
 103 The optional 4th arg PADDING, if non-nil, specifies a padding
 104 character (which should have a display width of 1) to add at the end
 105 of the result if STR doesn't reach column END-COLUMN, or if END-COLUMN
 106 comes in the middle of a character in STR.  PADDING is also added at
 107 the beginning of the result if column START-COLUMN appears in the
 108 middle of a character in STR.
 109
 110 If PADDING is nil, no padding is added in these cases, so
 111 the resulting string may be narrower than END-COLUMN.
 112
 113 If ELLIPSIS is non-nil, it should be a string which will replace the
 114 end of STR (including any padding) if it extends beyond END-COLUMN,
 115 unless the display width of STR is equal to or less than the display
 116 width of ELLIPSIS.  If it is non-nil and not a string, then ELLIPSIS
 117 defaults to \"...\"."
 118   (or start-column
 119       (setq start-column 0))
 120   (when (and ellipsis (not (stringp ellipsis)))
 121     (setq ellipsis "..."))
 122   (let ((str-len (length str))
 123         (str-width (string-width str))
 124         (ellipsis-len (if ellipsis (length ellipsis) 0))
 125         (ellipsis-width (if ellipsis (string-width ellipsis) 0))
 126         (idx 0)
 127         (column 0)
 128         (head-padding "") (tail-padding "")
 129         ch last-column last-idx from-idx)
 130     (condition-case nil
 131         (while (< column start-column)
 132           (setq ch (aref str idx)
 133                 column (+ column (char-width ch))
 134                 idx (1+ idx)))
 135       (args-out-of-range (setq idx str-len)))
 136     (if (< column start-column)
 137         (if padding (make-string end-column padding) "")
 138       (when (and padding (> column start-column))
 139         (setq head-padding (make-string (- column start-column) padding)))
 140       (setq from-idx idx)
 141       (when (>= end-column column)
 142         (if (and (< end-column str-width)
 143                  (> str-width ellipsis-width))
 144             (setq end-column (- end-column ellipsis-width))
 145           (setq ellipsis ""))
 146         (condition-case nil
 147             (while (< column end-column)
 148               (setq last-column column
 149                     last-idx idx
 150                     ch (aref str idx)
 151                     column (+ column (char-width ch))
 152                     idx (1+ idx)))
 153           (args-out-of-range (setq idx str-len)))
 154         (when (> column end-column)
 155           (setq column last-column
 156                 idx last-idx))
 157         (when (and padding (< column end-column))
 158           (setq tail-padding (make-string (- end-column column) padding))))
 159       (concat head-padding (substring str from-idx idx)
 160               tail-padding ellipsis))))
 161
 162 ;;; Test suite for truncate-string-to-width
 163 ;; (dolist (test '((("" 0) . "")
 164 ;;              (("x" 1) . "x")
 165 ;;              (("xy" 1) . "x")
 166 ;;              (("xy" 2 1) . "y")
 167 ;;              (("xy" 0) . "")
 168 ;;              (("xy" 3) . "xy")
 169 ;;              (("\e$AVP\e(B" 0) . "")
 170 ;;              (("\e$AVP\e(B" 1) . "")
 171 ;;              (("\e$AVP\e(B" 2) . "\e$AVP\e(B")
 172 ;;              (("\e$AVP\e(B" 1 nil ? ) . " ")
 173 ;;              (("\e$AVPND\e(B" 3 1 ? ) . "  ")
 174 ;;              (("x\e$AVP\e(Bx" 2) . "x")
 175 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 176 ;;              (("x\e$AVP\e(Bx" 3) . "x\e$AVP\e(B")
 177 ;;              (("x\e$AVP\e(Bx" 4 1) . "\e$AVP\e(Bx")
 178 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 8 1 ? ) . "or\e$(CGQ\e(Be\e$(C1[\e(B")
 179 ;;              (("kor\e$(CGQ\e(Be\e$(C1[\e(Ban" 7 2 ? ) . "r\e$(CGQ\e(Be ")
 180 ;;              (("" 0 nil nil "...") . "")
 181 ;;              (("x" 3 nil nil "...") . "x")
 182 ;;              (("\e$AVP\e(B" 3 nil nil "...") . "\e$AVP\e(B")
 183 ;;              (("foo" 3 nil nil "...") . "foo")
 184 ;;              (("foo" 2 nil nil "...") . "fo") ;; XEmacs failure?
 185 ;;              (("foobar" 6 0 nil "...") . "foobar")
 186 ;;              (("foobarbaz" 6 nil nil "...") . "foo...")
 187 ;;              (("foobarbaz" 7 2 nil "...") . "ob...")
 188 ;;              (("foobarbaz" 9 3 nil "...") . "barbaz")
 189 ;;              (("\e$B$3\e(Bh\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo" 15 1 ?  t) . " h\e$B$s\e(Be\e$B$K\e(Bl\e$B$A\e(Bl\e$B$O\e(Bo")
 190 ;;              (("\e$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$A\e(B...")
 191 ;;              (("x" 3 nil nil "\e$(0GnM$\e(B") . "x")
 192 ;;              (("\e$AVP\e(B" 2 nil nil "\e$(0GnM$\e(B") . "\e$AVP\e(B")
 193 ;;              (("\e$AVP\e(B" 1 nil ?x "\e$(0GnM$\e(B") . "x") ;; XEmacs error
 194 ;;              (("\e$AVPND\e(B" 3 nil ?  "\e$(0GnM$\e(B") . "\e$AVP\e(B ") ;; XEmacs error
 195 ;;              (("foobarbaz" 4 nil nil  "\e$(0GnM$\e(B") . "\e$(0GnM$\e(B")
 196 ;;              (("foobarbaz" 5 nil nil  "\e$(0GnM$\e(B") . "f\e$(0GnM$\e(B")
 197 ;;              (("foobarbaz" 6 nil nil  "\e$(0GnM$\e(B") . "fo\e$(0GnM$\e(B")
 198 ;;              (("foobarbaz" 8 3 nil "\e$(0GnM$\e(B") . "b\e$(0GnM$\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$O\e(Bo" 14 4 ?x "\e$BF|K\8l\e(B") . "xe\e$B$KF|K\8l\e(B")
 200 ;;              (("\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")
 201 ;;              ))
 202 ;;   (let (ret)
 203 ;;     (condition-case e
 204 ;;      (setq ret (apply #'truncate-string-to-width (car test)))
 205 ;;       (error (setq ret e)))
 206 ;;     (unless (equal ret (cdr test))
 207 ;;       (error "%s: expected %s, got %s"
 208 ;;           (prin1-to-string (cons 'truncate-string-to-width (car test)))
 209 ;;           (prin1-to-string (cdr test))
 210 ;;           (if (consp ret)
 211 ;;               (format "error: %s: %s" (car ret)
 212 ;;                       (prin1-to-string (cdr ret)))
 213 ;;             (prin1-to-string ret))))))
 214
 215 ;;; For backward compatibility ...
 216 ;;;###autoload
 217 (defalias 'truncate-string 'truncate-string-to-width)
 218
 219 ;;;###autoload
 220 (make-obsolete 'truncate-string 'truncate-string-to-width "20.1")
 221 \f
 222 ;;; Nested alist handler.  Nested alist is alist whose elements are
 223 ;;; also nested alist.
 224
 225 ;;;###autoload
 226 (defsubst nested-alist-p (obj)
 227   "Return t if OBJ is a nested alist.
 228
 229 Nested alist is a list of the form (ENTRY . BRANCHES), where ENTRY is
 230 any Lisp object, and BRANCHES is a list of cons cells of the form
 231 \(KEY-ELEMENT . NESTED-ALIST).
 232
 233 You can use a nested alist to store any Lisp object (ENTRY) for a key
 234 sequence KEYSEQ, where KEYSEQ is a sequence of KEY-ELEMENT.  KEYSEQ
 235 can be a string, a vector, or a list."
 236   (and obj (listp obj) (listp (cdr obj))))
 237
 238 ;;;###autoload
 239 (defun set-nested-alist (keyseq entry alist &optional len branches)
 240   "Set ENTRY for KEYSEQ in a nested alist ALIST.
 241 Optional 4th arg LEN non-nil means the first LEN elements in KEYSEQ
 242  is considered.
 243 Optional argument BRANCHES if non-nil is branches for a keyseq
 244 longer than KEYSEQ.
 245 See the documentation of `nested-alist-p' for more detail."
 246   (or (nested-alist-p alist)
 247       (error "Invalid argument %s" alist))
 248   (let ((islist (listp keyseq))
 249         (len (or len (length keyseq)))
 250         (i 0)
 251         key-elt slot)
 252     (while (< i len)
 253       (if (null (nested-alist-p alist))
 254           (error "Keyseq %s is too long for this nested alist" keyseq))
 255       (setq key-elt (if islist (nth i keyseq) (aref keyseq i)))
 256       (setq slot (assoc key-elt (cdr alist)))
 257       (if (null slot)
 258           (progn
 259             (setq slot (cons key-elt (list t)))
 260             (setcdr alist (cons slot (cdr alist)))))
 261       (setq alist (cdr slot))
 262       (setq i (1+ i)))
 263     (setcar alist entry)
 264     (if branches
 265         (setcdr (last alist) branches))))
 266
 267 ;;;###autoload
 268 (defun lookup-nested-alist (keyseq alist &optional len start nil-for-too-long)
 269   "Look up key sequence KEYSEQ in nested alist ALIST.  Return the definition.
 270 Optional 1st argument LEN specifies the length of KEYSEQ.
 271 Optional 2nd argument START specifies index of the starting key.
 272 The returned value is normally a nested alist of which
 273 car part is the entry for KEYSEQ.
 274 If ALIST is not deep enough for KEYSEQ, return number which is
 275  how many key elements at the front of KEYSEQ it takes
 276  to reach a leaf in ALIST.
 277 Optional 3rd argument NIL-FOR-TOO-LONG non-nil means return nil
 278  even if ALIST is not deep enough."
 279   (or (nested-alist-p alist)
 280       (error "Invalid argument %s" alist))
 281   (or len
 282       (setq len (length keyseq)))
 283   (let ((i (or start 0)))
 284     (if (catch 'lookup-nested-alist-tag
 285           (if (listp keyseq)
 286               (while (< i len)
 287                 (if (setq alist (cdr (assoc (nth i keyseq) (cdr alist))))
 288                     (setq i (1+ i))
 289                   (throw 'lookup-nested-alist-tag t))))
 290           (while (< i len)
 291             (if (setq alist (cdr (assoc (aref keyseq i) (cdr alist))))
 292                 (setq i (1+ i))
 293               (throw 'lookup-nested-alist-tag t))))
 294         ;; KEYSEQ is too long.
 295         (if nil-for-too-long nil i)
 296       alist)))
 297
 298 \f
 299 ;; Coding system related functions.
 300
 301 ;;;###autoload
 302 (defun coding-system-post-read-conversion (coding-system)
 303   "Return the value of CODING-SYSTEM's `post-read-conversion' property."
 304   (coding-system-get coding-system :post-read-conversion))
 305
 306 ;;;###autoload
 307 (defun coding-system-pre-write-conversion (coding-system)
 308   "Return the value of CODING-SYSTEM's `pre-write-conversion' property."
 309   (coding-system-get coding-system :pre-write-conversion))
 310
 311 ;;;###autoload
 312 (defun coding-system-translation-table-for-decode (coding-system)
 313   "Return the value of CODING-SYSTEM's `decode-translation-table' property."
 314   (coding-system-get coding-system :decode-translation-table))
 315
 316 ;;;###autoload
 317 (defun coding-system-translation-table-for-encode (coding-system)
 318   "Return the value of CODING-SYSTEM's `encode-translation-table' property."
 319   (coding-system-get coding-system :encode-translation-table))
 320
 321 ;;;###autoload
 322 (defun coding-system-equal (coding-system-1 coding-system-2)
 323   "Return t if and only if CODING-SYSTEM-1 and CODING-SYSTEM-2 are identical.
 324 Two coding systems are identical if two symbols are equal
 325 or one is an alias of the other."
 326   (or (eq coding-system-1 coding-system-2)
 327       (and (equal (coding-system-plist coding-system-1)
 328                   (coding-system-plist coding-system-2))
 329            (let ((eol-type-1 (coding-system-eol-type coding-system-1))
 330                  (eol-type-2 (coding-system-eol-type coding-system-2)))
 331              (or (eq eol-type-1 eol-type-2)
 332                  (and (vectorp eol-type-1) (vectorp eol-type-2)))))))
 333
 334 ;;;###autoload
 335 (defmacro with-coding-priority (coding-systems &rest body)
 336   "Execute BODY like `progn' with CODING-SYSTEMS at the front of priority list.
 337 CODING-SYSTEMS is a list of coding systems.  See
 338 `set-coding-priority'.  This affects the implicit sorting of lists of
 339 coding sysems returned by operations such as `find-coding-systems-region'."
 340   (let ((current (make-symbol "current")))
 341   `(let ((,current (coding-system-priority-list)))
 342      (apply #'set-coding-system-priority ,coding-systems)
 343      (unwind-protect
 344          (progn ,@body)
 345        (apply #'set-coding-system-priority ,current)))))
 346 (put 'with-coding-priority 'lisp-indent-function 1)
 347 (put 'with-coding-priority 'edebug-form-spec t)
 348
 349 ;;;###autoload
 350 (defmacro detect-coding-with-priority (from to priority-list)
 351   "Detect a coding system of the text between FROM and TO with PRIORITY-LIST.
 352 PRIORITY-LIST is an alist of coding categories vs the corresponding
 353 coding systems ordered by priority."
 354   `(with-coding-priority (mapcar #'cdr ,priority-list)
 355      (detect-coding-region ,from ,to)))
 356 (make-obsolete 'detect-coding-with-priority
 357                "Use with-coding-priority and detect-coding-region" "22.1")
 358
 359 ;;;###autoload
 360 (defun detect-coding-with-language-environment (from to lang-env)
 361   "Detect a coding system of the text between FROM and TO with LANG-ENV.
 362 The detection takes into account the coding system priorities for the
 363 language environment LANG-ENV."
 364   (let ((coding-priority (get-language-info lang-env 'coding-priority)))
 365     (if coding-priority
 366         (with-coding-priority coding-priority
 367           (detect-coding-region from to)))))
 368
 369 \f
 370 (provide 'mule-util)
 371
 372 ;; Local Variables:
 373 ;; coding: iso-2022-7bit
 374 ;; End:
 375
 376 ;;; mule-util.el ends here