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.
15 #include "support/std_sstream.h"
17 #include "BoostFormat.h"
21 #include <boost/tokenizer.hpp>
22 #include <boost/assert.hpp>
33 #ifndef CXX_GLOBAL_CSTD
44 int compare_no_case(string const & s, string const & s2)
46 string::const_iterator p = s.begin();
47 string::const_iterator p2 = s2.begin();
49 while (p != s.end() && p2 != s2.end()) {
50 int const lc1 = tolower(*p);
51 int const lc2 = tolower(*p2);
53 return (lc1 < lc2) ? -1 : 1;
58 if (s.size() == s2.size())
60 if (s.size() < s2.size())
67 int ascii_tolower(int c) {
68 if (c >= 'A' && c <= 'Z')
75 int compare_ascii_no_case(string const & s, string const & s2)
77 string::const_iterator p = s.begin();
78 string::const_iterator p2 = s2.begin();
80 while (p != s.end() && p2 != s2.end()) {
81 int const lc1 = ascii_tolower(*p);
82 int const lc2 = ascii_tolower(*p2);
84 return (lc1 < lc2) ? -1 : 1;
89 if (s.size() == s2.size())
91 if (s.size() < s2.size())
97 int compare_no_case(string const & s, string const & s2, unsigned int len)
99 string::const_iterator p = s.begin();
100 string::const_iterator p2 = s2.begin();
102 while (i < len && p != s.end() && p2 != s2.end()) {
103 int const lc1 = tolower(*p);
104 int const lc2 = tolower(*p2);
106 return (lc1 < lc2) ? -1 : 1;
112 if (s.size() >= len && s2.size() >= len)
114 if (s.size() < s2.size())
120 bool isStrInt(string const & str)
122 if (str.empty()) return false;
124 // Remove leading and trailing white space chars.
125 string const tmpstr = trim(str);
126 if (tmpstr.empty()) return false;
128 string::const_iterator cit = tmpstr.begin();
129 if ((*cit) == '-') ++cit;
130 string::const_iterator end = tmpstr.end();
131 for (; cit != end; ++cit) {
132 if (!isdigit((*cit))) return false;
138 bool isStrUnsignedInt(string const & str)
140 if (str.empty()) return false;
142 // Remove leading and trailing white space chars.
143 string const tmpstr = trim(str);
144 if (tmpstr.empty()) return false;
146 string::const_iterator cit = tmpstr.begin();
147 string::const_iterator end = tmpstr.end();
148 for (; cit != end; ++cit) {
149 if (!isdigit((*cit))) return false;
155 int strToInt(string const & str)
158 // Remove leading and trailing white space chars.
159 string const tmpstr = trim(str);
160 // Do the conversion proper.
168 unsigned int strToUnsignedInt(string const & str)
170 if (isStrUnsignedInt(str)) {
171 // Remove leading and trailing white space chars.
172 string const tmpstr = trim(str);
173 // Do the conversion proper.
181 bool isStrDbl(string const & str)
183 if (str.empty()) return false;
185 // Remove leading and trailing white space chars.
186 string const tmpstr = trim(str);
187 if (tmpstr.empty()) return false;
188 // if (1 < tmpstr.count('.')) return false;
190 string::const_iterator cit = tmpstr.begin();
191 bool found_dot(false);
192 if ((*cit) == '-') ++cit;
193 string::const_iterator end = tmpstr.end();
194 for (; cit != end; ++cit) {
211 double strToDbl(string const & str)
214 // Remove leading and trailing white space chars.
215 string const tmpstr = trim(str);
216 // Do the conversion proper.
217 return ::atof(tmpstr.c_str());
224 char lowercase(char c)
226 return char(tolower(c));
230 char uppercase(char c)
232 return char(toupper(c));
238 // since we cannot use std::tolower and std::toupper directly in the
239 // calls to std::transform yet, we use these helper clases. (Lgb)
241 struct local_lowercase {
242 char operator()(char c) const {
247 struct local_uppercase {
248 char operator()(char c) const {
253 struct local_ascii_lowercase {
254 char operator()(char c) const {
255 return ascii_tolower(c);
259 } // end of anon namespace
261 string const lowercase(string const & a)
264 transform(tmp.begin(), tmp.end(), tmp.begin(), local_lowercase());
268 string const uppercase(string const & a)
271 transform(tmp.begin(), tmp.end(), tmp.begin(), local_uppercase());
276 string const ascii_lowercase(string const & a)
279 transform(tmp.begin(), tmp.end(), tmp.begin(),
280 local_ascii_lowercase());
285 bool prefixIs(string const & a, string const & pre)
287 string::size_type const prelen = pre.length();
288 string::size_type const alen = a.length();
290 if (prelen > alen || a.empty())
293 #if !defined(USE_INCLUDED_STRING) && !defined(STD_STRING_IS_GOOD)
294 return ::strncmp(a.c_str(), pre.c_str(), prelen) == 0;
296 return a.compare(0, prelen, pre) == 0;
302 bool suffixIs(string const & a, char c)
304 if (a.empty()) return false;
305 return a[a.length() - 1] == c;
309 bool suffixIs(string const & a, string const & suf)
311 string::size_type const suflen = suf.length();
312 string::size_type const alen = a.length();
317 #if !defined(USE_INCLUDED_STRING) && !defined(STD_STRING_IS_GOOD)
318 string tmp(a, alen - suflen);
319 return ::strncmp(tmp.c_str(), suf.c_str(), suflen) == 0;
321 return a.compare(alen - suflen, suflen, suf) == 0;
327 bool contains(string const & a, string const & b)
331 return a.find(b) != string::npos;
335 bool contains(string const & a, char b)
339 return a.find(b) != string::npos;
343 bool containsOnly(string const & s, string const & cset)
345 return s.find_first_not_of(cset) == string::npos;
349 // ale970405+lasgoutt-970425
350 // rewritten to use new string (Lgb)
351 string const token(string const & a, char delim, int n)
353 if (a.empty()) return string();
355 string::size_type k = 0;
356 string::size_type i = 0;
358 // Find delimiter or end of string
360 if ((i = a.find(delim, i)) == string::npos)
364 // i is now the n'th delim (or string::npos)
365 if (i == string::npos) return string();
366 k = a.find(delim, i);
367 // k is now the n'th + 1 delim (or string::npos)
369 return a.substr(i, k - i);
373 // this could probably be faster and/or cleaner, but it seems to work (JMarc)
374 // rewritten to use new string (Lgb)
375 int tokenPos(string const & a, char delim, string const & tok)
381 while (!str.empty()) {
382 str = split(str, tmptok, delim);
391 string const subst(string const & a, char oldchar, char newchar)
394 string::iterator lit = tmp.begin();
395 string::iterator end = tmp.end();
396 for (; lit != end; ++lit)
397 if ((*lit) == oldchar)
403 string const subst(string const & a,
404 string const & oldstr, string const & newstr)
407 string::size_type i = 0;
408 string::size_type const olen = oldstr.length();
409 while ((i = lstr.find(oldstr, i)) != string::npos) {
410 lstr.replace(i, olen, newstr);
411 i += newstr.length(); // We need to be sure that we dont
412 // use the same i over and over again.
418 string const trim(string const & a, char const * p)
422 if (a.empty() || !*p)
425 string::size_type r = a.find_last_not_of(p);
426 string::size_type l = a.find_first_not_of(p);
428 // Is this the minimal test? (lgb)
429 if (r == string::npos && l == string::npos)
432 return a.substr(l, r - l + 1);
436 string const rtrim(string const & a, char const * p)
440 if (a.empty() || !*p)
443 string::size_type r = a.find_last_not_of(p);
445 // Is this test really needed? (Lgb)
446 if (r == string::npos)
449 return a.substr(0, r + 1);
453 string const ltrim(string const & a, char const * p)
457 if (a.empty() || !*p)
460 string::size_type l = a.find_first_not_of(p);
462 if (l == string::npos)
465 return a.substr(l, string::npos);
469 string const split(string const & a, string & piece, char delim)
472 string::size_type i = a.find(delim);
473 if (i == a.length() - 1) {
474 piece = a.substr(0, i);
475 } else if (i != string::npos) {
476 piece = a.substr(0, i);
477 tmp = a.substr(i + 1);
480 tmp = a.substr(i + 1);
488 string const split(string const & a, char delim)
491 string::size_type i = a.find(delim);
492 if (i != string::npos) // found delim
493 tmp = a.substr(i + 1);
499 string const rsplit(string const & a, string & piece, char delim)
502 string::size_type i = a.rfind(delim);
503 if (i != string::npos) { // delimiter was found
504 piece = a.substr(0, i);
505 tmp = a.substr(i + 1);
506 } else { // delimiter was not found
513 // This function escapes 8-bit characters and other problematic
514 // characters that cause problems in latex labels.
515 string const escape(string const & lab)
517 char hexdigit[16] = { '0', '1', '2', '3', '4', '5', '6', '7',
518 '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
520 for (string::size_type i = 0; i < lab.length(); ++i) {
521 unsigned char c= lab[i];
522 if (c >= 128 || c == '=' || c == '%') {
524 enc += hexdigit[c>>4];
525 enc += hexdigit[c & 15];
534 /// gives a vector of stringparts which have the delimiter delim
535 vector<string> const getVectorFromString(string const & str,
536 string const & delim)
538 // Lars would like this code to go, but for now his replacement (below)
539 // doesn't fullfil the same function. I have, therefore, reactivated the
540 // old code for now. Angus 11 Nov 2002.
545 string keys(rtrim(str));
547 string::size_type const idx = keys.find(delim);
548 if (idx == string::npos) {
549 vec.push_back(ltrim(keys));
552 string const key = trim(keys.substr(0, idx));
555 string::size_type const start = idx + delim.size();
556 keys = keys.substr(start);
560 boost::char_separator<char> sep(delim.c_str());
561 boost::tokenizer<boost::char_separator<char> > tokens(str, sep);
562 #ifndef USE_INCLUDED_STRING
563 return vector<string>(tokens.begin(), tokens.end());
566 using boost::tokenizer;
567 using boost::char_separator;
569 tokenizer<char_separator<char> >::iterator it = tokens.begin();
570 tokenizer<char_separator<char> >::iterator end = tokens.end();
571 for (; it != end; ++it) {
572 vec.push_back(STRCONV((*it)));
580 // the same vice versa
581 string const getStringFromVector(vector<string> const & vec,
582 string const & delim)
586 for (vector<string>::const_iterator it = vec.begin();
587 it != vec.end(); ++it) {
588 string item = trim(*it);
601 string bformat(string const & fmt, string const & arg1)
603 return (boost::format(fmt) % arg1).str();
607 string bformat(string const & fmt, string const & arg1, string const & arg2)
609 return (boost::format(fmt) % arg1 % arg2).str();
613 string bformat(string const & fmt, int arg1, int arg2)
615 return (boost::format(fmt) % arg1 % arg2).str();
619 string bformat(string const & fmt, string const & arg1, string const & arg2,
622 return (boost::format(fmt) % arg1 % arg2 % arg3).str();
626 string bformat(string const & fmt, string const & arg1, string const & arg2,
627 string const & arg3, string const & arg4)
629 return (boost::format(fmt) % arg1 % arg2 % arg3 % arg4).str();
633 string bformat(string const & fmt, string const & arg1, string const & arg2,
634 string const & arg3, string const & arg4, string const & arg5)
636 return (boost::format(fmt) % arg1 % arg2 % arg3 % arg4 % arg5).str();
641 string bformat(string const & fmt, string const & arg1)
643 BOOST_ASSERT(contains(fmt, "%1$s"));
644 string const str = subst(fmt, "%1$s", arg1);
645 return subst(str, "%%", "%");
649 string bformat(string const & fmt, string const & arg1, string const & arg2)
651 BOOST_ASSERT(contains(fmt, "%1$s"));
652 BOOST_ASSERT(contains(fmt, "%2$s"));
653 string str = subst(fmt, "%1$s", arg1);
654 str = subst(str, "%2$s", arg2);
655 return subst(str, "%%", "%");
659 string bformat(string const & fmt, int arg1, int arg2)
661 BOOST_ASSERT(contains(fmt, "%1$d"));
662 BOOST_ASSERT(contains(fmt, "%2$d"));
663 string str = subst(fmt, "%1$d", tostr(arg1));
664 str = subst(str, "%2$d", tostr(arg2));
665 return subst(str, "%%", "%");
669 string bformat(string const & fmt, string const & arg1, string const & arg2,
672 BOOST_ASSERT(contains(fmt, "%1$s"));
673 BOOST_ASSERT(contains(fmt, "%2$s"));
674 BOOST_ASSERT(contains(fmt, "%3$s"));
675 string str = subst(fmt, "%1$s", arg1);
676 str = subst(str, "%2$s", arg2);
677 str = subst(str, "%3$s", arg3);
678 return subst(str, "%%", "%");
682 string bformat(string const & fmt, string const & arg1, string const & arg2,
683 string const & arg3, string const & arg4)
685 BOOST_ASSERT(contains(fmt, "%1$s"));
686 BOOST_ASSERT(contains(fmt, "%2$s"));
687 BOOST_ASSERT(contains(fmt, "%3$s"));
688 BOOST_ASSERT(contains(fmt, "%4$s"));
689 string str = subst(fmt, "%1$s", arg1);
690 str = subst(str, "%2$s", arg2);
691 str = subst(str, "%3$s", arg3);
692 str = subst(str, "%4$s", arg4);
693 return subst(str, "%%", "%");
697 string bformat(string const & fmt, string const & arg1, string const & arg2,
698 string const & arg3, string const & arg4, string const & arg5)
700 BOOST_ASSERT(contains(fmt, "%1$s"));
701 BOOST_ASSERT(contains(fmt, "%2$s"));
702 BOOST_ASSERT(contains(fmt, "%3$s"));
703 BOOST_ASSERT(contains(fmt, "%4$s"));
704 BOOST_ASSERT(contains(fmt, "%5$s"));
705 string str = subst(fmt, "%1$s", arg1);
706 str = subst(str, "%2$s", arg2);
707 str = subst(str, "%3$s", arg3);
708 str = subst(str, "%4$s", arg4);
709 str = subst(str, "%5$s", arg5);
710 return subst(str, "%%", "%");
715 } // namespace support