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/lstrings.h"
15 #include "support/lyxlib.h"
16 #include "support/convert.h"
17 #include "support/qstring_helpers.h"
21 #include <boost/tokenizer.hpp>
22 #include <boost/assert.hpp>
24 #ifndef I_AM_NOT_AFRAID_OF_HEADER_LIBRARIES
26 #include <boost/format.hpp>
41 #ifndef CXX_GLOBAL_CSTD
51 int compare_no_case(docstring const & s, docstring const & s2)
53 docstring::const_iterator p = s.begin();
54 docstring::const_iterator p2 = s2.begin();
56 while (p != s.end() && p2 != s2.end()) {
57 char_type const lc1 = lowercase(*p);
58 char_type const lc2 = lowercase(*p2);
60 return (lc1 < lc2) ? -1 : 1;
65 if (s.size() == s2.size())
67 if (s.size() < s2.size())
75 template<typename Char>
76 Char ascii_tolower(Char c) {
77 if (c >= 'A' && c <= 'Z')
85 int compare_ascii_no_case(string const & s, string const & s2)
87 string::const_iterator p = s.begin();
88 string::const_iterator p2 = s2.begin();
90 while (p != s.end() && p2 != s2.end()) {
91 int const lc1 = ascii_tolower(*p);
92 int const lc2 = ascii_tolower(*p2);
94 return (lc1 < lc2) ? -1 : 1;
99 if (s.size() == s2.size())
101 if (s.size() < s2.size())
107 int compare_ascii_no_case(docstring const & s, docstring const & s2)
109 docstring::const_iterator p = s.begin();
110 docstring::const_iterator p2 = s2.begin();
112 while (p != s.end() && p2 != s2.end()) {
113 char_type const lc1 = ascii_tolower(*p);
114 char_type const lc2 = ascii_tolower(*p2);
116 return (lc1 < lc2) ? -1 : 1;
121 if (s.size() == s2.size())
123 if (s.size() < s2.size())
129 bool isStrInt(string const & str)
131 if (str.empty()) return false;
133 // Remove leading and trailing white space chars.
134 string const tmpstr = trim(str);
135 if (tmpstr.empty()) return false;
137 string::const_iterator cit = tmpstr.begin();
138 if ((*cit) == '-') ++cit;
139 string::const_iterator end = tmpstr.end();
140 for (; cit != end; ++cit) {
141 if (!isdigit((*cit))) return false;
147 bool isStrUnsignedInt(string const & str)
149 if (str.empty()) return false;
151 // Remove leading and trailing white space chars.
152 string const tmpstr = trim(str);
153 if (tmpstr.empty()) return false;
155 string::const_iterator cit = tmpstr.begin();
156 string::const_iterator end = tmpstr.end();
157 for (; cit != end; ++cit) {
158 if (!isdigit((*cit))) return false;
164 bool isStrDbl(string const & str)
166 if (str.empty()) return false;
168 // Remove leading and trailing white space chars.
169 string const tmpstr = trim(str);
170 if (tmpstr.empty()) return false;
171 // if (1 < tmpstr.count('.')) return false;
173 string::const_iterator cit = tmpstr.begin();
174 bool found_dot(false);
175 if ((*cit) == '-') ++cit;
176 string::const_iterator end = tmpstr.end();
177 for (; cit != end; ++cit) {
197 bool isHexChar(char_type c)
209 c == 'a' || c == 'A' ||
210 c == 'b' || c == 'B' ||
211 c == 'c' || c == 'C' ||
212 c == 'd' || c == 'D' ||
213 c == 'e' || c == 'E' ||
214 c == 'f' || c == 'F';
220 bool isHex(docstring const & str)
224 if (str.length() > 2 && str[0] == '0' &&
225 (str[1] == 'x' || str[1] == 'X'))
228 int const len = str.length();
230 for (; index < len; ++index) {
231 if (!isHexChar(str[index]))
238 int hexToInt(docstring const & str)
240 string s = to_ascii(str);
242 sscanf(s.c_str(), "%x", &h);
247 bool isAscii(docstring const & str)
249 int const len = str.length();
250 for (int i = 0; i < len; ++i)
257 bool isAscii(string const & str)
259 int const len = str.length();
260 for (int i = 0; i < len; ++i)
261 if (static_cast<unsigned char>(str[i]) >= 0x80)
267 char lowercase(char c)
269 BOOST_ASSERT(static_cast<unsigned char>(c) < 0x80);
270 return char(tolower(c));
274 char uppercase(char c)
276 BOOST_ASSERT(static_cast<unsigned char>(c) < 0x80);
277 return char(toupper(c));
281 char_type lowercase(char_type c)
284 // We don't know how to lowercase a non-utf16 char
286 return qchar_to_ucs4(ucs4_to_qchar(c).toLower());
290 char_type uppercase(char_type c)
293 // We don't know how to uppercase a non-utf16 char
295 return qchar_to_ucs4(ucs4_to_qchar(c).toUpper());
301 // since we cannot use std::tolower and std::toupper directly in the
302 // calls to std::transform yet, we use these helper clases. (Lgb)
304 struct local_lowercase {
305 char_type operator()(char_type c) const {
307 // We don't know how to lowercase a non-utf16 char
309 return qchar_to_ucs4(ucs4_to_qchar(c).toLower());
313 struct local_uppercase {
314 char_type operator()(char_type c) const {
316 // We don't know how to uppercase a non-utf16 char
318 return qchar_to_ucs4(ucs4_to_qchar(c).toUpper());
322 template<typename Char> struct local_ascii_lowercase {
323 Char operator()(Char c) const {
324 return ascii_tolower(c);
328 } // end of anon namespace
330 docstring const lowercase(docstring const & a)
333 transform(tmp.begin(), tmp.end(), tmp.begin(), local_lowercase());
338 docstring const uppercase(docstring const & a)
341 transform(tmp.begin(), tmp.end(), tmp.begin(), local_uppercase());
346 string const ascii_lowercase(string const & a)
349 transform(tmp.begin(), tmp.end(), tmp.begin(),
350 local_ascii_lowercase<char>());
355 docstring const ascii_lowercase(docstring const & a)
358 transform(tmp.begin(), tmp.end(), tmp.begin(),
359 local_ascii_lowercase<char_type>());
364 bool prefixIs(docstring const & a, char_type c)
372 bool prefixIs(string const & a, string const & pre)
374 string::size_type const prelen = pre.length();
375 string::size_type const alen = a.length();
377 if (prelen > alen || a.empty())
380 #if defined(STD_STRING_IS_GOOD)
381 return a.compare(0, prelen, pre) == 0;
383 return ::strncmp(a.c_str(), pre.c_str(), prelen) == 0;
389 bool prefixIs(docstring const & a, docstring const & pre)
391 docstring::size_type const prelen = pre.length();
392 docstring::size_type const alen = a.length();
394 if (prelen > alen || a.empty())
397 return a.compare(0, prelen, pre) == 0;
401 bool suffixIs(string const & a, char c)
403 if (a.empty()) return false;
404 return a[a.length() - 1] == c;
408 bool suffixIs(docstring const & a, char_type c)
412 return a[a.length() - 1] == c;
416 bool suffixIs(string const & a, string const & suf)
418 string::size_type const suflen = suf.length();
419 string::size_type const alen = a.length();
424 #if !defined(USE_INCLUDED_STRING) && !defined(STD_STRING_IS_GOOD)
425 string tmp(a, alen - suflen);
426 return ::strncmp(tmp.c_str(), suf.c_str(), suflen) == 0;
428 return a.compare(alen - suflen, suflen, suf) == 0;
434 bool containsOnly(string const & s, string const & cset)
436 return s.find_first_not_of(cset) == string::npos;
440 // ale970405+lasgoutt-970425
441 // rewritten to use new string (Lgb)
442 string const token(string const & a, char delim, int n)
444 if (a.empty()) return string();
446 string::size_type k = 0;
447 string::size_type i = 0;
449 // Find delimiter or end of string
451 if ((i = a.find(delim, i)) == string::npos)
455 // i is now the n'th delim (or string::npos)
456 if (i == string::npos) return string();
457 k = a.find(delim, i);
458 // k is now the n'th + 1 delim (or string::npos)
460 return a.substr(i, k - i);
464 docstring const token(docstring const & a, char_type delim, int n)
466 if (a.empty()) return docstring();
468 string::size_type k = 0;
469 string::size_type i = 0;
471 // Find delimiter or end of string
473 if ((i = a.find(delim, i)) == docstring::npos)
477 // i is now the n'th delim (or string::npos)
478 if (i == docstring::npos) return docstring();
479 k = a.find(delim, i);
480 // k is now the n'th + 1 delim (or string::npos)
482 return a.substr(i, k - i);
486 // this could probably be faster and/or cleaner, but it seems to work (JMarc)
487 // rewritten to use new string (Lgb)
488 int tokenPos(string const & a, char delim, string const & tok)
494 while (!str.empty()) {
495 str = split(str, tmptok, delim);
506 /// Substitute all \a oldchar with \a newchar
507 template<typename Ch> inline
508 std::basic_string<Ch> const subst_char(std::basic_string<Ch> const & a,
509 Ch oldchar, Ch newchar)
511 typedef std::basic_string<Ch> String;
513 typename String::iterator lit = tmp.begin();
514 typename String::iterator end = tmp.end();
515 for (; lit != end; ++lit)
516 if ((*lit) == oldchar)
521 /// Substitute all \a oldchar with \a newchar
522 docstring const subst_char(docstring const & a,
523 docstring::value_type oldchar, docstring::value_type newchar)
526 docstring::iterator lit = tmp.begin();
527 docstring::iterator end = tmp.end();
528 for (; lit != end; ++lit)
529 if ((*lit) == oldchar)
535 /// substitutes all instances of \a oldstr with \a newstr
536 template<typename String> inline
537 String const subst_string(String const & a,
538 String const & oldstr, String const & newstr)
540 BOOST_ASSERT(!oldstr.empty());
542 typename String::size_type i = 0;
543 typename String::size_type const olen = oldstr.length();
544 while ((i = lstr.find(oldstr, i)) != string::npos) {
545 lstr.replace(i, olen, newstr);
546 i += newstr.length(); // We need to be sure that we dont
547 // use the same i over and over again.
552 docstring const subst_string(docstring const & a,
553 docstring const & oldstr, docstring const & newstr)
555 BOOST_ASSERT(!oldstr.empty());
557 docstring::size_type i = 0;
558 docstring::size_type const olen = oldstr.length();
559 while ((i = lstr.find(oldstr, i)) != string::npos) {
560 lstr.replace(i, olen, newstr);
561 i += newstr.length(); // We need to be sure that we dont
562 // use the same i over and over again.
570 string const subst(string const & a, char oldchar, char newchar)
572 return subst_char(a, oldchar, newchar);
576 docstring const subst(docstring const & a,
577 char_type oldchar, char_type newchar)
579 return subst_char(a, oldchar, newchar);
583 string const subst(string const & a,
584 string const & oldstr, string const & newstr)
586 return subst_string(a, oldstr, newstr);
590 docstring const subst(docstring const & a,
591 docstring const & oldstr, docstring const & newstr)
593 return subst_string(a, oldstr, newstr);
597 docstring const trim(docstring const & a, char const * p)
601 if (a.empty() || !*p)
604 docstring s = lyx::from_ascii(p);
605 docstring::size_type r = a.find_last_not_of(s);
606 docstring::size_type l = a.find_first_not_of(s);
608 // Is this the minimal test? (lgb)
609 if (r == docstring::npos && l == docstring::npos)
612 return a.substr(l, r - l + 1);
616 string const trim(string const & a, char const * p)
620 if (a.empty() || !*p)
623 string::size_type r = a.find_last_not_of(p);
624 string::size_type l = a.find_first_not_of(p);
626 // Is this the minimal test? (lgb)
627 if (r == string::npos && l == string::npos)
630 return a.substr(l, r - l + 1);
634 string const rtrim(string const & a, char const * p)
638 if (a.empty() || !*p)
641 string::size_type r = a.find_last_not_of(p);
643 // Is this test really needed? (Lgb)
644 if (r == string::npos)
647 return a.substr(0, r + 1);
651 docstring const rtrim(docstring const & a, char const * p)
655 if (a.empty() || !*p)
658 docstring::size_type r = a.find_last_not_of(from_ascii(p));
660 // Is this test really needed? (Lgb)
661 if (r == docstring::npos)
664 return a.substr(0, r + 1);
668 string const ltrim(string const & a, char const * p)
671 if (a.empty() || !*p)
673 string::size_type l = a.find_first_not_of(p);
674 if (l == string::npos)
676 return a.substr(l, string::npos);
680 docstring const ltrim(docstring const & a, char const * p)
683 if (a.empty() || !*p)
685 size_t l = a.find_first_not_of(from_ascii(p));
686 if (l == docstring::npos)
688 return a.substr(l, docstring::npos);
693 template<typename String, typename Char> inline
694 String const doSplit(String const & a, String & piece, Char delim)
697 typename String::size_type i = a.find(delim);
698 if (i == a.length() - 1) {
699 piece = a.substr(0, i);
700 } else if (i != String::npos) {
701 piece = a.substr(0, i);
702 tmp = a.substr(i + 1);
705 tmp = a.substr(i + 1);
712 template<typename Char> inline
713 docstring const doSplit(docstring const & a, docstring & piece, Char delim)
716 typename docstring::size_type i = a.find(delim);
717 if (i == a.length() - 1) {
718 piece = a.substr(0, i);
719 } else if (i != docstring::npos) {
720 piece = a.substr(0, i);
721 tmp = a.substr(i + 1);
724 tmp = a.substr(i + 1);
734 string const split(string const & a, string & piece, char delim)
736 return doSplit(a, piece, delim);
740 docstring const split(docstring const & a, docstring & piece, char_type delim)
742 return doSplit(a, piece, delim);
746 string const split(string const & a, char delim)
749 string::size_type i = a.find(delim);
750 if (i != string::npos) // found delim
751 tmp = a.substr(i + 1);
757 string const rsplit(string const & a, string & piece, char delim)
760 string::size_type i = a.rfind(delim);
761 if (i != string::npos) { // delimiter was found
762 piece = a.substr(0, i);
763 tmp = a.substr(i + 1);
764 } else { // delimiter was not found
771 docstring const escape(docstring const & lab)
773 char_type hexdigit[16] = { '0', '1', '2', '3', '4', '5', '6', '7',
774 '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
776 for (docstring::size_type i = 0; i < lab.length(); ++i) {
777 char_type c = lab[i];
778 if (c >= 128 || c == '=' || c == '%') {
779 // Although char_type is a 32 bit type we know that
780 // UCS4 occupies only 21 bits, so we don't need to
781 // encode bigger values. Test for 2^24 because we
782 // can encode that with the 6 hex digits that are
783 // needed for 21 bits anyway.
784 BOOST_ASSERT(c < (1 << 24));
786 enc += hexdigit[(c>>20) & 15];
787 enc += hexdigit[(c>>16) & 15];
788 enc += hexdigit[(c>>12) & 15];
789 enc += hexdigit[(c>> 8) & 15];
790 enc += hexdigit[(c>> 4) & 15];
791 enc += hexdigit[ c & 15];
802 template<typename String> vector<String> const
803 getVectorFromStringT(String const & str, String const & delim)
805 // Lars would like this code to go, but for now his replacement (below)
806 // doesn't fullfil the same function. I have, therefore, reactivated the
807 // old code for now. Angus 11 Nov 2002.
812 String keys = rtrim(str);
814 typename String::size_type const idx = keys.find(delim);
815 if (idx == String::npos) {
816 vec.push_back(ltrim(keys));
819 String const key = trim(keys.substr(0, idx));
822 typename String::size_type const start = idx + delim.size();
823 keys = keys.substr(start);
827 typedef boost::char_separator<typename String::value_type> Separator;
828 typedef boost::tokenizer<Separator, typename String::const_iterator, String> Tokenizer;
829 Separator sep(delim.c_str());
830 Tokenizer tokens(str, sep);
831 return vector<String>(tokens.begin(), tokens.end());
838 vector<string> const getVectorFromString(string const & str,
839 string const & delim)
841 return getVectorFromStringT<string>(str, delim);
845 vector<docstring> const getVectorFromString(docstring const & str,
846 docstring const & delim)
848 return getVectorFromStringT<docstring>(str, delim);
852 // the same vice versa
853 string const getStringFromVector(vector<string> const & vec,
854 string const & delim)
858 for (vector<string>::const_iterator it = vec.begin();
859 it != vec.end(); ++it) {
860 string item = trim(*it);
871 int findToken(char const * const str[], string const & search_token)
875 while (str[i][0] && str[i] != search_token)
883 docstring const externalLineEnding(docstring const & str)
885 #if defined(__APPLE__)
886 // The MAC clipboard uses \r for lineendings, and we use \n
887 return subst(str, '\n', '\r');
888 #elif defined (_WIN32) || (defined (__CYGWIN__) && defined (X_DISPLAY_MISSING))
889 // Windows clipboard uses \r\n for lineendings, and we use \n
890 return subst(str, lyx::from_ascii("\n"), lyx::from_ascii("\r\n"));
897 docstring const internalLineEnding(docstring const & str)
899 docstring const s = subst(str,
900 lyx::from_ascii("\r\n"), lyx::from_ascii("\n"));
901 return subst(s, '\r', '\n');
905 #ifndef I_AM_NOT_AFRAID_OF_HEADER_LIBRARIES
909 docstring bformat(docstring const & fmt, int arg1)
911 return (boost::basic_format<char_type>(fmt) % arg1).str();
916 docstring bformat(docstring const & fmt, long arg1)
918 return (boost::basic_format<char_type>(fmt) % arg1).str();
923 docstring bformat(docstring const & fmt, unsigned int arg1)
925 return (boost::basic_format<char_type>(fmt) % arg1).str();
930 docstring bformat<docstring>(docstring const & fmt, docstring arg1)
932 return (boost::basic_format<char_type>(fmt) % arg1).str();
937 docstring bformat(docstring const & fmt, char * arg1)
939 return (boost::basic_format<char_type>(fmt) % arg1).str();
944 docstring bformat(docstring const & fmt, int arg1, int arg2)
946 return (boost::basic_format<char_type>(fmt) % arg1 % arg2).str();
951 docstring bformat(docstring const & fmt, docstring arg1, docstring arg2)
953 return (boost::basic_format<char_type>(fmt) % arg1 % arg2).str();
958 docstring bformat(docstring const & fmt, char const * arg1, docstring arg2)
960 return (boost::basic_format<char_type>(fmt) % arg1 % arg2).str();
965 docstring bformat(docstring const & fmt, docstring arg1, docstring arg2, docstring arg3)
967 return (boost::basic_format<char_type>(fmt) % arg1 % arg2 % arg3).str();
972 docstring bformat(docstring const & fmt,
973 docstring arg1, docstring arg2, docstring arg3, docstring arg4)
975 return (boost::basic_format<char_type>(fmt) % arg1 % arg2 % arg3 % arg4).str();
981 docstring bformat(docstring const & fmt, int arg1)
983 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$d")));
984 docstring const str = subst(fmt, lyx::from_ascii("%1$d"), convert<docstring>(arg1));
985 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
990 docstring bformat(docstring const & fmt, long arg1)
992 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$d")));
993 docstring const str = subst(fmt, lyx::from_ascii("%1$d"), convert<docstring>(arg1));
994 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
999 docstring bformat(docstring const & fmt, unsigned int arg1)
1001 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$d")));
1002 docstring const str = subst(fmt, lyx::from_ascii("%1$d"), convert<docstring>(arg1));
1003 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1008 docstring bformat(docstring const & fmt, docstring arg1)
1010 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1011 docstring const str = subst(fmt, lyx::from_ascii("%1$s"), arg1);
1012 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1017 docstring bformat(docstring const & fmt, char * arg1)
1019 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1020 docstring const str = subst(fmt, lyx::from_ascii("%1$s"), lyx::from_ascii(arg1));
1021 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1026 docstring bformat(docstring const & fmt, docstring arg1, docstring arg2)
1028 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1029 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%2$s")));
1030 docstring str = subst(fmt, lyx::from_ascii("%1$s"), arg1);
1031 str = subst(str, lyx::from_ascii("%2$s"), arg2);
1032 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1037 docstring bformat(docstring const & fmt, char const * arg1, docstring arg2)
1039 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1040 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%2$s")));
1041 docstring str = subst(fmt, lyx::from_ascii("%1$s"), lyx::from_ascii(arg1));
1042 str = subst(fmt, lyx::from_ascii("%2$s"), arg2);
1043 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1048 docstring bformat(docstring const & fmt, int arg1, int arg2)
1050 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$d")));
1051 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%2$d")));
1052 docstring str = subst(fmt, lyx::from_ascii("%1$d"), convert<docstring>(arg1));
1053 str = subst(str, lyx::from_ascii("%2$d"), convert<docstring>(arg2));
1054 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1059 docstring bformat(docstring const & fmt, docstring arg1, docstring arg2, docstring arg3)
1061 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1062 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%2$s")));
1063 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%3$s")));
1064 docstring str = subst(fmt, lyx::from_ascii("%1$s"), arg1);
1065 str = subst(str, lyx::from_ascii("%2$s"), arg2);
1066 str = subst(str, lyx::from_ascii("%3$s"), arg3);
1067 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1072 docstring bformat(docstring const & fmt,
1073 docstring arg1, docstring arg2, docstring arg3, docstring arg4)
1075 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%1$s")));
1076 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%2$s")));
1077 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%3$s")));
1078 BOOST_ASSERT(contains(fmt, lyx::from_ascii("%4$s")));
1079 docstring str = subst(fmt, lyx::from_ascii("%1$s"), arg1);
1080 str = subst(str, lyx::from_ascii("%2$s"), arg2);
1081 str = subst(str, lyx::from_ascii("%3$s"), arg3);
1082 str = subst(str, lyx::from_ascii("%4$s"), arg4);
1083 return subst(str, lyx::from_ascii("%%"), lyx::from_ascii("%"));
1089 } // namespace support