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
9 * Full author contact details are available in file CREDITS.
14 #include "support/std_string.h"
16 #include "support/std_sstream.h"
18 #include "BoostFormat.h"
22 #include <boost/tokenizer.hpp>
23 #include <boost/assert.hpp>
34 #ifndef CXX_GLOBAL_CSTD
45 int compare_no_case(string const & s, string const & s2)
47 string::const_iterator p = s.begin();
48 string::const_iterator p2 = s2.begin();
50 while (p != s.end() && p2 != s2.end()) {
51 int const lc1 = tolower(*p);
52 int const lc2 = tolower(*p2);
54 return (lc1 < lc2) ? -1 : 1;
59 if (s.size() == s2.size())
61 if (s.size() < s2.size())
68 int ascii_tolower(int c) {
69 if (c >= 'A' && c <= 'Z')
76 int compare_ascii_no_case(string const & s, string const & s2)
78 string::const_iterator p = s.begin();
79 string::const_iterator p2 = s2.begin();
81 while (p != s.end() && p2 != s2.end()) {
82 int const lc1 = ascii_tolower(*p);
83 int const lc2 = ascii_tolower(*p2);
85 return (lc1 < lc2) ? -1 : 1;
90 if (s.size() == s2.size())
92 if (s.size() < s2.size())
98 int compare_no_case(string const & s, string const & s2, unsigned int len)
100 string::const_iterator p = s.begin();
101 string::const_iterator p2 = s2.begin();
103 while (i < len && p != s.end() && p2 != s2.end()) {
104 int const lc1 = tolower(*p);
105 int const lc2 = tolower(*p2);
107 return (lc1 < lc2) ? -1 : 1;
113 if (s.size() >= len && s2.size() >= len)
115 if (s.size() < s2.size())
121 bool isStrInt(string const & str)
123 if (str.empty()) return false;
125 // Remove leading and trailing white space chars.
126 string const tmpstr = trim(str);
127 if (tmpstr.empty()) return false;
129 string::const_iterator cit = tmpstr.begin();
130 if ((*cit) == '-') ++cit;
131 string::const_iterator end = tmpstr.end();
132 for (; cit != end; ++cit) {
133 if (!isdigit((*cit))) return false;
139 bool isStrUnsignedInt(string const & str)
141 if (str.empty()) return false;
143 // Remove leading and trailing white space chars.
144 string const tmpstr = trim(str);
145 if (tmpstr.empty()) return false;
147 string::const_iterator cit = tmpstr.begin();
148 string::const_iterator end = tmpstr.end();
149 for (; cit != end; ++cit) {
150 if (!isdigit((*cit))) return false;
156 int strToInt(string const & str)
159 // Remove leading and trailing white space chars.
160 string const tmpstr = trim(str);
161 // Do the conversion proper.
169 unsigned int strToUnsignedInt(string const & str)
171 if (isStrUnsignedInt(str)) {
172 // Remove leading and trailing white space chars.
173 string const tmpstr = trim(str);
174 // Do the conversion proper.
182 bool isStrDbl(string const & str)
184 if (str.empty()) return false;
186 // Remove leading and trailing white space chars.
187 string const tmpstr = trim(str);
188 if (tmpstr.empty()) return false;
189 // if (1 < tmpstr.count('.')) return false;
191 string::const_iterator cit = tmpstr.begin();
192 bool found_dot(false);
193 if ((*cit) == '-') ++cit;
194 string::const_iterator end = tmpstr.end();
195 for (; cit != end; ++cit) {
212 double strToDbl(string const & str)
215 // Remove leading and trailing white space chars.
216 string const tmpstr = trim(str);
217 // Do the conversion proper.
218 return ::atof(tmpstr.c_str());
225 char lowercase(char c)
227 return char(tolower(c));
231 char uppercase(char c)
233 return char(toupper(c));
239 // since we cannot use std::tolower and std::toupper directly in the
240 // calls to std::transform yet, we use these helper clases. (Lgb)
242 struct local_lowercase {
243 char operator()(char c) const {
248 struct local_uppercase {
249 char operator()(char c) const {
254 struct local_ascii_lowercase {
255 char operator()(char c) const {
256 return ascii_tolower(c);
260 } // end of anon namespace
262 string const lowercase(string const & a)
265 transform(tmp.begin(), tmp.end(), tmp.begin(), local_lowercase());
269 string const uppercase(string const & a)
272 transform(tmp.begin(), tmp.end(), tmp.begin(), local_uppercase());
277 string const ascii_lowercase(string const & a)
280 transform(tmp.begin(), tmp.end(), tmp.begin(),
281 local_ascii_lowercase());
286 bool prefixIs(string const & a, string const & pre)
288 string::size_type const prelen = pre.length();
289 string::size_type const alen = a.length();
291 if (prelen > alen || a.empty())
294 #if !defined(USE_INCLUDED_STRING) && !defined(STD_STRING_IS_GOOD)
295 return ::strncmp(a.c_str(), pre.c_str(), prelen) == 0;
297 return a.compare(0, prelen, pre) == 0;
303 bool suffixIs(string const & a, char c)
305 if (a.empty()) return false;
306 return a[a.length() - 1] == c;
310 bool suffixIs(string const & a, string const & suf)
312 string::size_type const suflen = suf.length();
313 string::size_type const alen = a.length();
318 #if !defined(USE_INCLUDED_STRING) && !defined(STD_STRING_IS_GOOD)
319 string tmp(a, alen - suflen);
320 return ::strncmp(tmp.c_str(), suf.c_str(), suflen) == 0;
322 return a.compare(alen - suflen, suflen, suf) == 0;
328 bool contains(string const & a, string const & b)
332 return a.find(b) != string::npos;
336 bool contains(string const & a, char b)
340 return a.find(b) != string::npos;
344 bool containsOnly(string const & s, string const & cset)
346 return s.find_first_not_of(cset) == string::npos;
350 // ale970405+lasgoutt-970425
351 // rewritten to use new string (Lgb)
352 string const token(string const & a, char delim, int n)
354 if (a.empty()) return string();
356 string::size_type k = 0;
357 string::size_type i = 0;
359 // Find delimiter or end of string
361 if ((i = a.find(delim, i)) == string::npos)
365 // i is now the n'th delim (or string::npos)
366 if (i == string::npos) return string();
367 k = a.find(delim, i);
368 // k is now the n'th + 1 delim (or string::npos)
370 return a.substr(i, k - i);
374 // this could probably be faster and/or cleaner, but it seems to work (JMarc)
375 // rewritten to use new string (Lgb)
376 int tokenPos(string const & a, char delim, string const & tok)
382 while (!str.empty()) {
383 str = split(str, tmptok, delim);
392 string const subst(string const & a, char oldchar, char newchar)
395 string::iterator lit = tmp.begin();
396 string::iterator end = tmp.end();
397 for (; lit != end; ++lit)
398 if ((*lit) == oldchar)
404 string const subst(string const & a,
405 string const & oldstr, string const & newstr)
408 string::size_type i = 0;
409 string::size_type const olen = oldstr.length();
410 while ((i = lstr.find(oldstr, i)) != string::npos) {
411 lstr.replace(i, olen, newstr);
412 i += newstr.length(); // We need to be sure that we dont
413 // use the same i over and over again.
419 string const trim(string const & a, char const * p)
423 if (a.empty() || !*p)
426 string::size_type r = a.find_last_not_of(p);
427 string::size_type l = a.find_first_not_of(p);
429 // Is this the minimal test? (lgb)
430 if (r == string::npos && l == string::npos)
433 return a.substr(l, r - l + 1);
437 string const rtrim(string const & a, char const * p)
441 if (a.empty() || !*p)
444 string::size_type r = a.find_last_not_of(p);
446 // Is this test really needed? (Lgb)
447 if (r == string::npos)
450 return a.substr(0, r + 1);
454 string const ltrim(string const & a, char const * p)
458 if (a.empty() || !*p)
461 string::size_type l = a.find_first_not_of(p);
463 if (l == string::npos)
466 return a.substr(l, string::npos);
470 string const split(string const & a, string & piece, char delim)
473 string::size_type i = a.find(delim);
474 if (i == a.length() - 1) {
475 piece = a.substr(0, i);
476 } else if (i != string::npos) {
477 piece = a.substr(0, i);
478 tmp = a.substr(i + 1);
481 tmp = a.substr(i + 1);
489 string const split(string const & a, char delim)
492 string::size_type i = a.find(delim);
493 if (i != string::npos) // found delim
494 tmp = a.substr(i + 1);
500 string const rsplit(string const & a, string & piece, char delim)
503 string::size_type i = a.rfind(delim);
504 if (i != string::npos) { // delimiter was found
505 piece = a.substr(0, i);
506 tmp = a.substr(i + 1);
507 } else { // delimiter was not found
514 // This function escapes 8-bit characters and other problematic
515 // characters that cause problems in latex labels.
516 string const escape(string const & lab)
518 char hexdigit[16] = { '0', '1', '2', '3', '4', '5', '6', '7',
519 '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
521 for (string::size_type i = 0; i < lab.length(); ++i) {
522 unsigned char c= lab[i];
523 if (c >= 128 || c == '=' || c == '%') {
525 enc += hexdigit[c>>4];
526 enc += hexdigit[c & 15];
535 /// gives a vector of stringparts which have the delimiter delim
536 vector<string> const getVectorFromString(string const & str,
537 string const & delim)
539 // Lars would like this code to go, but for now his replacement (below)
540 // doesn't fullfil the same function. I have, therefore, reactivated the
541 // old code for now. Angus 11 Nov 2002.
546 string keys(rtrim(str));
548 string::size_type const idx = keys.find(delim);
549 if (idx == string::npos) {
550 vec.push_back(ltrim(keys));
553 string const key = trim(keys.substr(0, idx));
556 string::size_type const start = idx + delim.size();
557 keys = keys.substr(start);
561 boost::char_separator<char> sep(delim.c_str());
562 boost::tokenizer<boost::char_separator<char> > tokens(str, sep);
563 #ifndef USE_INCLUDED_STRING
564 return vector<string>(tokens.begin(), tokens.end());
567 using boost::tokenizer;
568 using boost::char_separator;
570 tokenizer<char_separator<char> >::iterator it = tokens.begin();
571 tokenizer<char_separator<char> >::iterator end = tokens.end();
572 for (; it != end; ++it) {
573 vec.push_back(STRCONV((*it)));
581 // the same vice versa
582 string const getStringFromVector(vector<string> const & vec,
583 string const & delim)
587 for (vector<string>::const_iterator it = vec.begin();
588 it != vec.end(); ++it) {
589 string item = trim(*it);
602 string bformat(string const & fmt, string const & arg1)
604 return (boost::format(fmt) % arg1).str();
608 string bformat(string const & fmt, string const & arg1, string const & arg2)
610 return (boost::format(fmt) % arg1 % arg2).str();
614 string bformat(string const & fmt, int arg1, int arg2)
616 return (boost::format(fmt) % arg1 % arg2).str();
620 string bformat(string const & fmt, string const & arg1, string const & arg2,
623 return (boost::format(fmt) % arg1 % arg2 % arg3).str();
627 string bformat(string const & fmt, string const & arg1, string const & arg2,
628 string const & arg3, string const & arg4)
630 return (boost::format(fmt) % arg1 % arg2 % arg3 % arg4).str();
634 string bformat(string const & fmt, string const & arg1, string const & arg2,
635 string const & arg3, string const & arg4, string const & arg5)
637 return (boost::format(fmt) % arg1 % arg2 % arg3 % arg4 % arg5).str();
642 string bformat(string const & fmt, string const & arg1)
644 BOOST_ASSERT(contains(fmt, "%1$s"));
645 string const str = subst(fmt, "%1$s", arg1);
646 return subst(str, "%%", "%");
650 string bformat(string const & fmt, string const & arg1, string const & arg2)
652 BOOST_ASSERT(contains(fmt, "%1$s"));
653 BOOST_ASSERT(contains(fmt, "%2$s"));
654 string str = subst(fmt, "%1$s", arg1);
655 str = subst(str, "%2$s", arg2);
656 return subst(str, "%%", "%");
660 string bformat(string const & fmt, int arg1, int arg2)
662 BOOST_ASSERT(contains(fmt, "%1$d"));
663 BOOST_ASSERT(contains(fmt, "%2$d"));
664 string str = subst(fmt, "%1$d", tostr(arg1));
665 str = subst(str, "%2$d", tostr(arg2));
666 return subst(str, "%%", "%");
670 string bformat(string const & fmt, string const & arg1, string const & arg2,
673 BOOST_ASSERT(contains(fmt, "%1$s"));
674 BOOST_ASSERT(contains(fmt, "%2$s"));
675 BOOST_ASSERT(contains(fmt, "%3$s"));
676 string str = subst(fmt, "%1$s", arg1);
677 str = subst(str, "%2$s", arg2);
678 str = subst(str, "%3$s", arg3);
679 return subst(str, "%%", "%");
683 string bformat(string const & fmt, string const & arg1, string const & arg2,
684 string const & arg3, string const & arg4)
686 BOOST_ASSERT(contains(fmt, "%1$s"));
687 BOOST_ASSERT(contains(fmt, "%2$s"));
688 BOOST_ASSERT(contains(fmt, "%3$s"));
689 BOOST_ASSERT(contains(fmt, "%4$s"));
690 string str = subst(fmt, "%1$s", arg1);
691 str = subst(str, "%2$s", arg2);
692 str = subst(str, "%3$s", arg3);
693 str = subst(str, "%4$s", arg4);
694 return subst(str, "%%", "%");
698 string bformat(string const & fmt, string const & arg1, string const & arg2,
699 string const & arg3, string const & arg4, string const & arg5)
701 BOOST_ASSERT(contains(fmt, "%1$s"));
702 BOOST_ASSERT(contains(fmt, "%2$s"));
703 BOOST_ASSERT(contains(fmt, "%3$s"));
704 BOOST_ASSERT(contains(fmt, "%4$s"));
705 BOOST_ASSERT(contains(fmt, "%5$s"));
706 string str = subst(fmt, "%1$s", arg1);
707 str = subst(str, "%2$s", arg2);
708 str = subst(str, "%3$s", arg3);
709 str = subst(str, "%4$s", arg4);
710 str = subst(str, "%5$s", arg5);
711 return subst(str, "%%", "%");
716 } // namespace support