lisp/nxml/nxml-uchnm.el

   1 ;;; nxml-uchnm.el --- support for Unicode standard cha names in nxml-mode
   2
   3 ;; Copyright (C) 2003, 2007, 2008 Free Software Foundation, Inc.
   4
   5 ;; Author: James Clark
   6 ;; Keywords: XML
   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 3, 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., 51 Franklin Street, Fifth Floor,
  23 ;; Boston, MA 02110-1301, USA.
  24
  25 ;;; Commentary:
  26
  27 ;; This enables the use of the character names defined in the Unicode
  28 ;; Standard.  The use of the names can be controlled on a per-block
  29 ;; basis, so as both to reduce memory usage and loading time,
  30 ;; and to make completion work better.
  31
  32 ;;; Code:
  33
  34 (require 'nxml-mode)
  35
  36 (defconst nxml-unicode-blocks
  37   '(("Basic Latin" #x0000 #x007F)
  38     ("Latin-1 Supplement" #x0080 #x00FF)
  39     ("Latin Extended-A" #x0100 #x017F)
  40     ("Latin Extended-B" #x0180 #x024F)
  41     ("IPA Extensions" #x0250 #x02AF)
  42     ("Spacing Modifier Letters" #x02B0 #x02FF)
  43     ("Combining Diacritical Marks" #x0300 #x036F)
  44     ("Greek and Coptic" #x0370 #x03FF)
  45     ("Cyrillic" #x0400 #x04FF)
  46     ("Cyrillic Supplementary" #x0500 #x052F)
  47     ("Armenian" #x0530 #x058F)
  48     ("Hebrew" #x0590 #x05FF)
  49     ("Arabic" #x0600 #x06FF)
  50     ("Syriac" #x0700 #x074F)
  51     ("Thaana" #x0780 #x07BF)
  52     ("Devanagari" #x0900 #x097F)
  53     ("Bengali" #x0980 #x09FF)
  54     ("Gurmukhi" #x0A00 #x0A7F)
  55     ("Gujarati" #x0A80 #x0AFF)
  56     ("Oriya" #x0B00 #x0B7F)
  57     ("Tamil" #x0B80 #x0BFF)
  58     ("Telugu" #x0C00 #x0C7F)
  59     ("Kannada" #x0C80 #x0CFF)
  60     ("Malayalam" #x0D00 #x0D7F)
  61     ("Sinhala" #x0D80 #x0DFF)
  62     ("Thai" #x0E00 #x0E7F)
  63     ("Lao" #x0E80 #x0EFF)
  64     ("Tibetan" #x0F00 #x0FFF)
  65     ("Myanmar" #x1000 #x109F)
  66     ("Georgian" #x10A0 #x10FF)
  67     ("Hangul Jamo" #x1100 #x11FF)
  68     ("Ethiopic" #x1200 #x137F)
  69     ("Cherokee" #x13A0 #x13FF)
  70     ("Unified Canadian Aboriginal Syllabics" #x1400 #x167F)
  71     ("Ogham" #x1680 #x169F)
  72     ("Runic" #x16A0 #x16FF)
  73     ("Tagalog" #x1700 #x171F)
  74     ("Hanunoo" #x1720 #x173F)
  75     ("Buhid" #x1740 #x175F)
  76     ("Tagbanwa" #x1760 #x177F)
  77     ("Khmer" #x1780 #x17FF)
  78     ("Mongolian" #x1800 #x18AF)
  79     ("Latin Extended Additional" #x1E00 #x1EFF)
  80     ("Greek Extended" #x1F00 #x1FFF)
  81     ("General Punctuation" #x2000 #x206F)
  82     ("Superscripts and Subscripts" #x2070 #x209F)
  83     ("Currency Symbols" #x20A0 #x20CF)
  84     ("Combining Diacritical Marks for Symbols" #x20D0 #x20FF)
  85     ("Letterlike Symbols" #x2100 #x214F)
  86     ("Number Forms" #x2150 #x218F)
  87     ("Arrows" #x2190 #x21FF)
  88     ("Mathematical Operators" #x2200 #x22FF)
  89     ("Miscellaneous Technical" #x2300 #x23FF)
  90     ("Control Pictures" #x2400 #x243F)
  91     ("Optical Character Recognition" #x2440 #x245F)
  92     ("Enclosed Alphanumerics" #x2460 #x24FF)
  93     ("Box Drawing" #x2500 #x257F)
  94     ("Block Elements" #x2580 #x259F)
  95     ("Geometric Shapes" #x25A0 #x25FF)
  96     ("Miscellaneous Symbols" #x2600 #x26FF)
  97     ("Dingbats" #x2700 #x27BF)
  98     ("Miscellaneous Mathematical Symbols-A" #x27C0 #x27EF)
  99     ("Supplemental Arrows-A" #x27F0 #x27FF)
 100     ("Braille Patterns" #x2800 #x28FF)
 101     ("Supplemental Arrows-B" #x2900 #x297F)
 102     ("Miscellaneous Mathematical Symbols-B" #x2980 #x29FF)
 103     ("Supplemental Mathematical Operators" #x2A00 #x2AFF)
 104     ("CJK Radicals Supplement" #x2E80 #x2EFF)
 105     ("Kangxi Radicals" #x2F00 #x2FDF)
 106     ("Ideographic Description Characters" #x2FF0 #x2FFF)
 107     ("CJK Symbols and Punctuation" #x3000 #x303F)
 108     ("Hiragana" #x3040 #x309F)
 109     ("Katakana" #x30A0 #x30FF)
 110     ("Bopomofo" #x3100 #x312F)
 111     ("Hangul Compatibility Jamo" #x3130 #x318F)
 112     ("Kanbun" #x3190 #x319F)
 113     ("Bopomofo Extended" #x31A0 #x31BF)
 114     ("Katakana Phonetic Extensions" #x31F0 #x31FF)
 115     ("Enclosed CJK Letters and Months" #x3200 #x32FF)
 116     ("CJK Compatibility" #x3300 #x33FF)
 117     ("CJK Unified Ideographs Extension A" #x3400 #x4DBF)
 118     ;;("CJK Unified Ideographs" #x4E00 #x9FFF)
 119     ("Yi Syllables" #xA000 #xA48F)
 120     ("Yi Radicals" #xA490 #xA4CF)
 121     ;;("Hangul Syllables" #xAC00 #xD7AF)
 122     ;;("High Surrogates" #xD800 #xDB7F)
 123     ;;("High Private Use Surrogates" #xDB80 #xDBFF)
 124     ;;("Low Surrogates" #xDC00 #xDFFF)
 125     ;;("Private Use Area" #xE000 #xF8FF)
 126     ;;("CJK Compatibility Ideographs" #xF900 #xFAFF)
 127     ("Alphabetic Presentation Forms" #xFB00 #xFB4F)
 128     ("Arabic Presentation Forms-A" #xFB50 #xFDFF)
 129     ("Variation Selectors" #xFE00 #xFE0F)
 130     ("Combining Half Marks" #xFE20 #xFE2F)
 131     ("CJK Compatibility Forms" #xFE30 #xFE4F)
 132     ("Small Form Variants" #xFE50 #xFE6F)
 133     ("Arabic Presentation Forms-B" #xFE70 #xFEFF)
 134     ("Halfwidth and Fullwidth Forms" #xFF00 #xFFEF)
 135     ("Specials" #xFFF0 #xFFFF)
 136     ("Old Italic" #x10300 #x1032F)
 137     ("Gothic" #x10330 #x1034F)
 138     ("Deseret" #x10400 #x1044F)
 139     ("Byzantine Musical Symbols" #x1D000 #x1D0FF)
 140     ("Musical Symbols" #x1D100 #x1D1FF)
 141     ("Mathematical Alphanumeric Symbols" #x1D400 #x1D7FF)
 142     ;;("CJK Unified Ideographs Extension B" #x20000 #x2A6DF)
 143     ;;("CJK Compatibility Ideographs Supplement" #x2F800 #x2FA1F)
 144     ("Tags" #xE0000 #xE007F)
 145     ;;("Supplementary Private Use Area-A" #xF0000 #xFFFFF)
 146     ;;("Supplementary Private Use Area-B" #x100000 #x10FFFF)
 147     )
 148   "List of Unicode blocks.
 149 For each block there is a list (NAME FIRST LAST), where
 150 NAME is a string giving the offical name of the block,
 151 FIRST is the first code-point and LAST is the last code-point.
 152 Blocks containing only characters with algorithmic names or no names
 153 are omitted.")
 154
 155 (defun nxml-unicode-block-char-name-set (name)
 156   "Return a symbol for a block whose offical Unicode name is NAME.
 157 The symbol is generated by downcasing and replacing each space
 158 by a hyphen."
 159   (intern (replace-regexp-in-string " " "-" (downcase name))))
 160
 161 ;; This is intended to be a superset of the coverage
 162 ;; of existing standard entity sets.
 163 (defvar nxml-enabled-unicode-blocks-default
 164   '(basic-latin
 165     latin-1-supplement
 166     latin-extended-a
 167     latin-extended-b
 168     ipa-extensions
 169     spacing-modifier-letters
 170     combining-diacritical-marks
 171     greek-and-coptic
 172     cyrillic
 173     general-punctuation
 174     superscripts-and-subscripts
 175     currency-symbols
 176     combining-diacritical-marks-for-symbols
 177     letterlike-symbols
 178     number-forms
 179     arrows
 180     mathematical-operators
 181     miscellaneous-technical
 182     control-pictures
 183     optical-character-recognition
 184     enclosed-alphanumerics
 185     box-drawing
 186     block-elements
 187     geometric-shapes
 188     miscellaneous-symbols
 189     dingbats
 190     miscellaneous-mathematical-symbols-a
 191     supplemental-arrows-a
 192     supplemental-arrows-b
 193     miscellaneous-mathematical-symbols-b
 194     supplemental-mathematical-operators
 195     cjk-symbols-and-punctuation
 196     alphabetic-presentation-forms
 197     variation-selectors
 198     small-form-variants
 199     specials
 200     mathematical-alphanumeric-symbols)
 201   "Default value for `nxml-enabled-unicode-blocks'.")
 202
 203 (mapc (lambda (block)
 204         (nxml-autoload-char-name-set
 205          (nxml-unicode-block-char-name-set (car block))
 206          (expand-file-name
 207           (format "nxml/%05X-%05X"
 208                   (nth 1 block)
 209                   (nth 2 block))
 210           data-directory)))
 211       nxml-unicode-blocks)
 212
 213 ;; Internal flag to control whether customize reloads the character tables.
 214 ;; Should be set the first time the
 215 (defvar nxml-internal-unicode-char-name-sets-enabled nil)
 216
 217 (defcustom nxml-enabled-unicode-blocks nxml-enabled-unicode-blocks-default
 218   "List of Unicode blocks for which Unicode character names are enabled.
 219 Each block is identified by a symbol derived from the name
 220 of the block by downcasing and replacing each space by a hyphen."
 221   :group 'nxml
 222   :set (lambda (sym value)
 223          (set-default 'nxml-enabled-unicode-blocks value)
 224          (when nxml-internal-unicode-char-name-sets-enabled
 225            (nxml-enable-unicode-char-name-sets)))
 226   :type (cons 'set
 227               (mapcar (lambda (block)
 228                         `(const :tag ,(format "%s (%04X-%04X)"
 229                                               (nth 0 block)
 230                                               (nth 1 block)
 231                                               (nth 2 block))
 232                                 ,(nxml-unicode-block-char-name-set
 233                                   (nth 0 block))))
 234                       nxml-unicode-blocks)))
 235
 236 ;;;###autoload
 237 (defun nxml-enable-unicode-char-name-sets ()
 238   "Enable the use of Unicode standard names for characters.
 239 The Unicode blocks for which names are enabled is controlled by
 240 the variable `nxml-enabled-unicode-blocks'."
 241   (interactive)
 242   (setq nxml-internal-unicode-char-name-sets-enabled t)
 243   (mapc (lambda (block)
 244           (nxml-disable-char-name-set
 245            (nxml-unicode-block-char-name-set (car block))))
 246         nxml-unicode-blocks)
 247   (mapc (lambda (nameset)
 248           (nxml-enable-char-name-set nameset))
 249         nxml-enabled-unicode-blocks))
 250
 251 (provide 'nxml-uchnm)
 252
 253 ;; arch-tag: 440248c3-b604-467c-8b50-e83662c659a3
 254 ;;; nxml-uchnm.el ends here