1 /* This file is part of
2 * ======================================================
4 * LyX, The Document Processor
6 * Copyright 2001 The LyX Team.
8 * ======================================================
11 * \author Angus Leeming <a.leeming@ic.ac.uk>
12 * \author Herbert Voss <voss@perce.de>
18 #pragma implementation
23 #include "gettext.h" // for _()
24 #include "helper_funcs.h"
25 #include "support/lstrings.h"
26 #include "support/LAssert.h"
28 #include <boost/regex.hpp>
42 using namespace biblio;
44 char const * const citeCommands[] = {
45 "cite", "citet", "citep", "citealt", "citealp", "citeauthor",
46 "citeyear", "citeyearpar" };
48 unsigned int const nCiteCommands =
49 sizeof(citeCommands) / sizeof(char *);
51 CiteStyle const citeStyles[] = {
52 CITE, CITET, CITEP, CITEALT, CITEALP,
53 CITEAUTHOR, CITEYEAR, CITEYEARPAR };
55 unsigned int const nCiteStyles =
56 sizeof(citeStyles) / sizeof(CiteStyle);
58 CiteStyle const citeStylesFull[] = {
59 CITET, CITEP, CITEALT, CITEALP, CITEAUTHOR };
61 unsigned int const nCiteStylesFull =
62 sizeof(citeStylesFull) / sizeof(CiteStyle);
64 CiteStyle const citeStylesUCase[] = {
65 CITET, CITEP, CITEALT, CITEALP, CITEAUTHOR };
67 unsigned int const nCiteStylesUCase =
68 sizeof(citeStylesUCase) / sizeof(CiteStyle);
71 // The functions doing the dirty work for the search.
72 vector<string>::const_iterator
73 simpleSearch(InfoMap const & theMap,
74 vector<string> const & keys,
76 vector<string>::const_iterator start,
84 vector<string> searchwords = getVectorFromString(tmp, " ");
86 // Loop over all keys from start...
87 for (vector<string>::const_iterator it = start;
88 // End condition is direction-dependent.
89 (dir == FORWARD) ? (it<keys.end()) : (it>=keys.begin());
90 // increment is direction-dependent.
91 (dir == FORWARD) ? (++it) : (--it)) {
94 InfoMap::const_iterator info = theMap.find(*it);
95 if (info != theMap.end())
96 data += " " + info->second;
98 data = lowercase(data);
102 // Loop over all search words...
103 for (vector<string>::const_iterator sit = searchwords.begin();
104 sit != searchwords.end(); ++sit) {
105 if (data.find(*sit) == string::npos) {
111 if (found) return it;
118 vector<string>::const_iterator
119 regexSearch(InfoMap const & theMap,
120 vector<string> const & keys,
122 vector<string>::const_iterator start,
125 boost::regex reg(expr);
127 for (vector<string>::const_iterator it = start;
128 // End condition is direction-dependent.
129 (dir == FORWARD) ? (it < keys.end()) : (it >= keys.begin());
130 // increment is direction-dependent.
131 (dir == FORWARD) ? (++it) : (--it)) {
134 InfoMap::const_iterator info = theMap.find(*it);
135 if (info != theMap.end())
136 data += " " + info->second;
138 if (boost::regex_match(data, reg)) {
150 string const familyName(string const & name)
152 // Very simple parser
155 // possible authorname combinations are:
156 // "Surname, FirstName"
158 // "FirstName Surname"
160 string::size_type idx = fname.find(",");
161 if (idx != string::npos)
162 return frontStrip(fname.substr(0,idx));
163 idx = fname.rfind(".");
164 if (idx != string::npos)
165 fname = frontStrip(fname.substr(idx+1));
166 // test if we have a LaTeX Space in front
167 if (fname[0] == '\\')
168 return fname.substr(2);
174 string const getAbbreviatedAuthor(InfoMap const & map, string const & key)
176 lyx::Assert(!map.empty());
178 InfoMap::const_iterator it = map.find(key);
182 string::size_type const pos = it->second.find("TheBibliographyRef");
183 if (pos != string::npos) {
189 strip(frontStrip(it->second.substr(0, pos-1)));
194 split(opt, authors, '(');
198 string author = parseBibTeX(it->second, "author");
200 author = parseBibTeX(it->second, "editor");
202 if (author.empty()) {
203 author = parseBibTeX(it->second, "key");
209 vector<string> authors = getVectorFromString(author, " and");
211 if (!authors.empty()) {
214 for (vector<string>::iterator it = authors.begin();
215 it != authors.end(); ++it) {
216 *it = familyName(strip(*it));
220 if (authors.size() == 2)
221 author += _(" and ") + authors[1];
222 else if (authors.size() > 2)
223 author += _(" et al.");
230 string const getYear(InfoMap const & map, string const & key)
232 lyx::Assert(!map.empty());
234 InfoMap::const_iterator it = map.find(key);
238 string::size_type const pos = it->second.find("TheBibliographyRef");
239 if (pos != string::npos) {
245 strip(frontStrip(it->second.substr(0, pos-1)));
250 string const tmp = split(opt, authors, '(');
252 split(tmp, year, ')');
257 string year = parseBibTeX(it->second, "year");
265 // A functor for use with std::sort, leading to case insensitive sorting
266 struct compareNoCase: public std::binary_function<string, string, bool>
268 bool operator()(string const & s1, string const & s2) const {
269 return compare_no_case(s1, s2) < 0;
273 vector<string> const getKeys(InfoMap const & map)
275 vector<string> bibkeys;
277 for (InfoMap::const_iterator it = map.begin(); it != map.end(); ++it) {
278 bibkeys.push_back(it->first);
281 sort(bibkeys.begin(), bibkeys.end(), compareNoCase());
286 string const getInfo(InfoMap const & map, string const & key)
288 lyx::Assert(!map.empty());
290 InfoMap::const_iterator it = map.find(key);
294 // is the entry a BibTeX one or one from lyx-layout "bibliography"?
295 string const separator("TheBibliographyRef");
296 string::size_type const pos = it->second.find(separator);
297 if (pos != string::npos) {
298 string::size_type const pos2 = pos + separator.size();
299 string const info = strip(frontStrip(it->second.substr(pos2)));
303 // Search for all possible "required" keys
304 string author = parseBibTeX(it->second, "author");
306 author = parseBibTeX(it->second, "editor");
308 string year = parseBibTeX(it->second, "year");
309 string title = parseBibTeX(it->second, "title");
310 string booktitle = parseBibTeX(it->second, "booktitle");
311 string chapter = parseBibTeX(it->second, "chapter");
312 string number = parseBibTeX(it->second, "number");
313 string volume = parseBibTeX(it->second, "volume");
314 string pages = parseBibTeX(it->second, "pages");
316 string media = parseBibTeX(it->second, "journal");
318 media = parseBibTeX(it->second, "publisher");
320 media = parseBibTeX(it->second, "school");
322 media = parseBibTeX(it->second, "institution");
324 ostringstream result;
326 result << author << ", ";
329 if (!booktitle.empty())
330 result << ", in " << booktitle;
331 if (!chapter.empty())
332 result << ", Ch. " << chapter;
334 result << ", " << media;
336 result << ", vol. " << volume;
338 result << ", no. " << number;
340 result << ", pp. " << pages;
342 result << ", " << year;
344 string const result_str = strip(result.str().c_str());
345 if (!result_str.empty())
348 // This should never happen (or at least be very unusual!)
353 vector<string>::const_iterator
354 searchKeys(InfoMap const & theMap,
355 vector<string> const & keys,
357 vector<string>::const_iterator start,
362 // Preliminary checks
363 if (start < keys.begin() || start >= keys.end())
366 string search_expr = frontStrip(strip(expr));
367 if (search_expr.empty())
371 return simpleSearch(theMap, keys, search_expr, start, dir,
374 return regexSearch(theMap, keys, search_expr, start, dir);
378 string const parseBibTeX(string data, string const & findkey)
381 // at first we delete all characters right of '%' and
382 // replace tabs through a space and remove leading spaces
383 // we read the data line by line so that the \n are
387 string dummy = token(data,'\n', Entries);
388 while (!dummy.empty()) {
389 dummy = subst(dummy, '\t', ' '); // no tabs
390 dummy = frontStrip(dummy); // no leading spaces
391 // ignore lines with a beginning '%' or ignore all right of %
392 string::size_type const idx =
393 dummy.empty() ? string::npos : dummy.find('%');
394 if (idx != string::npos)
395 dummy.erase(idx, string::npos);
396 // do we have a new token or a new line of
397 // the same one? In the first case we ignore
398 // the \n and in the second we replace it
400 if (!dummy.empty()) {
401 if (!contains(dummy, "="))
402 data_ += (' ' + dummy);
406 dummy = token(data, '\n', ++Entries);
409 // replace double commas with "" for easy scanning
410 data = subst(data_, ",,", "\"\"");
416 // now get only the important line of the bibtex entry.
417 // all entries are devided by ',' except the last one.
418 data += ','; // now we have same behaviour for all entries
419 // because the last one is "blah ... }"
422 // parsing of title and booktitle is different from the
423 // others, because booktitle contains title
425 dummy = token(data, ',', Entries++);
426 if (!dummy.empty()) {
427 found = contains(lowercase(dummy), findkey);
428 if (findkey == "title" &&
429 contains(lowercase(dummy), "booktitle"))
432 } while (!found && !dummy.empty());
437 // we are not sure, if we get all, because "key= "blah, blah" is
439 // Therefore we read all until the next "=" character, which follows a
442 dummy = token(data, ',', Entries++);
443 while (!contains(dummy, '=') && !dummy.empty()) {
444 keyvalue += (',' + dummy);
445 dummy = token(data, ',', Entries++);
448 // replace double "" with originals ,, (two commas)
449 // leaving us with the all-important line
450 data = subst(keyvalue, "\"\"", ",,");
454 data = strip(data, ' ');
455 // 2. if there is no opening '{' then a closing '{' is probably cruft.
456 if (!contains(data, '{'))
457 data = strip(data, '}');
458 // happens, when last keyword
459 string::size_type const idx =
460 !data.empty() ? data.find('=') : string::npos;
462 if (idx == string::npos)
465 data = data.substr(idx);
466 data = frontStrip(strip(data));
468 if (data.length() < 2 || data[0] != '=') { // a valid entry?
471 // delete '=' and the following spaces
472 data = frontStrip(frontStrip(data,'='));
473 if (data.length() < 2) {
474 return data; // not long enough to find delimiters
476 string::size_type keypos = 1;
478 if (data[0] == '{') {
480 } else if (data[0] == '"') {
483 // no {} and no "", pure data but with a
484 // possible ',' at the end
485 return strip(data,',');
487 string tmp = data.substr(keypos);
488 while (tmp.find('{') != string::npos &&
489 tmp.find('}') != string::npos &&
490 tmp.find('{') < tmp.find('}') &&
491 tmp.find('{') < tmp.find(enclosing)) {
493 keypos += tmp.find('{') + 1;
494 tmp = data.substr(keypos);
495 keypos += tmp.find('}') + 1;
496 tmp = data.substr(keypos);
498 if (tmp.find(enclosing) == string::npos)
501 keypos += tmp.find(enclosing);
502 return data.substr(1, keypos - 1);
509 CitationStyle const getCitationStyle(string const & command)
511 if (command.empty()) return CitationStyle();
514 string cmd = command;
517 cs.forceUCase = true;
521 size_t n = cmd.size()-1;
524 cmd = cmd.substr(0,n);
527 char const * const * const last = citeCommands + nCiteCommands;
528 char const * const * const ptr = std::find(citeCommands, last, cmd);
531 size_t idx = ptr - citeCommands;
532 cs.style = citeStyles[idx];
539 string const getCiteCommand(CiteStyle command, bool full, bool forceUCase)
541 string cite = citeCommands[command];
543 CiteStyle const * last = citeStylesFull + nCiteStylesFull;
544 if (std::find(citeStylesFull, last, command) != last)
549 CiteStyle const * last = citeStylesUCase + nCiteStylesUCase;
550 if (std::find(citeStylesUCase, last, command) != last)
558 vector<CiteStyle> const getCiteStyles(bool usingNatbib)
560 unsigned int nStyles = 1;
561 unsigned int start = 0;
563 nStyles = nCiteStyles - 1;
567 vector<CiteStyle> styles(nStyles);
569 vector<CiteStyle>::size_type i = 0;
571 for (; i != styles.size(); ++i, ++j) {
572 styles[i] = citeStyles[j];
580 getNumericalStrings(string const & key,
581 InfoMap const & map, vector<CiteStyle> const & styles)
584 return vector<string>();
587 string const author = getAbbreviatedAuthor(map, key);
588 string const year = getYear(map, key);
589 if (author.empty() || year.empty())
590 return vector<string>();
592 vector<string> vec(styles.size());
593 for (vector<string>::size_type i = 0; i != vec.size(); ++i) {
603 str = author + " [#ID]";
607 str = author + " #ID";
623 str = "(" + year + ")";
635 getAuthorYearStrings(string const & key,
636 InfoMap const & map, vector<CiteStyle> const & styles)
639 return vector<string>();
642 string const author = getAbbreviatedAuthor(map, key);
643 string const year = getYear(map, key);
644 if (author.empty() || year.empty())
645 return vector<string>();
647 vector<string> vec(styles.size());
648 for (vector<string>::size_type i = 0; i != vec.size(); ++i) {
654 str = author + " (" + year + ")";
658 str = "(" + author + ", " + year + ")";
662 str = author + " " + year ;
666 str = author + ", " + year ;
678 str = "(" + year + ")";
688 } // namespace biblio