void setIgnoreFormat(string const & type, bool value, bool fromUser)
{
- ignoreFormats.setIgnoreFormat(type, value, fromUser);
+ ignoreFormats.setIgnoreFormat(type, value, fromUser);
}
return temp2;
}
+static void buildAccentsMap();
+
string correctRegex(string t, bool withformat)
{
/* Convert \backslash => \
* and \{, \}, \[, \] => {, }, [, ]
*/
string s("");
- regex wordre("(\\\\)*(\\\\((backslash|mathcircumflex) ?|[\\[\\]\\{\\}]))");
+ regex wordre("(\\\\)*(\\\\(([A-Za-z]+|[\\{\\}])( |\\{\\})?|[\\[\\]\\{\\}]))");
size_t lastpos = 0;
smatch sub;
bool backslashed = false;
+ if (accents.empty())
+ buildAccentsMap();
+
+ //LYXERR0("correctRegex input '" << t << "'");
for (sregex_iterator it(t.begin(), t.end(), wordre), end; it != end; ++it) {
sub = *it;
string replace;
else {
if (sub.str(4) == "backslash") {
replace = "\\";
- if (withformat) {
+ {
// transforms '\backslash \{' into '\{'
- // and '\{' into '{'
string next = t.substr(sub.position(2) + sub.str(2).length(), 2);
if ((next == "\\{") || (next == "\\}")) {
replace = "";
replace = "^";
else if (backslashed) {
backslashed = false;
- if (withformat && (sub.str(3) == "{"))
- replace = accents["braceleft"];
- else if (withformat && (sub.str(3) == "}"))
- replace = accents["braceright"];
+ if (withformat) {
+ if (sub.str(3) == "{")
+ replace = accents["braceleft"];
+ else if (sub.str(3) == "}")
+ replace = accents["braceright"];
+ else {
+ // else part should not exist
+ LASSERT(1, /**/);
+ }
+ }
else {
- // else part should not exist
- LASSERT(1, /**/);
+ if (sub.str(3) == "{")
+ replace = "\\{";
+ else if (sub.str(3) == "}")
+ replace = "\\}";
+ else {
+ // else part should not exist
+ LASSERT(1, /**/);
+ }
+ }
+ }
+ else if (sub.str(4) == "{") // transforms '\{' into '{'
+ replace = "{";
+ else if (sub.str(4) == "}")
+ replace = "}";
+ else {
+ AccentsIterator it_ac = accents.find(sub.str(4));
+ if (it_ac == accents.end()) {
+ replace = sub.str(2);
+ }
+ else {
+ replace = it_ac->second;
}
}
- else
- replace = sub.str(3);
}
if (lastpos < (size_t) sub.position(2))
s += t.substr(lastpos, sub.position(2) - lastpos);
return t;
else if (lastpos < t.length())
s += t.substr(lastpos, t.length() - lastpos);
+ //LYXERR0("correctRegex output '" << s << "'");
return s;
}
if (lastpos == s.size())
break;
}
- size_t end_pos = s.find("\\endregexp{}}", regex_pos + 8);
+ size_t end_pos = s.find("\\endregexp", regex_pos + 8);
result += correctRegex(s.substr(regex_pos + 8, end_pos -(regex_pos + 8)), withformat);
lastpos = end_pos + 13;
}
** @todo Normalization should also expand macros, if the corresponding
** search option was checked.
**/
- string normalize(docstring const & s) const;
+ string normalize(docstring const & s, bool ignore_fomat) const;
// normalized string to search
string par_as_string;
// regular expression to use for searching
return ods.str();
}
+static string latexNamesToUtf8(docstring strIn)
+{
+ string addtmp = to_utf8(strIn);
+ static regex const rmAcc("(\\\\)*("
+ "\\\\([A-Za-z]+\\{.\\})" // e.g. "ddot{A}" == sub.str(3)
+ "|\\\\([A-Za-z]+)( |\\{\\})?" // e.g. "LyX", "LyX{}", "LyX " == sub.str(4)
+ ")"
+ );
+ size_t lastpos = 0;
+ smatch sub;
+ string replace;
+ string add("");
+ if (accents.empty())
+ buildAccentsMap();
+ for (sregex_iterator it_add(addtmp.begin(), addtmp.end(), rmAcc), end; it_add != end; ++it_add) {
+ sub = *it_add;
+ if ((sub.position(2) - sub.position(0)) % 2 == 1) {
+ continue;
+ }
+ else {
+ string key;
+ if (sub.length(3) > 0)
+ key = sub.str(3);
+ else
+ key = sub.str(4);
+ AccentsIterator it_ac = accents.find(key);
+ if (it_ac == accents.end()) {
+ replace = sub.str(2);
+ }
+ else {
+ replace = it_ac->second;
+ }
+ }
+ if (lastpos < (size_t) sub.position(2))
+ add += addtmp.substr(lastpos, sub.position(2) - lastpos);
+ add += replace;
+ lastpos = sub.position(2) + sub.length(2);
+ }
+ if (lastpos == 0)
+ add = addtmp;
+ else if (addtmp.length() > lastpos)
+ add += addtmp.substr(lastpos, addtmp.length() - lastpos);
+ LYXERR(Debug::FIND, "Adding to search string: '"
+ << add << "'");
+ return add;
+}
static docstring stringifySearchBuffer(Buffer & buffer, FindAndReplaceOptions const & opt)
{
if (ignoreFormats.getNonContent()) {
runparams.for_searchAdv |= OutputParams::SearchNonOutput;
}
+ string t("");
for (pos_type pit = pos_type(0); pit < (pos_type)buffer.paragraphs().size(); ++pit) {
Paragraph const & par = buffer.paragraphs().at(pit);
+ string add = latexNamesToUtf8(par.asString(pos_type(0), par.size(),
+ option,
+ &runparams));
LYXERR(Debug::FIND, "Adding to search string: '"
- << par.asString(pos_type(0), par.size(),
- option,
- &runparams)
- << "'");
- str += par.asString(pos_type(0), par.size(),
- option,
- &runparams);
+ << add << "'");
+ t += add;
}
// Even in ignore-format we have to remove "\text{}, \lyxmathsym{}" parts
- string t = to_utf8(str);
while (regex_replace(t, t, "\\\\(text|lyxmathsym|ensuremath)\\{([^\\}]*)\\}", "$2"));
str = from_utf8(t);
}
addAccents("\\o", getutf8(0x00f8));
addAccents("\\textcrlambda", getutf8(0x019b));
addAccents("\\j", getutf8(0x0237));
+ addAccents("\\textrevepsilon", getutf8(0x025c));
+ addAccents("\\textbaru", getutf8(0x0289));
addAccents("\\textquoteleft", getutf8(0x02bb));
addAccents("\\textGamma", getutf8(0x0393));
addAccents("\\Gamma", getutf8(0x0393));
{
if (accents.empty())
buildAccentsMap();
- static regex const accre("\\\\(([\\S]|grave|breve|ddot|dot|acute|dacute|mathring|check|hat|bar|tilde|subdot|ogonek|"
- "cedilla|subring|textsubring|subhat|textsubcircum|subtilde|textsubtilde|dgrave|textdoublegrave|rcap|textroundcap|slashed)\\{[^\\{\\}]+\\}"
- "|((i|imath|jmath|cdot|[a-z]+space)|((backslash )?([lL]y[xX]|[tT]e[xX]|[lL]a[tT]e[xX]e?|lyxarrow))|(textquote|brace|guillemot)(left|right)|textasciicircum|mathcircumflex|sim)(?![a-zA-Z]))");
+ static regex const accre("\\\\("
+ "([\\S]|[A-Za-z]+)\\{[^\\{\\}]+\\}"
+ "|("
+ "(backslash ([lL]y[xX]|[tT]e[xX]|[lL]a[tT]e[xX]e?|lyxarrow))"
+ "|[A-Za-z]+"
+ ")"
+ "(?![a-zA-Z]))");
smatch sub;
for (sregex_iterator itacc(par.begin(), par.end(), accre), end; itacc != end; ++itacc) {
sub = *itacc;
// Remove the key with all parameters and following spaces
size_t pos;
size_t start;
- if (interval_.par[actual._dataEnd-1] == ' ')
+ if (interval_.par[actual._dataEnd-1] == ' ' || interval_.par[actual._dataEnd-1] == '}')
start = actual._dataEnd;
else
start = actual._dataEnd+1;
balanced--;
if (balanced < 0)
break;
- }
- skip = 1;
}
+ skip = 1;
+ }
if (balanced != 0) {
regexIsValid = false;
regexError = "Unbalanced curly brackets in regexp \"" + regexp_str + "\"";
previous_single_replace = true;
}
// When using regexp, braces are hacked already by escape_for_regex()
- par_as_string = normalize(ds);
+ par_as_string = normalize(ds, opt.ignoreformat);
open_braces = 0;
close_wildcards = 0;
string lead_as_regexp;
if (lead_size > 0) {
lead_as_regexp = string2regex(par_as_string.substr(0, lead_size));
+ (void)regex_replace(par_as_string_nolead, par_as_string_nolead, "\\$$", "");
(void)regex_replace(par_as_string_nolead, par_as_string_nolead, "}$", "");
par_as_string = par_as_string_nolead;
LYXERR(Debug::FIND, "lead_as_regexp is '" << lead_as_regexp << "'");
docstring docstr = stringifyFromForSearch(opt, cur, len);
string str;
- str = normalize(docstr);
+ str = normalize(docstr, opt.ignoreformat);
if (!opt.ignoreformat) {
str = correctlanguagesetting(str, false, !opt.ignoreformat);
// remove closing '}' and '\n' to allow for use of '$' in regex
}
#endif
-string MatchStringAdv::normalize(docstring const & s) const
+string MatchStringAdv::normalize(docstring const & s, bool ignore_format) const
{
string t;
t = lyx::to_utf8(s);
// Remove \n at begin
while (!t.empty() && t[0] == '\n')
t = t.substr(1);
- // Remove \n at end
- while (!t.empty() && t[t.size() - 1] == '\n')
- t = t.substr(0, t.size() - 1);
+ // Remove [%]*\n at end
+ while (!t.empty() && t[t.size() - 1] == '\n') {
+ int count = 1;
+ if (!ignore_format) {
+ while ((t.size() > 1 + count) && (t[t.size() - 1 - count] == '%'))
+ count++;
+ }
+ t = t.substr(0, t.size() - count);
+ }
size_t pos;
// Handle all other '\n'
while ((pos = t.find("\n")) != string::npos) {
t.replace(pos-2, 3, "");
}
}
- else if (!isAlnumASCII(t[pos+1]) || !isAlnumASCII(t[pos-1])) {
- // '\n' adjacent to non-alpha-numerics, discard
- t.replace(pos, 1, "");
- }
else {
- // Replace all other \n with spaces
- t.replace(pos, 1, " ");
+ if (!isAlnumASCII(t[pos+1]) || !isAlnumASCII(t[pos-1])) {
+ // '\n' adjacent to non-alpha-numerics, discard
+ t.replace(pos, 1, "");
+ }
+ else {
+ // Replace all other \n with spaces
+ t.replace(pos, 1, " ");
+ }
+ if (!ignore_format) {
+ int count = 0;
+ while ((pos > count + 1) && (t[pos - 1 -count] == '%')) {
+ count++;
+ }
+ if (count > 0) {
+ t.replace(pos - count, count, "");
+ }
+ }
}
}
// Remove stale empty \emph{}, \textbf{} and similar blocks from latexify
runparams.for_searchAdv |= OutputParams::SearchNonOutput;
}
LYXERR(Debug::FIND, "Stringifying with cur: "
- << cur << ", from pos: " << cur.pos() << ", end: " << end);
- return par.asString(cur.pos(), end,
- option,
- &runparams);
+ << cur << ", from pos: " << cur.pos() << ", end: " << end);
+ return from_utf8(latexNamesToUtf8(par.asString(cur.pos(), end,
+ option,
+ &runparams)));
} else if (cur.inMathed()) {
CursorSlice cs = cur.top();
MathData md = cs.cell();
? md.end()
: md.begin() + cs.pos() + len );
MathData md2;
- for (MathData::const_iterator it = md.begin() + cs.pos();
- it != it_end; ++it)
+ for (MathData::const_iterator it = md.begin() + cs.pos(); it != it_end; ++it)
md2.push_back(*it);
- docstring s = asString(md2);
+ docstring s = from_utf8(latexNamesToUtf8(asString(md2)));
LYXERR(Debug::FIND, "Stringified math: '" << s << "'");
return s;
}
projects / features.git / blobdiff