4 * This file is part of LyX, the document processor.
5 * Licence details can be found in the file COPYING.
7 * \author Lars Gullik Bjønnes
8 * \author Jean-Marc Lasgouttes
10 * Full author contact details are available in file CREDITS.
12 * A collection of string helper functions that works with string.
13 * Some of these would certainly benefit from a rewrite/optimization.
19 #include "support/docstring.h"
27 /// Compare \p s and \p s2, ignoring the case.
28 /// Does not depend on the locale.
29 int compare_no_case(docstring const & s, docstring const & s2);
31 /// Compare \p s and \p s2 using the collating rules of the current locale.
32 int compare_locale(docstring const & s, docstring const & s2);
34 /// Compare \p s and \p s2, ignoring the case of ASCII characters only.
35 int compare_ascii_no_case(std::string const & s, std::string const & s2);
37 /// Compare \p s and \p s2, ignoring the case of ASCII characters only.
38 int compare_ascii_no_case(docstring const & s, docstring const & s2);
41 bool isStrInt(std::string const & str);
43 /// does the std::string represent an unsigned integer value ?
44 bool isStrUnsignedInt(std::string const & str);
47 bool isStrDbl(std::string const & str);
49 /// does the string contain a digit?
50 bool hasDigitASCII(docstring const & str);
52 bool isHexChar(char_type);
54 bool isHex(docstring const & str);
56 unsigned int hexToInt(docstring const & str);
58 /// is \p str pure ascii?
59 bool isAscii(docstring const & str);
61 /// is \p str pure ascii?
62 bool isAscii(std::string const & str);
65 * Changes the case of \p c to lowercase.
66 * Don't use this for non-ASCII characters, since it depends on the locale.
67 * This overloaded function is only implemented because the char_type variant
68 * would be used otherwise, and we assert in this function that \p c is in
71 char lowercase(char c);
74 * Changes the case of \p c to uppercase.
75 * Don't use this for non-ASCII characters, since it depends on the locale.
76 * This overloaded function is only implemented because the char_type variant
77 * would be used otherwise, and we assert in this function that \p c is in
80 char uppercase(char c);
82 /// Changes the case of \p c to lowercase.
83 /// Does not depend on the locale.
84 char_type lowercase(char_type c);
86 /// Changes the case of \p c to uppercase.
87 /// Does not depend on the locale.
88 char_type uppercase(char_type c);
90 /// Checks if the supplied character is lower-case
91 bool isLowerCase(char_type ch);
93 /// Checks if the supplied character is upper-case
94 bool isUpperCase(char_type ch);
96 /// same as lowercase(), but ignores locale
97 std::string const ascii_lowercase(std::string const &);
98 docstring const ascii_lowercase(docstring const &);
100 /// Changes the case of \p s to lowercase.
101 /// Does not depend on the locale.
102 docstring const lowercase(docstring const & s);
103 // Currently unused, but the code is there if needed.
104 // std::string const lowercase(std::string const & s);
106 /// Changes the case of \p s to uppercase.
107 /// Does not depend on the locale.
108 docstring const uppercase(docstring const & s);
110 /// Returns the superscript of \p c or \p c if no superscript exists.
111 /// Does not depend on the locale.
112 char_type superscript(char_type c);
114 /// Returns the subscript of \p c or \p c if no subscript exists.
115 /// Does not depend on the locale.
116 char_type subscript(char_type c);
118 /// Does str start with c?
119 bool prefixIs(docstring const & str, char_type c);
121 /// Does str start with pre?
122 bool prefixIs(std::string const & str, std::string const & pre);
123 bool prefixIs(docstring const & str, docstring const & pre);
125 /// Does the string end with this char?
126 bool suffixIs(std::string const &, char);
127 bool suffixIs(docstring const &, char_type);
129 /// Does the string end with this suffix?
130 bool suffixIs(std::string const &, std::string const &);
131 bool suffixIs(docstring const &, docstring const &);
133 /// Is b contained in a?
134 inline bool contains(std::string const & a, std::string const & b)
136 return a.find(b) != std::string::npos;
139 inline bool contains(docstring const & a, docstring const & b)
141 return a.find(b) != docstring::npos;
144 inline bool contains(std::string const & a, char b)
146 return a.find(b) != std::string::npos;
149 inline bool contains(docstring const & a, char_type b)
151 return a.find(b) != docstring::npos;
154 /// Returns true if the first argument is made of ascii chars given in the
156 bool containsOnly(std::string const &, std::string const &);
158 bool containsOnly(docstring const &, std::string const &);
160 /** Extracts a token from this string at the nth delim.
161 Doesn't modify the original string. Similar to strtok.
164 token("a;bc;d", ';', 1) == "bc";
165 token("a;bc;d", ';', 2) == "d";
168 std::string const token(std::string const & a, char delim, int n);
170 docstring const token(docstring const & a, char_type delim, int n);
172 /** Search a token in this string using the delim.
173 Doesn't modify the original string. Returns -1 in case of
177 tokenPos("a;bc;d", ';', "bc") == 1;
178 tokenPos("a;bc;d", ';', "d") == 2;
181 int tokenPos(std::string const & a, char delim, std::string const & tok);
182 int tokenPos(docstring const & a, char_type delim, docstring const & tok);
185 docstring capitalize(docstring const & s);
187 /// Substitute all \a oldchar with \a newchar
188 std::string const subst(std::string const & a, char oldchar, char newchar);
190 /// Substitute all \a oldchar with \a newchar
191 docstring const subst(docstring const & a, char_type oldchar, char_type newchar);
193 /// substitutes all instances of \a oldstr with \a newstr
194 std::string const subst(std::string const & a,
195 std::string const & oldstr, std::string const & newstr);
197 /// substitutes all instances of \a oldstr with \a newstr
198 docstring const subst(docstring const & a,
199 docstring const & oldstr, docstring const & newstr);
201 /// Count all occurrences of char \a chr inside \a str
202 int count_char(std::string const & str, char chr);
204 /// Count all occurrences of char \a chr inside \a str
205 int count_char(docstring const & str, docstring::value_type chr);
207 /// get an approximate word count
208 int wordCount(docstring const &);
210 /** Count all occurrences of binary chars inside \a str.
211 It is assumed that \a str is utf-8 encoded and that a binary char
212 belongs to the unicode class names Zl, Zp, Cc, Cf, Cs, Co, or Cn
213 (excluding white space characters such as '\t', '\n', '\v', '\f', '\r').
214 See http://www.unicode.org/Public/6.2.0/ucd/UnicodeData.txt
216 int count_bin_chars(std::string const & str);
218 /** Trims characters off the end and beginning of a string.
220 trim("ccabccc", "c") == "ab".
223 docstring const trim(docstring const & a, char const * p = " ");
225 /** Trims characters off the end and beginning of a string.
227 trim("ccabccc", "c") == "ab".
230 std::string const trim(std::string const & a, char const * p = " ");
232 /** Trims characters off the end of a string, removing any character
235 rtrim("abcde", "dec") == "ab".
238 std::string const rtrim(std::string const & a, char const * p = " ");
239 docstring const rtrim(docstring const & a, char const * p = " ");
241 /** Trims characters off the beginning of a string.
243 ("abbabcdef", "ab") = "cdef"
246 std::string const ltrim(std::string const & a, char const * p = " ");
247 docstring const ltrim(docstring const & a, char const * p = " ");
249 /** Splits the string given in the first argument at the first occurrence
250 of the third argument, delim.
251 What precedes delim is returned in the second argument, piece; this
252 will be the whole of the string if no delimiter is found.
253 The return value is what follows delim, if anything. So the return
254 value is the null string if no delimiter is found.
255 'a' and 'piece' must be different variables.
259 ret = split(s1, s2, ';') -> ret = "bc", s2 == "a"
262 std::string const split(std::string const & a, std::string & piece, char delim);
263 docstring const split(docstring const & a, docstring & piece, char_type delim);
265 /// Same as split but does not return a piece
266 std::string const split(std::string const & a, char delim);
268 /// Same as split but uses the last delim.
269 std::string const rsplit(std::string const & a, std::string & piece, char delim);
270 docstring const rsplit(docstring const & a, docstring & piece, char_type delim);
271 docstring const rsplit(docstring const & a, char_type delim);
273 /// Escapes non ASCII chars and other problematic characters that cause
274 /// problems in latex labels.
275 docstring const escape(docstring const & lab);
277 /// Group contents of an argument if needed
278 docstring const protectArgument(docstring & arg, char const l = '[',
281 /// Truncates a string with an ellipsis at the end. Leaves str unchanged and
282 /// returns false if it is shorter than len. Otherwise resizes str to len, with
283 /// U+2026 HORIZONTAL ELLIPSIS at the end, and returns true.
284 /// If mid is true, the ellipsis will be put to the mid of the string, and the first
285 /// and last half is appended/prepended.
287 /// Warning (Unicode): The cases where we want to truncate the text and it does
288 /// not end up converted into a QString for UI display must be really
289 /// rare. Whenever possible, we should prefer calling QFontMetrics::elidedText()
290 /// instead, which takes into account the actual length on the screen and the
291 /// layout direction (RTL or LTR). Or a similar function taking into account the
292 /// font metrics from the buffer view, which still has to be defined. Or set up
293 /// the widgets such that Qt elides the string automatically with the exact
294 /// needed width. Recall that not only graphemes vary greatly in width, but also
295 /// can be made of several code points. See:
296 /// <http://utf8everywhere.org/#myth.strlen>
298 /// What is acceptable is when we know that the string is probably going to be
299 /// elided by Qt anyway, and len is chosen such that our own ellipsis will only
300 /// be displayed in worst-case scenarios.
302 /// FIXME: apply those principles in the current code.
304 bool truncateWithEllipsis(docstring & str, size_t const len,
305 bool const mid = false);
307 /// Word-wraps the provided docstring, returning a line-broken string
308 /// of width no wider than width, with the string broken at spaces.
309 /// If the string cannot be broken appropriately, it returns something
310 /// with "..." at the end, again no wider than width.
311 /// We assume here that str does not contain newlines.
312 /// If indent is positive, then the first line is indented that many
313 /// spaces. If it is negative, then successive lines are indented, as
314 /// if the first line were "outdented".
316 /// Warning (Unicode): uses truncateWithEllipsis() internally. Therefore it is
317 /// subject to the same warning and FIXME as above.
319 docstring wrap(docstring const & str, int const indent = 0,
320 size_t const width = 80);
322 /// Like the preceding, except it is intended to operate on strings
323 /// that may contain embedded newlines.
324 /// \param numlines Don't return more than numlines lines. If numlines
325 /// is 0, we return everything.
327 /// Warning (Unicode): uses truncateWithEllipsis() internally. Therefore it is
328 /// subject to the same warning and FIXME as above.
330 docstring wrapParas(docstring const & str, int const indent = 0,
331 size_t const width = 80, size_t const maxlines = 10);
333 /// gives a vector of stringparts which have the delimiter delim
334 /// If \p keepempty is true, empty strings will be pushed to the vector as well
335 /// If \p trimit is true, leading and trailing whitespace will be trimmed from
336 /// all values. Note that this can affect what counts as "empty".
337 /// NOTE: If you want to split a string on whitespace, then do:
338 /// getVectorFromString(str, " ", false, true);
339 std::vector<std::string> const getVectorFromString(std::string const & str,
340 std::string const & delim = std::string(","),
341 bool keepempty = false, bool trimit = true);
342 std::vector<docstring> const getVectorFromString(docstring const & str,
343 docstring const & delim = from_ascii(","),
344 bool keepempty = false, bool trimit = true);
346 /// the same vice versa
347 std::string const getStringFromVector(std::vector<std::string> const & vec,
348 std::string const & delim = std::string(","));
349 docstring const getStringFromVector(std::vector<docstring> const & vec,
350 docstring const & delim = from_ascii(","));
352 /// Search \p search_token in \p str and return the position if it is
353 /// found, else -1. The last item in \p str must be "".
354 int findToken(char const * const str[], std::string const & search_token);
357 /// Format a floating point number with at least 6 significant digits, but
358 /// without scientific notation.
359 /// Scientific notation would be invalid in some contexts, such as lengths for
360 /// LaTeX. Simply using std::ostream with std::fixed would produce results
361 /// like "1000000.000000", and precision control would not be that easy either.
362 std::string formatFPNumber(double);
364 /// Returns an URI/URL-style percent-encoded copy of the string \p in.
365 /// \p ex defines a string of characters that are excluded from the transformation
366 docstring to_percent_encoding(docstring const & in, docstring const & ex = docstring());
368 /// Returns a string decoded from an URI/URL-style percent-encoded string \p in.
369 std::string from_percent_encoding(std::string const & in);
371 docstring bformat(docstring const & fmt, int arg1);
372 docstring bformat(docstring const & fmt, long arg1);
373 #ifdef HAVE_LONG_LONG_INT
374 docstring bformat(docstring const & fmt, long long arg1);
376 docstring bformat(docstring const & fmt, unsigned int arg1);
377 docstring bformat(docstring const & fmt, docstring const & arg1);
378 docstring bformat(docstring const & fmt, char * arg1);
379 docstring bformat(docstring const & fmt, docstring const & arg1, docstring const & arg2);
380 docstring bformat(docstring const & fmt, docstring const & arg1, int arg2);
381 docstring bformat(docstring const & fmt, char const * arg1, docstring const & arg2);
382 docstring bformat(docstring const & fmt, int arg1, int arg2);
383 docstring bformat(docstring const & fmt, docstring const & arg1, docstring const & arg2, docstring const & arg3);
384 docstring bformat(docstring const & fmt, docstring const & arg1, docstring const & arg2, docstring const & arg3, docstring const & arg4);
385 docstring bformat(docstring const & fmt, docstring const & arg1, docstring const & arg2, docstring const & arg3, docstring const & arg4, docstring const & arg5);
388 } // namespace support