]> git.sesse.net Git - casparcg/blob - core/producer/text/utils/texture_font.cpp
* made it optional to normalize coordinates of a text-producer in order for it to...
[casparcg] / core / producer / text / utils / texture_font.cpp
1 #include "..\..\..\StdAfx.h"
2
3 #include "texture_atlas.h"
4 #include "texture_font.h"
5
6 #include <map>
7 #include <ft2build.h>
8 #include FT_FREETYPE_H
9 #include FT_GLYPH_H
10
11 namespace caspar { namespace core { namespace text {
12
13 struct freetype_exception : virtual caspar_exception
14 {
15         freetype_exception() {}
16         explicit freetype_exception(const char* msg) : caspar_exception(msg) {}
17 };
18
19 struct unicode_range
20 {
21         unicode_range() : first(0), last(0) {}
22         unicode_range(int f, int l) : first(f), last(l) {}
23
24         int first;
25         int last;
26 };
27
28 unicode_range get_range(unicode_block block);
29
30
31 struct texture_font::impl
32 {
33 private:
34         struct glyph_info
35         {
36                 glyph_info(int w, int h, float l, float t, float r, float b) : width(w), height(h), left(l), top(t), right(r), bottom(b)
37                 {}
38
39                 float left, top, right, bottom;
40                 int width, height;
41         };
42
43         FT_Library              lib_;
44         FT_Face                 face_;
45         texture_atlas   atlas_;
46         float                   size_;
47         bool                    normalize_;
48         std::map<int, glyph_info> glyphs_;
49
50 public:
51         impl::impl(texture_atlas& atlas, const std::wstring& filename, float size, bool normalize_coordinates) : lib_(nullptr), face_(nullptr), atlas_(atlas), size_(size), normalize_(normalize_coordinates)
52         {
53                 try
54                 {
55                         FT_Error err;
56                         err = FT_Init_FreeType(&lib_);
57                         if(err) throw freetype_exception("Failed to initialize freetype");
58
59                         err = FT_New_Face(lib_, u8(filename).c_str(), 0, &face_);
60                         if(err) throw freetype_exception("Failed to load font");
61
62                         err = FT_Set_Char_Size(face_, (FT_F26Dot6)(size*64), 0, 72, 72);
63                         if(err) throw freetype_exception("Failed to set font size");
64                 }
65                 catch(std::exception& ex)
66                 {
67                         if(face_ != nullptr)
68                                 FT_Done_Face(face_);
69                         if(lib_ != nullptr)
70                                 FT_Done_FreeType(lib_);
71
72                         throw ex;
73                 }
74         }
75
76         ~impl()
77         {
78                 if(face_ != nullptr)
79                         FT_Done_Face(face_);
80                 if(lib_ != nullptr)
81                         FT_Done_FreeType(lib_);
82         }
83
84         int count_glyphs_in_range(unicode_block block)
85         { 
86                 unicode_range range = get_range(block);
87
88                 //TODO: extract info from freetype
89
90                 //very pesimistic, assumes a glyph for each charcode
91                 return range.last - range.first;
92         }
93
94         void impl::load_glyphs(unicode_block block, const color<float>& col)
95         {
96                 FT_Error err;
97                 int flags = FT_LOAD_RENDER | FT_LOAD_FORCE_AUTOHINT | FT_LOAD_TARGET_NORMAL;
98                 unicode_range range = get_range(block);
99
100                 for(int i = range.first; i <= range.last; ++i)
101                 {
102                         FT_UInt glyph_index = FT_Get_Char_Index(face_, i);
103                         if(!glyph_index)        //ignore codes that doesn't have a glyph for now. Might want to map these to a special glyph later.
104                                 continue;
105                         
106                         err = FT_Load_Glyph(face_, glyph_index, flags);
107                         if(err) continue;       //igonore glyphs that fail to load
108
109                         const FT_Bitmap& bitmap = face_->glyph->bitmap; //shorthand notation
110
111                         auto region = atlas_.get_region(bitmap.width+1, bitmap.rows+1);
112                         if(region.x < 0)
113                         {
114                                 //the glyph doesn't fit in the texture-atlas. ignore it for now.
115                                 //we might want to restart with a bigger atlas in the future
116                                 continue;
117                         }
118
119                         atlas_.set_region(region.x, region.y, bitmap.width, bitmap.rows, bitmap.buffer, bitmap.pitch, col);
120                         glyphs_.insert(std::pair<int, glyph_info>(i, glyph_info(bitmap.width, bitmap.rows, 
121                                                                                 region.x / (float)atlas_.width(), 
122                                                                                 region.y / (float)atlas_.height(), 
123                                                                                 (region.x + bitmap.width) / (float)atlas_.width(), 
124                                                                                 (region.y + bitmap.rows) / (float)atlas_.height())));
125                 }
126         }
127
128         std::vector<float> create_vertex_stream(const std::wstring& str, int x, int y, int parent_width, int parent_height, string_metrics* metrics)
129         {
130                 //TODO: detect glyphs that aren't in the atlas and load them (and maybe that entire unicode_block on the fly
131
132                 std::vector<float> result(16*str.length(), 0);
133
134                 bool use_kerning = (face_->face_flags & FT_FACE_FLAG_KERNING) == FT_FACE_FLAG_KERNING;
135                 int index = 0;
136                 FT_UInt previous = 0;
137                 float pos_x = (float)x;
138                 float pos_y = (float)y;
139
140                 int maxBearingY = 0;
141                 int maxHeight = 0;
142
143                 auto end = str.end();
144                 for(auto it = str.begin(); it != end; ++it, ++index)
145                 {
146                         auto glyph_it = glyphs_.find(*it);
147                         if(glyph_it != glyphs_.end())
148                         {       
149                                 const glyph_info& coords = glyph_it->second;
150
151                                 FT_UInt glyph_index = FT_Get_Char_Index(face_, (*it));
152
153                                 if(use_kerning && previous && glyph_index)
154                                 {
155                                         FT_Vector delta;
156                                         FT_Get_Kerning(face_, previous, glyph_index, FT_KERNING_DEFAULT, &delta);
157
158                                         pos_x += delta.x / 64.0f;
159                                 }
160
161                                 FT_Load_Glyph(face_, glyph_index, FT_LOAD_NO_BITMAP | FT_LOAD_FORCE_AUTOHINT | FT_LOAD_TARGET_NORMAL);
162
163                                 float left = (pos_x + face_->glyph->metrics.horiBearingX/64.0f) / parent_width ;
164                                 float right = ((pos_x + face_->glyph->metrics.horiBearingX/64.0f) + coords.width) / parent_width;
165
166                                 float top = (pos_y - face_->glyph->metrics.horiBearingY/64.0f) / parent_height;
167                                 float bottom = ((pos_y - face_->glyph->metrics.horiBearingY/64.0f) + coords.height) / parent_height;
168
169                                 //vertex 1 top left
170                                 result[index*16 + 0] = left;                    //vertex.x
171                                 result[index*16 + 1] = top;                             //vertex.y
172                                 result[index*16 + 2] = coords.left;             //texcoord.r
173                                 result[index*16 + 3] = coords.top;              //texcoord.s
174
175                                 //vertex 2 top right
176                                 result[index*16 + 4] = right;                   //vertex.x
177                                 result[index*16 + 5] = top;                             //vertex.y
178                                 result[index*16 + 6] = coords.right;    //texcoord.r
179                                 result[index*16 + 7] = coords.top;              //texcoord.s
180
181                                 //vertex 3 bottom right
182                                 result[index*16 + 8] = right;                   //vertex.x
183                                 result[index*16 + 9] = bottom;                  //vertex.y
184                                 result[index*16 + 10] = coords.right;   //texcoord.r
185                                 result[index*16 + 11] = coords.bottom;  //texcoord.s
186
187                                 //vertex 4 bottom left
188                                 result[index*16 + 12] = left;                   //vertex.x
189                                 result[index*16 + 13] = bottom;                 //vertex.y
190                                 result[index*16 + 14] = coords.left;    //texcoord.r
191                                 result[index*16 + 15] = coords.bottom;  //texcoord.s
192
193                                 int bearingY = face_->glyph->metrics.horiBearingY >> 6;
194                                 if(bearingY > maxBearingY)
195                                         maxBearingY = bearingY;
196
197                                 if(coords.height > maxHeight)
198                                         maxHeight = coords.height;
199
200                                 pos_x += face_->glyph->advance.x / 64.0f;
201                                 previous = glyph_index;
202                         }
203                         else
204                         {
205                                 //TODO: maybe we should try to load the glyph on the fly if it is missing.
206                         }
207                 }
208
209                 if(normalize_)
210                 {
211                         float ratio_x = parent_width/(pos_x - x);
212                         float ratio_y = parent_height/(float)(maxHeight);
213                         for(index = 0; index < result.size(); index += 4)
214                         {
215                                 result[index + 0] *= ratio_x;
216                                 result[index + 1] *= ratio_y;
217                         }
218                 }
219
220                 if(metrics != nullptr)
221                 {
222                         metrics->width = (int)(pos_x - x + 0.5f);
223                         metrics->bearingY = maxBearingY;
224                         metrics->height = maxHeight;
225                 }
226                 return result;
227         }
228
229         string_metrics measure_string(const std::wstring& str)
230         {
231                 string_metrics result;
232                 
233                 bool use_kerning = (face_->face_flags & FT_FACE_FLAG_KERNING) == FT_FACE_FLAG_KERNING;
234                 int index = 0;
235                 FT_UInt previous = 0;
236                 float pos_x = 0;
237 //              float pos_y = 0;
238
239                 auto end = str.end();
240                 for(auto it = str.begin(); it != end; ++it, ++index)
241                 {
242                         auto glyph_it = glyphs_.find(*it);
243                         if(glyph_it != glyphs_.end())
244                         {       
245                                 const glyph_info& coords = glyph_it->second;
246
247                                 FT_UInt glyph_index = FT_Get_Char_Index(face_, (*it));
248
249                                 if(use_kerning && previous && glyph_index)
250                                 {
251                                         FT_Vector delta;
252                                         FT_Get_Kerning(face_, previous, glyph_index, FT_KERNING_DEFAULT, &delta);
253
254                                         pos_x += delta.x / 64.0f;
255                                 }
256
257                                 FT_Load_Glyph(face_, glyph_index, FT_LOAD_NO_BITMAP | FT_LOAD_FORCE_AUTOHINT | FT_LOAD_TARGET_NORMAL);
258
259                                 int bearingY = face_->glyph->metrics.horiBearingY >> 6;
260                                 if(bearingY > result.bearingY)
261                                         result.bearingY = bearingY;
262
263                                 if(coords.height > result.height)
264                                         result.height = coords.height;
265
266                                 pos_x += face_->glyph->advance.x / 64.0f;
267                                 previous = glyph_index;
268                         }
269                 }
270
271                 result.width = (int)(pos_x+.5f);
272                 return result;
273         }
274 }; 
275
276 texture_font::texture_font(texture_atlas& atlas, const std::wstring& filename, float size, bool normalize_coordinates) : impl_(new impl(atlas, filename, size, normalize_coordinates)) {}
277 void texture_font::load_glyphs(unicode_block range, const color<float>& col) { impl_->load_glyphs(range, col); }
278 std::vector<float> texture_font::create_vertex_stream(const std::wstring& str, int x, int y, int parent_width, int parent_height, string_metrics* metrics) { return impl_->create_vertex_stream(str, x, y, parent_width, parent_height, metrics); }
279 string_metrics texture_font::measure_string(const std::wstring& str) { return impl_->measure_string(str); }
280
281 unicode_range get_range(unicode_block block)
282 {
283         switch(block)
284         {
285                 case Basic_Latin: return unicode_range(0x0000, 0x007F); 
286                 case Latin_1_Supplement: return unicode_range(0x0080, 0x00FF); 
287                 case Latin_Extended_A: return           unicode_range(0x0100, 0x017F); 
288                 case Latin_Extended_B: return           unicode_range(0x0180, 0x024F); 
289                 case IPA_Extensions: return             unicode_range(0x0250, 0x02AF); 
290                 case Spacing_Modifier_Letters: return           unicode_range(0x02B0, 0x02FF); 
291                 case Combining_Diacritical_Marks: return                unicode_range(0x0300, 0x036F); 
292                 case Greek_and_Coptic: return           unicode_range(0x0370, 0x03FF); 
293                 case Cyrillic: return           unicode_range(0x0400, 0x04FF); 
294                 case Cyrillic_Supplement: return                unicode_range(0x0500, 0x052F); 
295                 case Armenian: return           unicode_range(0x0530, 0x058F); 
296                 case Hebrew: return             unicode_range(0x0590, 0x05FF); 
297                 case Arabic: return             unicode_range(0x0600, 0x06FF); 
298                 case Syriac: return             unicode_range(0x0700, 0x074F); 
299                 case Arabic_Supplement: return          unicode_range(0x0750, 0x077F); 
300                 case Thaana: return             unicode_range(0x0780, 0x07BF); 
301                 case NKo: return                unicode_range(0x07C0, 0x07FF); 
302                 case Samaritan: return          unicode_range(0x0800, 0x083F); 
303                 case Mandaic: return            unicode_range(0x0840, 0x085F); 
304                 case Arabic_Extended_A: return          unicode_range(0x08A0, 0x08FF); 
305                 case Devanagari: return         unicode_range(0x0900, 0x097F); 
306                 case Bengali: return            unicode_range(0x0980, 0x09FF); 
307                 case Gurmukhi: return           unicode_range(0x0A00, 0x0A7F); 
308                 case Gujarati: return           unicode_range(0x0A80, 0x0AFF); 
309                 case Oriya: return              unicode_range(0x0B00, 0x0B7F); 
310                 case Tamil: return              unicode_range(0x0B80, 0x0BFF); 
311                 case Telugu: return             unicode_range(0x0C00, 0x0C7F); 
312                 case Kannada: return            unicode_range(0x0C80, 0x0CFF); 
313                 case Malayalam: return          unicode_range(0x0D00, 0x0D7F); 
314                 case Sinhala: return            unicode_range(0x0D80, 0x0DFF); 
315                 case Thai: return               unicode_range(0x0E00, 0x0E7F); 
316                 case Lao: return                unicode_range(0x0E80, 0x0EFF); 
317                 case Tibetan: return            unicode_range(0x0F00, 0x0FFF); 
318                 case Myanmar: return            unicode_range(0x1000, 0x109F); 
319                 case Georgian: return           unicode_range(0x10A0, 0x10FF); 
320                 case Hangul_Jamo: return                unicode_range(0x1100, 0x11FF); 
321                 case Ethiopic: return           unicode_range(0x1200, 0x137F); 
322                 case Ethiopic_Supplement: return                unicode_range(0x1380, 0x139F); 
323                 case Cherokee: return           unicode_range(0x13A0, 0x13FF); 
324                 case Unified_Canadian_Aboriginal_Syllabics: return              unicode_range(0x1400, 0x167F); 
325                 case Ogham: return              unicode_range(0x1680, 0x169F); 
326                 case Runic: return              unicode_range(0x16A0, 0x16FF); 
327                 case Tagalog: return            unicode_range(0x1700, 0x171F); 
328                 case Hanunoo: return            unicode_range(0x1720, 0x173F); 
329                 case Buhid: return              unicode_range(0x1740, 0x175F); 
330                 case Tagbanwa: return           unicode_range(0x1760, 0x177F); 
331                 case Khmer: return              unicode_range(0x1780, 0x17FF); 
332                 case Mongolian: return          unicode_range(0x1800, 0x18AF); 
333                 case Unified_Canadian_Aboriginal_Syllabics_Extended: return             unicode_range(0x18B0, 0x18FF); 
334                 case Limbu: return              unicode_range(0x1900, 0x194F); 
335                 case Tai_Le: return             unicode_range(0x1950, 0x197F); 
336                 case New_Tai_Lue: return                unicode_range(0x1980, 0x19DF); 
337                 case Khmer_Symbols: return              unicode_range(0x19E0, 0x19FF); 
338                 case Buginese: return           unicode_range(0x1A00, 0x1A1F); 
339                 case Tai_Tham: return           unicode_range(0x1A20, 0x1AAF); 
340                 case Balinese: return           unicode_range(0x1B00, 0x1B7F); 
341                 case Sundanese: return          unicode_range(0x1B80, 0x1BBF); 
342                 case Batak: return              unicode_range(0x1BC0, 0x1BFF); 
343                 case Lepcha: return             unicode_range(0x1C00, 0x1C4F); 
344                 case Ol_Chiki: return           unicode_range(0x1C50, 0x1C7F); 
345                 case Sundanese_Supplement: return               unicode_range(0x1CC0, 0x1CCF); 
346                 case Vedic_Extensions: return           unicode_range(0x1CD0, 0x1CFF); 
347                 case Phonetic_Extensions: return                unicode_range(0x1D00, 0x1D7F); 
348                 case Phonetic_Extensions_Supplement: return             unicode_range(0x1D80, 0x1DBF); 
349                 case Combining_Diacritical_Marks_Supplement: return             unicode_range(0x1DC0, 0x1DFF); 
350                 case Latin_Extended_Additional: return          unicode_range(0x1E00, 0x1EFF); 
351                 case Greek_Extended: return             unicode_range(0x1F00, 0x1FFF); 
352                 case General_Punctuation: return                unicode_range(0x2000, 0x206F); 
353                 case Superscripts_and_Subscripts: return                unicode_range(0x2070, 0x209F); 
354                 case Currency_Symbols: return           unicode_range(0x20A0, 0x20CF); 
355                 case Combining_Diacritical_Marks_for_Symbols: return            unicode_range(0x20D0, 0x20FF); 
356                 case Letterlike_Symbols: return         unicode_range(0x2100, 0x214F); 
357                 case Number_Forms: return               unicode_range(0x2150, 0x218F); 
358                 case Arrows: return             unicode_range(0x2190, 0x21FF); 
359                 case Mathematical_Operators: return             unicode_range(0x2200, 0x22FF); 
360                 case Miscellaneous_Technical: return            unicode_range(0x2300, 0x23FF); 
361                 case Control_Pictures: return           unicode_range(0x2400, 0x243F); 
362                 case Optical_Character_Recognition: return              unicode_range(0x2440, 0x245F); 
363                 case Enclosed_Alphanumerics: return             unicode_range(0x2460, 0x24FF); 
364                 case Box_Drawing: return                unicode_range(0x2500, 0x257F); 
365                 case Block_Elements: return             unicode_range(0x2580, 0x259F); 
366                 case Geometric_Shapes: return           unicode_range(0x25A0, 0x25FF); 
367                 case Miscellaneous_Symbols: return              unicode_range(0x2600, 0x26FF); 
368                 case Dingbats: return           unicode_range(0x2700, 0x27BF); 
369                 case Miscellaneous_Mathematical_Symbols_A: return               unicode_range(0x27C0, 0x27EF); 
370                 case Supplemental_Arrows_A: return              unicode_range(0x27F0, 0x27FF); 
371                 case Braille_Patterns: return           unicode_range(0x2800, 0x28FF); 
372                 case Supplemental_Arrows_B: return              unicode_range(0x2900, 0x297F); 
373                 case Miscellaneous_Mathematical_Symbols_B: return               unicode_range(0x2980, 0x29FF); 
374                 case Supplemental_Mathematical_Operators: return                unicode_range(0x2A00, 0x2AFF); 
375                 case Miscellaneous_Symbols_and_Arrows: return           unicode_range(0x2B00, 0x2BFF); 
376                 case Glagolitic: return         unicode_range(0x2C00, 0x2C5F); 
377                 case Latin_Extended_C: return           unicode_range(0x2C60, 0x2C7F); 
378                 case Coptic: return             unicode_range(0x2C80, 0x2CFF); 
379                 case Georgian_Supplement: return                unicode_range(0x2D00, 0x2D2F); 
380                 case Tifinagh: return           unicode_range(0x2D30, 0x2D7F); 
381                 case Ethiopic_Extended: return          unicode_range(0x2D80, 0x2DDF); 
382                 case Cyrillic_Extended_A: return                unicode_range(0x2DE0, 0x2DFF); 
383                 case Supplemental_Punctuation: return           unicode_range(0x2E00, 0x2E7F); 
384                 case CJK_Radicals_Supplement: return            unicode_range(0x2E80, 0x2EFF); 
385                 case Kangxi_Radicals: return            unicode_range(0x2F00, 0x2FDF); 
386                 case Ideographic_Description_Characters: return         unicode_range(0x2FF0, 0x2FFF); 
387                 case CJK_Symbols_and_Punctuation: return                unicode_range(0x3000, 0x303F); 
388                 case Hiragana: return           unicode_range(0x3040, 0x309F); 
389                 case Katakana: return           unicode_range(0x30A0, 0x30FF); 
390                 case Bopomofo: return           unicode_range(0x3100, 0x312F); 
391                 case Hangul_Compatibility_Jamo: return          unicode_range(0x3130, 0x318F); 
392                 case Kanbun: return             unicode_range(0x3190, 0x319F); 
393                 case Bopomofo_Extended: return          unicode_range(0x31A0, 0x31BF); 
394                 case CJK_Strokes: return                unicode_range(0x31C0, 0x31EF); 
395                 case Katakana_Phonetic_Extensions: return               unicode_range(0x31F0, 0x31FF); 
396                 case Enclosed_CJK_Letters_and_Months: return            unicode_range(0x3200, 0x32FF); 
397                 case CJK_Compatibility: return          unicode_range(0x3300, 0x33FF); 
398                 case CJK_Unified_Ideographs_Extension_A: return         unicode_range(0x3400, 0x4DBF); 
399                 case Yijing_Hexagram_Symbols: return            unicode_range(0x4DC0, 0x4DFF); 
400                 case CJK_Unified_Ideographs: return             unicode_range(0x4E00, 0x9FFF); 
401                 case Yi_Syllables: return               unicode_range(0xA000, 0xA48F); 
402                 case Yi_Radicals: return                unicode_range(0xA490, 0xA4CF); 
403                 case Lisu: return               unicode_range(0xA4D0, 0xA4FF); 
404                 case Vai: return                unicode_range(0xA500, 0xA63F); 
405                 case Cyrillic_Extended_B: return                unicode_range(0xA640, 0xA69F); 
406                 case Bamum: return              unicode_range(0xA6A0, 0xA6FF); 
407                 case Modifier_Tone_Letters: return              unicode_range(0xA700, 0xA71F); 
408                 case Latin_Extended_D: return           unicode_range(0xA720, 0xA7FF); 
409                 case Syloti_Nagri: return               unicode_range(0xA800, 0xA82F); 
410                 case Common_Indic_Number_Forms: return          unicode_range(0xA830, 0xA83F); 
411                 case Phags_pa: return           unicode_range(0xA840, 0xA87F); 
412                 case Saurashtra: return         unicode_range(0xA880, 0xA8DF); 
413                 case Devanagari_Extended: return                unicode_range(0xA8E0, 0xA8FF); 
414                 case Kayah_Li: return           unicode_range(0xA900, 0xA92F); 
415                 case Rejang: return             unicode_range(0xA930, 0xA95F); 
416                 case Hangul_Jamo_Extended_A: return             unicode_range(0xA960, 0xA97F); 
417                 case Javanese: return           unicode_range(0xA980, 0xA9DF); 
418                 case Cham: return               unicode_range(0xAA00, 0xAA5F); 
419                 case Myanmar_Extended_A: return         unicode_range(0xAA60, 0xAA7F); 
420                 case Tai_Viet: return           unicode_range(0xAA80, 0xAADF); 
421                 case Meetei_Mayek_Extensions: return            unicode_range(0xAAE0, 0xAAFF); 
422                 case Ethiopic_Extended_A: return                unicode_range(0xAB00, 0xAB2F); 
423                 case Meetei_Mayek: return               unicode_range(0xABC0, 0xABFF); 
424                 case Hangul_Syllables: return           unicode_range(0xAC00, 0xD7AF); 
425                 case Hangul_Jamo_Extended_B: return             unicode_range(0xD7B0, 0xD7FF); 
426                 case High_Surrogates: return            unicode_range(0xD800, 0xDB7F); 
427                 case High_Private_Use_Surrogates: return                unicode_range(0xDB80, 0xDBFF); 
428                 case Low_Surrogates: return             unicode_range(0xDC00, 0xDFFF); 
429                 case Private_Use_Area: return           unicode_range(0xE000, 0xF8FF); 
430                 case CJK_Compatibility_Ideographs: return               unicode_range(0xF900, 0xFAFF); 
431                 case Alphabetic_Presentation_Forms: return              unicode_range(0xFB00, 0xFB4F); 
432                 case Arabic_Presentation_Forms_A: return                unicode_range(0xFB50, 0xFDFF); 
433                 case Variation_Selectors: return                unicode_range(0xFE00, 0xFE0F); 
434                 case Vertical_Forms: return             unicode_range(0xFE10, 0xFE1F); 
435                 case Combining_Half_Marks: return               unicode_range(0xFE20, 0xFE2F); 
436                 case CJK_Compatibility_Forms: return            unicode_range(0xFE30, 0xFE4F); 
437                 case Small_Form_Variants: return                unicode_range(0xFE50, 0xFE6F); 
438                 case Arabic_Presentation_Forms_B: return                unicode_range(0xFE70, 0xFEFF); 
439                 case Halfwidth_and_Fullwidth_Forms: return              unicode_range(0xFF00, 0xFFEF); 
440                 case Specials: return           unicode_range(0xFFF0, 0xFFFF); 
441                 case Linear_B_Syllabary: return         unicode_range(0x10000, 0x1007F); 
442                 case Linear_B_Ideograms: return         unicode_range(0x10080, 0x100FF); 
443                 case Aegean_Numbers: return             unicode_range(0x10100, 0x1013F); 
444                 case Ancient_Greek_Numbers: return              unicode_range(0x10140, 0x1018F); 
445                 case Ancient_Symbols: return            unicode_range(0x10190, 0x101CF); 
446                 case Phaistos_Disc: return              unicode_range(0x101D0, 0x101FF); 
447                 case Lycian: return             unicode_range(0x10280, 0x1029F); 
448                 case Carian: return             unicode_range(0x102A0, 0x102DF); 
449                 case Old_Italic: return         unicode_range(0x10300, 0x1032F); 
450                 case Gothic: return             unicode_range(0x10330, 0x1034F); 
451                 case Ugaritic: return           unicode_range(0x10380, 0x1039F); 
452                 case Old_Persian: return                unicode_range(0x103A0, 0x103DF); 
453                 case Deseret: return            unicode_range(0x10400, 0x1044F); 
454                 case Shavian: return            unicode_range(0x10450, 0x1047F); 
455                 case Osmanya: return            unicode_range(0x10480, 0x104AF); 
456                 case Cypriot_Syllabary: return          unicode_range(0x10800, 0x1083F); 
457                 case Imperial_Aramaic: return           unicode_range(0x10840, 0x1085F); 
458                 case Phoenician: return         unicode_range(0x10900, 0x1091F); 
459                 case Lydian: return             unicode_range(0x10920, 0x1093F); 
460                 case Meroitic_Hieroglyphs: return               unicode_range(0x10980, 0x1099F); 
461                 case Meroitic_Cursive: return           unicode_range(0x109A0, 0x109FF); 
462                 case Kharoshthi: return         unicode_range(0x10A00, 0x10A5F); 
463                 case Old_South_Arabian: return          unicode_range(0x10A60, 0x10A7F); 
464                 case Avestan: return            unicode_range(0x10B00, 0x10B3F); 
465                 case Inscriptional_Parthian: return             unicode_range(0x10B40, 0x10B5F); 
466                 case Inscriptional_Pahlavi: return              unicode_range(0x10B60, 0x10B7F); 
467                 case Old_Turkic: return         unicode_range(0x10C00, 0x10C4F); 
468                 case Rumi_Numeral_Symbols: return               unicode_range(0x10E60, 0x10E7F); 
469                 case Brahmi: return             unicode_range(0x11000, 0x1107F); 
470                 case Kaithi: return             unicode_range(0x11080, 0x110CF); 
471                 case Sora_Sompeng: return               unicode_range(0x110D0, 0x110FF); 
472                 case Chakma: return             unicode_range(0x11100, 0x1114F); 
473                 case Sharada: return            unicode_range(0x11180, 0x111DF); 
474                 case Takri: return              unicode_range(0x11680, 0x116CF); 
475                 case Cuneiform: return          unicode_range(0x12000, 0x123FF); 
476                 case Cuneiform_Numbers_and_Punctuation: return          unicode_range(0x12400, 0x1247F); 
477                 case Egyptian_Hieroglyphs: return               unicode_range(0x13000, 0x1342F); 
478                 case Bamum_Supplement: return           unicode_range(0x16800, 0x16A3F); 
479                 case Miao: return               unicode_range(0x16F00, 0x16F9F); 
480                 case Kana_Supplement: return            unicode_range(0x1B000, 0x1B0FF); 
481                 case Byzantine_Musical_Symbols: return          unicode_range(0x1D000, 0x1D0FF); 
482                 case Musical_Symbols: return            unicode_range(0x1D100, 0x1D1FF); 
483                 case Ancient_Greek_Musical_Notation: return             unicode_range(0x1D200, 0x1D24F); 
484                 case Tai_Xuan_Jing_Symbols: return              unicode_range(0x1D300, 0x1D35F); 
485                 case Counting_Rod_Numerals: return              unicode_range(0x1D360, 0x1D37F); 
486                 case Mathematical_Alphanumeric_Symbols: return          unicode_range(0x1D400, 0x1D7FF); 
487                 case Arabic_Mathematical_Alphabetic_Symbols: return             unicode_range(0x1EE00, 0x1EEFF); 
488                 case Mahjong_Tiles: return              unicode_range(0x1F000, 0x1F02F); 
489                 case Domino_Tiles: return               unicode_range(0x1F030, 0x1F09F); 
490                 case Playing_Cards: return              unicode_range(0x1F0A0, 0x1F0FF); 
491                 case Enclosed_Alphanumeric_Supplement: return           unicode_range(0x1F100, 0x1F1FF); 
492                 case Enclosed_Ideographic_Supplement: return            unicode_range(0x1F200, 0x1F2FF); 
493                 case Miscellaneous_Symbols_And_Pictographs: return              unicode_range(0x1F300, 0x1F5FF); 
494                 case Emoticons: return          unicode_range(0x1F600, 0x1F64F); 
495                 case Transport_And_Map_Symbols: return          unicode_range(0x1F680, 0x1F6FF); 
496                 case Alchemical_Symbols: return         unicode_range(0x1F700, 0x1F77F); 
497                 case CJK_Unified_Ideographs_Extension_B: return         unicode_range(0x20000, 0x2A6DF); 
498                 case CJK_Unified_Ideographs_Extension_C: return         unicode_range(0x2A700, 0x2B73F); 
499                 case CJK_Unified_Ideographs_Extension_D: return         unicode_range(0x2B740, 0x2B81F); 
500                 case CJK_Compatibility_Ideographs_Supplement: return            unicode_range(0x2F800, 0x2FA1F); 
501                 case Tags: return               unicode_range(0xE0000, 0xE007F); 
502                 case Variation_Selectors_Supplement: return             unicode_range(0xE0100, 0xE01EF); 
503                 case Supplementary_Private_Use_Area_A: return           unicode_range(0xF0000, 0xFFFFF); 
504                 case Supplementary_Private_Use_Area_B: return           unicode_range(0x100000, 0x10FFFF);
505         }
506         return unicode_range(0,0);
507 }
508
509 }}}