3 * This file is part of LyX, the document processor.
4 * Licence details can be found in the file COPYING.
6 * \author Lars Gullik Bjønnes
7 * \author Jean-Marc Lasgouttes
10 * Full author contact details are available in file CREDITS.
17 #include "LaTeXFeatures.h"
21 #include "support/debug.h"
22 #include "support/FileName.h"
23 #include "support/lstrings.h"
24 #include "support/unicode.h"
26 #include <boost/cstdint.hpp>
31 using namespace lyx::support;
39 char_type arabic_table[172][4] = {
40 {0xfe80, 0xfe80, 0xfe80, 0xfe80}, // 0x0621 = hamza
41 {0xfe81, 0xfe82, 0xfe81, 0xfe82}, // 0x0622 = ligature madda on alef
42 {0xfe83, 0xfe84, 0xfe83, 0xfe84}, // 0x0623 = ligature hamza on alef
43 {0xfe85, 0xfe86, 0xfe85, 0xfe86}, // 0x0624 = ligature hamza on waw
44 {0xfe87, 0xfe88, 0xfe87, 0xfe88}, // 0x0625 = ligature hamza under alef
45 {0xfe89, 0xfe8a, 0xfe8b, 0xfe8c}, // 0x0626 = ligature hamza on ya
46 {0xfe8d, 0xfe8e, 0xfe8d, 0xfe8e}, // 0x0627 = alef
47 {0xfe8f, 0xfe90, 0xfe91, 0xfe92}, // 0x0628 = baa
48 {0xfe93, 0xfe94, 0xfe93, 0xfe94}, // 0x0629 = taa marbuta
49 {0xfe95, 0xfe96, 0xfe97, 0xfe98}, // 0x062a = taa
50 {0xfe99, 0xfe9a, 0xfe9b, 0xfe9c}, // 0x062b = thaa
51 {0xfe9d, 0xfe9e, 0xfe9f, 0xfea0}, // 0x062c = jeem
52 {0xfea1, 0xfea2, 0xfea3, 0xfea4}, // 0x062d = haa
53 {0xfea5, 0xfea6, 0xfea7, 0xfea8}, // 0x062e = khaa
54 {0xfea9, 0xfeaa, 0xfea9, 0xfeaa}, // 0x062f = dal
56 {0xfeab, 0xfeac, 0xfeab, 0xfeac}, // 0x0630 = thal
57 {0xfead, 0xfeae, 0xfead, 0xfeae}, // 0x0631 = ra
58 {0xfeaf, 0xfeb0, 0xfeaf, 0xfeb0}, // 0x0632 = zain
59 {0xfeb1, 0xfeb2, 0xfeb3, 0xfeb4}, // 0x0633 = seen
60 {0xfeb5, 0xfeb6, 0xfeb7, 0xfeb8}, // 0x0634 = sheen
61 {0xfeb9, 0xfeba, 0xfebb, 0xfebc}, // 0x0635 = sad
62 {0xfebd, 0xfebe, 0xfebf, 0xfec0}, // 0x0636 = dad
63 {0xfec1, 0xfec2, 0xfec3, 0xfec4}, // 0x0637 = tah
64 {0xfec5, 0xfec6, 0xfec7, 0xfec8}, // 0x0638 = zah
65 {0xfec9, 0xfeca, 0xfecb, 0xfecc}, // 0x0639 = ain
66 {0xfecd, 0xfece, 0xfecf, 0xfed0}, // 0x063a = ghain
67 {0, 0, 0, 0}, // 0x063b
68 {0, 0, 0, 0}, // 0x063c
69 {0, 0, 0, 0}, // 0x063d
70 {0, 0, 0, 0}, // 0x063e
71 {0, 0, 0, 0}, // 0x063f
73 {0, 0, 0, 0}, // 0x0640
74 {0xfed1, 0xfed2, 0xfed3, 0xfed4}, // 0x0641 = fa
75 {0xfed5, 0xfed6, 0xfed7, 0xfed8}, // 0x0642 = qaf
76 {0xfed9, 0xfeda, 0xfedb, 0xfedc}, // 0x0643 = kaf
77 {0xfedd, 0xfede, 0xfedf, 0xfee0}, // 0x0644 = lam
78 {0xfee1, 0xfee2, 0xfee3, 0xfee4}, // 0x0645 = meem
79 {0xfee5, 0xfee6, 0xfee7, 0xfee8}, // 0x0646 = noon
80 {0xfee9, 0xfeea, 0xfeeb, 0xfeec}, // 0x0647 = ha
81 {0xfeed, 0xfeee, 0xfeed, 0xfeee}, // 0x0648 = waw
82 {0xfeef, 0xfef0, 0xfeef, 0xfef0}, // 0x0649 = alef maksura
83 {0xfef1, 0xfef2, 0xfef3, 0xfef4}, // 0x064a = ya
84 {0x065b, 0x065b, 0x065b, 0x065b}, // 0x064b = fathatan
85 {0x065c, 0x065c, 0x065c, 0x065c}, // 0x064c = dammatan
86 {0x064d, 0x064d, 0x064d, 0x064d}, // 0x064d = kasratan
87 {0x064e, 0x064e, 0x064e, 0x064e}, // 0x064e = fatha
88 {0x064f, 0x064f, 0x064f, 0x064f}, // 0x064f = damma
90 {0x0650, 0x0650, 0x0650, 0x0650}, // 0x0650 = kasra
91 {0x0651, 0x0651, 0x0651, 0x0651}, // 0x0651 = shadda
92 {0x0652, 0x0652, 0x0652, 0x0652}, // 0x0652 = sukun
94 {0, 0, 0, 0}, // 0x0653
95 {0, 0, 0, 0}, // 0x0654
96 {0, 0, 0, 0}, // 0x0655
97 {0, 0, 0, 0}, // 0x0656
98 {0, 0, 0, 0}, // 0x0657
99 {0, 0, 0, 0}, // 0x0658
100 {0, 0, 0, 0}, // 0x0659
101 {0, 0, 0, 0}, // 0x065a
102 {0, 0, 0, 0}, // 0x065b
103 {0, 0, 0, 0}, // 0x065c
104 {0, 0, 0, 0}, // 0x065d
105 {0, 0, 0, 0}, // 0x065e
106 {0, 0, 0, 0}, // 0x065f
107 {0, 0, 0, 0}, // 0x0660
108 {0, 0, 0, 0}, // 0x0661
109 {0, 0, 0, 0}, // 0x0662
110 {0, 0, 0, 0}, // 0x0663
111 {0, 0, 0, 0}, // 0x0664
112 {0, 0, 0, 0}, // 0x0665
113 {0, 0, 0, 0}, // 0x0666
114 {0, 0, 0, 0}, // 0x0667
115 {0, 0, 0, 0}, // 0x0668
116 {0, 0, 0, 0}, // 0x0669
117 {0, 0, 0, 0}, // 0x066a
118 {0, 0, 0, 0}, // 0x066b
119 {0, 0, 0, 0}, // 0x066c
120 {0, 0, 0, 0}, // 0x066d
121 {0, 0, 0, 0}, // 0x066e
122 {0, 0, 0, 0}, // 0x066f
123 {0, 0, 0, 0}, // 0x0670
124 {0, 0, 0, 0}, // 0x0671
125 {0, 0, 0, 0}, // 0x0672
126 {0, 0, 0, 0}, // 0x0673
127 {0, 0, 0, 0}, // 0x0674
128 {0, 0, 0, 0}, // 0x0675
129 {0, 0, 0, 0}, // 0x0676
130 {0, 0, 0, 0}, // 0x0677
131 {0, 0, 0, 0}, // 0x0678
132 {0, 0, 0, 0}, // 0x0679
133 {0, 0, 0, 0}, // 0x067a
134 {0, 0, 0, 0}, // 0x067b
135 {0, 0, 0, 0}, // 0x067c
136 {0, 0, 0, 0}, // 0x067d
137 {0xfb56, 0xfb57, 0xfb58, 0xfb59}, // 0x067e = peh
138 {0, 0, 0, 0}, // 0x067f
139 {0, 0, 0, 0}, // 0x0680
140 {0, 0, 0, 0}, // 0x0681
141 {0, 0, 0, 0}, // 0x0682
142 {0, 0, 0, 0}, // 0x0683
143 {0, 0, 0, 0}, // 0x0684
144 {0, 0, 0, 0}, // 0x0685
145 {0xfb7a, 0xfb7b, 0xfb7c, 0xfb7d}, // 0x0686 = tcheh
146 {0, 0, 0, 0}, // 0x0687
147 {0, 0, 0, 0}, // 0x0688
148 {0, 0, 0, 0}, // 0x0689
149 {0, 0, 0, 0}, // 0x068a
150 {0, 0, 0, 0}, // 0x068b
151 {0, 0, 0, 0}, // 0x068c
152 {0, 0, 0, 0}, // 0x068d
153 {0, 0, 0, 0}, // 0x068e
154 {0, 0, 0, 0}, // 0x068f
155 {0, 0, 0, 0}, // 0x0690
156 {0, 0, 0, 0}, // 0x0691
157 {0, 0, 0, 0}, // 0x0692
158 {0, 0, 0, 0}, // 0x0693
159 {0, 0, 0, 0}, // 0x0694
160 {0, 0, 0, 0}, // 0x0695
161 {0, 0, 0, 0}, // 0x0696
162 {0, 0, 0, 0}, // 0x0697
163 {0xfb8a, 0xfb8b, 0xfb8a, 0xfb8b}, // 0x0698 = jeh
164 {0, 0, 0, 0}, // 0x0699
165 {0, 0, 0, 0}, // 0x069a
166 {0, 0, 0, 0}, // 0x069b
167 {0, 0, 0, 0}, // 0x069c
168 {0, 0, 0, 0}, // 0x069d
169 {0, 0, 0, 0}, // 0x069e
170 {0, 0, 0, 0}, // 0x069f
171 {0, 0, 0, 0}, // 0x06a0
172 {0, 0, 0, 0}, // 0x06a1
173 {0, 0, 0, 0}, // 0x06a2
174 {0, 0, 0, 0}, // 0x06a3
175 {0, 0, 0, 0}, // 0x06a4
176 {0, 0, 0, 0}, // 0x06a5
177 {0, 0, 0, 0}, // 0x06a6
178 {0, 0, 0, 0}, // 0x06a7
179 {0, 0, 0, 0}, // 0x06a8
180 {0xfb8e, 0xfb8f, 0xfb90, 0xfb91}, // 0x06a9 = farsi kaf
181 {0, 0, 0, 0}, // 0x06aa
182 {0, 0, 0, 0}, // 0x06ab
183 {0, 0, 0, 0}, // 0x06ac
184 {0, 0, 0, 0}, // 0x06ad
185 {0, 0, 0, 0}, // 0x06ae
186 {0xfb92, 0xfb93, 0xfb94, 0xfb95}, // 0x06af = gaf
187 {0, 0, 0, 0}, // 0x06b0
188 {0, 0, 0, 0}, // 0x06b1
189 {0, 0, 0, 0}, // 0x06b2
190 {0, 0, 0, 0}, // 0x06b3
191 {0, 0, 0, 0}, // 0x06b4
192 {0, 0, 0, 0}, // 0x06b5
193 {0, 0, 0, 0}, // 0x06b6
194 {0, 0, 0, 0}, // 0x06b7
195 {0, 0, 0, 0}, // 0x06b8
196 {0, 0, 0, 0}, // 0x06b9
197 {0, 0, 0, 0}, // 0x06ba
198 {0, 0, 0, 0}, // 0x06bb
199 {0, 0, 0, 0}, // 0x06bc
200 {0, 0, 0, 0}, // 0x06bd
201 {0, 0, 0, 0}, // 0x06be
202 {0, 0, 0, 0}, // 0x06bf
203 {0, 0, 0, 0}, // 0x06c0
204 {0, 0, 0, 0}, // 0x06c1
205 {0, 0, 0, 0}, // 0x06c2
206 {0, 0, 0, 0}, // 0x06c3
207 {0, 0, 0, 0}, // 0x06c4
208 {0, 0, 0, 0}, // 0x06c5
209 {0, 0, 0, 0}, // 0x06c6
210 {0, 0, 0, 0}, // 0x06c7
211 {0, 0, 0, 0}, // 0x06c8
212 {0, 0, 0, 0}, // 0x06c9
213 {0, 0, 0, 0}, // 0x06ca
214 {0, 0, 0, 0}, // 0x06cb
215 {0xfbfc, 0xfbfd, 0xfbfe, 0xfbff} // 0x06cc = farsi yeh
219 char_type const arabic_start = 0x0621;
220 char_type const arabic_end = 0x06cc;
223 /// Information about a single UCS4 character
225 /// LaTeX command (text mode) for this character
226 docstring textcommand;
227 /// LaTeX command (math mode) for this character
228 docstring mathcommand;
229 /// Needed LaTeX preamble (or feature) for text mode
231 /// Needed LaTeX preamble (or feature) for math mode
233 /// Is this a combining character?
235 /// Is \c textpreamble a feature known by LaTeXFeatures, or a raw LaTeX
238 /// Is \c mathpreamble a feature known by LaTeXFeatures, or a raw LaTeX
241 /// Always force the LaTeX command, even if the encoding contains
247 typedef map<char_type, CharInfo> CharInfoMap;
248 CharInfoMap unicodesymbols;
250 typedef std::set<char_type> CharSet;
254 /// The highest code point in UCS4 encoding (1<<20 + 1<<16)
255 char_type const max_ucs4 = 0x110000;
260 EncodingException::EncodingException(char_type c)
261 : failed_char(c), par_id(0), pos(0)
266 const char * EncodingException::what() const throw()
268 return "Could not find LaTeX command for a character";
272 Encoding::Encoding(string const & n, string const & l, string const & g,
273 string const & i, bool f, Encoding::Package p)
274 : name_(n), latexName_(l), guiName_(g), iconvName_(i), fixedwidth_(f), package_(p)
277 // ASCII can encode 128 code points and nothing else
278 start_encodable_ = 128;
280 } else if (i == "UTF-8") {
281 // UTF8 can encode all UCS4 code points
282 start_encodable_ = max_ucs4;
290 void Encoding::init() const
295 start_encodable_ = 0;
296 // temporarily switch off lyxerr, since we will generate iconv errors
299 // We do not need to check all UCS4 code points, it is enough
300 // if we check all 256 code points of this encoding.
301 for (unsigned short j = 0; j < 256; ++j) {
302 char const c = char(j);
303 vector<char_type> const ucs4 = eightbit_to_ucs4(&c, 1, iconvName_);
304 if (ucs4.size() != 1)
306 char_type const uc = ucs4[0];
307 CharInfoMap::const_iterator const it = unicodesymbols.find(uc);
308 if (it == unicodesymbols.end() || !it->second.force)
309 encodable_.insert(uc);
312 // We do not know how many code points this encoding has, and
313 // they do not have a direct representation as a single byte,
314 // therefore we need to check all UCS4 code points.
315 // This is expensive!
316 for (char_type c = 0; c < max_ucs4; ++c) {
317 vector<char> const eightbit = ucs4_to_eightbit(&c, 1, iconvName_);
318 if (!eightbit.empty()) {
319 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
320 if (it == unicodesymbols.end() || !it->second.force)
321 encodable_.insert(c);
326 CharSet::iterator it = encodable_.find(start_encodable_);
327 while (it != encodable_.end()) {
328 encodable_.erase(it);
330 it = encodable_.find(start_encodable_);
336 docstring Encoding::latexChar(char_type c, bool for_mathed) const
338 // assure the used encoding is properly initialized
341 if (iconvName_ == "UTF-8" && package_ == none)
342 return docstring(1, c);
343 if (c < start_encodable_ && !encodings.isForced(c))
344 return docstring(1, c);
345 if (encodable_.find(c) != encodable_.end())
346 return docstring(1, c);
350 // c cannot (or should not) be encoded in this encoding
351 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
352 if (it == unicodesymbols.end())
353 throw EncodingException(c);
354 // at least one of mathcommand and textcommand is nonempty
355 if (it->second.textcommand.empty())
356 return "\\ensuremath{" + it->second.mathcommand + '}';
357 return it->second.textcommand;
361 vector<char_type> Encoding::symbolsList() const
363 // assure the used encoding is properly initialized
366 // first all encodable characters
367 vector<char_type> symbols(encodable_.begin(), encodable_.end());
368 // add those below start_encodable_
369 for (char_type c = 0; c < start_encodable_; ++c)
370 symbols.push_back(c);
371 // now the ones from the unicodesymbols file
372 CharInfoMap::const_iterator const end = unicodesymbols.end();
373 CharInfoMap::const_iterator it = unicodesymbols.begin();
374 for (; it != end; ++it)
375 symbols.push_back(it->first);
380 bool Encodings::latexMathChar(char_type c, Encoding const * encoding,
384 command = encoding->latexChar(c, true);
385 if (!command.empty())
388 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
389 if (it == unicodesymbols.end())
390 throw EncodingException(c);
391 if (it->second.mathcommand.empty()) {
392 if (it->second.textcommand.empty())
393 throw EncodingException(c);
394 command = it->second.textcommand;
397 command = it->second.mathcommand;
402 char_type Encodings::fromLaTeXCommand(docstring const & cmd, bool & combining)
404 CharInfoMap::const_iterator const end = unicodesymbols.end();
405 CharInfoMap::const_iterator it = unicodesymbols.begin();
406 for (; it != end; ++it) {
407 docstring const math = it->second.mathcommand;
408 docstring const text = it->second.textcommand;
409 if (math == cmd || text == cmd) {
410 combining = it->second.combining;
418 docstring Encodings::fromLaTeXCommand(docstring const & cmd, docstring & rem)
422 size_t const cmdend = cmd.size();
423 CharInfoMap::const_iterator const uniend = unicodesymbols.end();
424 for (size_t j = 0; j < cmdend; ++j) {
425 // Also get the char after a backslash
426 if (j + 1 < cmdend && cmd[j] == '\\')
428 // If a macro argument follows, get it, too
429 if (j + 1 < cmdend && cmd[j + 1] == '{') {
432 while (k < cmdend && count && k != docstring::npos) {
433 k = cmd.find_first_of(from_ascii("{}"), k + 1);
439 if (k != docstring::npos)
442 // Start with this substring and try augmenting it when it is
443 // the prefix of some command in the unicodesymbols file
444 docstring const subcmd = cmd.substr(i, j - i + 1);
446 CharInfoMap::const_iterator it = unicodesymbols.begin();
447 size_t unicmd_size = 0;
449 for (; it != uniend; ++it) {
450 docstring const math = it->second.mathcommand;
451 docstring const text = it->second.textcommand;
452 size_t cur_size = max(math.size(), text.size());
453 // The current math or text unicode command cannot
454 // match, or we already matched a longer one
455 if (cur_size < subcmd.size() || cur_size <= unicmd_size)
458 docstring tmp = subcmd;
460 while (prefixIs(math, tmp) || prefixIs(text, tmp)) {
462 if (k >= cmdend || cur_size <= tmp.size())
470 // The last added char caused a mismatch, because
471 // we didn't exhaust the chars in cmd and didn't
472 // exceed the maximum size of the current unicmd
473 if (k < cmdend && cur_size > tmp.size())
474 tmp.resize(tmp.size() - 1);
476 // If this is an exact match, we found a (longer)
477 // matching command in the unicodesymbols file
478 if (math == tmp || text == tmp) {
482 unicmd_size = cur_size;
487 else if (j + 1 == cmdend)
488 // No luck. Return what remains
495 void Encodings::validate(char_type c, LaTeXFeatures & features, bool for_mathed)
497 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
498 if (it != unicodesymbols.end()) {
499 // at least one of mathcommand and textcommand is nonempty
500 bool const use_math = (for_mathed && !it->second.mathcommand.empty()) ||
501 (!for_mathed && it->second.textcommand.empty());
503 if (!it->second.mathpreamble.empty()) {
504 if (it->second.mathfeature)
505 features.require(it->second.mathpreamble);
507 features.addPreambleSnippet(it->second.mathpreamble);
510 if (!it->second.textpreamble.empty()) {
511 if (it->second.textfeature)
512 features.require(it->second.textpreamble);
514 features.addPreambleSnippet(it->second.textpreamble);
517 features.require("relsize");
518 features.require("lyxmathsym");
525 bool Encodings::isHebrewComposeChar(char_type c)
527 return c <= 0x05c2 && c >= 0x05b0 && c != 0x05be && c != 0x05c0;
531 // Special Arabic letters are ones that do not get connected from left
532 // they are hamza, alef_madda, alef_hamza, waw_hamza, alef_hamza_under,
533 // alef, tah_marbota, dal, thal, rah, zai, wow, alef_maksoura
535 bool Encodings::isArabicSpecialChar(char_type c)
537 return (c >= 0x0621 && c <= 0x0625) || (c >= 0x0630 && c <= 0x0632)
538 || c == 0x0627 || c == 0x0629 || c == 0x062f || c == 0x0648
539 || c == 0x0649 || c == 0x0698;
543 bool Encodings::isArabicComposeChar(char_type c)
545 return c >= 0x064b && c <= 0x0652;
549 bool Encodings::isArabicChar(char_type c)
551 return c >= arabic_start && c <= arabic_end
552 && arabic_table[c-arabic_start][0];
556 char_type Encodings::transformChar(char_type c, Encodings::LetterForm form)
558 return isArabicChar(c) ? arabic_table[c-arabic_start][form] : c;
562 bool Encodings::isCombiningChar(char_type c)
564 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
565 if (it != unicodesymbols.end())
566 return it->second.combining;
571 bool Encodings::isKnownScriptChar(char_type const c, string & preamble)
573 CharInfoMap::const_iterator const it = unicodesymbols.find(c);
575 if (it == unicodesymbols.end())
578 if (it->second.textpreamble != "textgreek" && it->second.textpreamble != "textcyr")
581 if (preamble.empty()) {
582 preamble = it->second.textpreamble;
585 return it->second.textpreamble == preamble;
589 bool Encodings::isForced(char_type c)
591 return (!forced.empty() && forced.find(c) != forced.end());
595 Encoding const * Encodings::fromLyXName(string const & name) const
597 EncodingList::const_iterator const it = encodinglist.find(name);
598 return it != encodinglist.end() ? &it->second : 0;
602 Encoding const * Encodings::fromLaTeXName(string const & name) const
604 // We don't use find_if because it makes copies of the pairs in
606 // This linear search is OK since we don't have many encodings.
607 // Users could even optimize it by putting the encodings they use
608 // most at the top of lib/encodings.
609 EncodingList::const_iterator const end = encodinglist.end();
610 for (EncodingList::const_iterator it = encodinglist.begin(); it != end; ++it)
611 if (it->second.latexName() == name)
617 Encodings::Encodings()
622 void Encodings::read(FileName const & encfile, FileName const & symbolsfile)
624 // We must read the symbolsfile first, because the Encoding
625 // constructor depends on it.
627 symbolslex.setFile(symbolsfile);
628 bool getNextToken = true;
629 while (symbolslex.isOK()) {
635 if (!symbolslex.next(true))
640 istringstream is(symbolslex.getString());
641 // reading symbol directly does not work if
642 // char_type == wchar_t.
644 if(!(is >> hex >> tmp))
648 if (!symbolslex.next(true))
650 info.textcommand = symbolslex.getDocString();
651 if (!symbolslex.next(true))
653 info.textpreamble = symbolslex.getString();
654 if (!symbolslex.next(true))
656 flags = symbolslex.getString();
658 info.combining = false;
659 info.textfeature = false;
661 while (!flags.empty()) {
663 flags = split(flags, flag, ',');
664 if (flag == "combining")
665 info.combining = true;
666 else if (flag == "force") {
668 forced.insert(symbol);
670 lyxerr << "Ignoring unknown flag `" << flag
671 << "' for symbol `0x"
672 << hex << symbol << dec
675 // mathcommand and mathpreamble have been added for 1.6.0.
676 // make them optional so that old files still work.
677 int const lineno = symbolslex.lineNumber();
678 bool breakout = false;
679 if (symbolslex.next(true)) {
680 if (symbolslex.lineNumber() != lineno) {
681 // line in old format without mathcommand and mathpreamble
682 getNextToken = false;
684 info.mathcommand = symbolslex.getDocString();
685 if (symbolslex.next(true)) {
686 if (symbolslex.lineNumber() != lineno) {
687 // line in new format with mathcommand only
688 getNextToken = false;
690 // line in new format with mathcommand and mathpreamble
691 info.mathpreamble = symbolslex.getString();
700 if (!info.textpreamble.empty())
701 info.textfeature = info.textpreamble[0] != '\\';
702 if (!info.mathpreamble.empty())
703 info.mathfeature = info.mathpreamble[0] != '\\';
705 LYXERR(Debug::INFO, "Read unicode symbol " << symbol << " '"
706 << to_utf8(info.textcommand) << "' '" << info.textpreamble
707 << "' " << info.combining << ' ' << info.textfeature
708 << " '" << to_utf8(info.mathcommand) << "' '"
709 << info.mathpreamble << "' " << info.mathfeature);
711 // we assume that at least one command is nonempty when using unicodesymbols
712 if (!info.textcommand.empty() || !info.mathcommand.empty())
713 unicodesymbols[symbol] = info;
719 // Now read the encodings
725 LexerKeyword encodingtags[] = {
726 { "encoding", et_encoding },
730 Lexer lex(encodingtags);
731 lex.setFile(encfile);
732 lex.setContext("Encodings::read");
738 string const name = lex.getString();
740 string const latexname = lex.getString();
742 string const guiname = lex.getString();
744 string const iconvname = lex.getString();
746 string const width = lex.getString();
747 bool fixedwidth = false;
748 if (width == "fixed")
750 else if (width == "variable")
753 lex.printError("Unknown width");
756 string const p = lex.getString();
757 Encoding::Package package = Encoding::none;
759 package = Encoding::none;
760 else if (p == "inputenc")
761 package = Encoding::inputenc;
763 package = Encoding::CJK;
764 else if (p == "japanese")
765 package = Encoding::japanese;
767 lex.printError("Unknown package");
769 LYXERR(Debug::INFO, "Reading encoding " << name);
770 encodinglist[name] = Encoding(name, latexname,
771 guiname, iconvname, fixedwidth, package);
773 if (lex.lex() != et_end)
774 lex.printError("Missing end");
778 lex.printError("Misplaced end");
780 case Lexer::LEX_FEOF:
783 lex.printError("Unknown tag");