namespace lyx {
+
+// Helper class for deciding what should be ignored
+class IgnoreFormats {
+ public:
+ ///
+ IgnoreFormats()
+ : ignoreFamily_(false), ignoreSeries_(false),
+ ignoreShape_(false), ignoreUnderline_(false),
+ ignoreMarkUp_(false), ignoreStrikeOut_(false),
+ ignoreSectioning_(false), ignoreFrontMatter_(true),
+ ignoreColor_(false), ignoreLanguage_(false) {}
+ ///
+ bool getFamily() { return ignoreFamily_; };
+ ///
+ bool getSeries() { return ignoreSeries_; };
+ ///
+ bool getShape() { return ignoreShape_; };
+ ///
+ bool getUnderline() { return ignoreUnderline_; };
+ ///
+ bool getMarkUp() { return ignoreMarkUp_; };
+ ///
+ bool getStrikeOut() { return ignoreStrikeOut_; };
+ ///
+ bool getSectioning() { return ignoreSectioning_; };
+ ///
+ bool getFrontMatter() { return ignoreFrontMatter_; };
+ ///
+ bool getColor() { return ignoreColor_; };
+ ///
+ bool getLanguage() { return ignoreLanguage_; };
+ ///
+ void setIgnoreFormat(string type, bool value);
+
+private:
+ ///
+ bool ignoreFamily_;
+ ///
+ bool ignoreSeries_;
+ ///
+ bool ignoreShape_;
+ ///
+ bool ignoreUnderline_;
+ ///
+ bool ignoreMarkUp_;
+ ///
+ bool ignoreStrikeOut_;
+ ///
+ bool ignoreSectioning_;
+ ///
+ bool ignoreFrontMatter_;
+ ///
+ bool ignoreColor_;
+ ///
+ bool ignoreLanguage_;
+};
+
+
+void IgnoreFormats::setIgnoreFormat(string type, bool value)
+{
+ if (type == "color") {
+ ignoreColor_ = value;
+ }
+ else if (type == "language") {
+ ignoreLanguage_ = value;
+ }
+ else if (type == "sectioning") {
+ ignoreSectioning_ = value;
+ ignoreFrontMatter_ = value;
+ }
+ else if (type == "font") {
+ ignoreSeries_ = value;
+ ignoreShape_ = value;
+ ignoreFamily_ = value;
+ }
+ else if (type == "series") {
+ ignoreSeries_ = value;
+ }
+ else if (type == "shape") {
+ ignoreShape_ = value;
+ }
+ else if (type == "family") {
+ ignoreFamily_ = value;
+ }
+ else if (type == "markup") {
+ ignoreMarkUp_ = value;
+ }
+ else if (type == "underline") {
+ ignoreUnderline_ = value;
+ }
+ else if (type == "strike") {
+ ignoreStrikeOut_ = value;
+ }
+}
+
+// The global variable that can be changed from outside
+IgnoreFormats ignoreFormats;
+
+
+void setIgnoreFormat(string type, bool value)
+{
+ ignoreFormats.setIgnoreFormat(type, value);
+}
+
+
namespace {
bool parse_bool(docstring & howto)
escape_map.push_back(P("\\]", "\\{\\]\\}"));
escape_map.push_back(P("\\^", "(?:\\^|\\\\textasciicircum\\{\\}|\\\\textasciicircum|\\\\mathcircumflex)"));
escape_map.push_back(P("%", "\\\\\\%"));
+ escape_map.push_back(P("#", "\\\\#"));
}
return escape_map;
}
string par_as_string;
// regular expression to use for searching
lyx::regex regexp;
- // same as regexp, but prefixed with a ".*"
+ // same as regexp, but prefixed with a ".*?"
lyx::regex regexp2;
// leading format material as string
string lead_as_string;
otexstream os(ods);
runparams.nice = true;
runparams.flavor = OutputParams::LATEX;
- runparams.linelen = 80; //lyxrc.plaintext_linelen;
+ runparams.linelen = 100000; //lyxrc.plaintext_linelen;
// No side effect of file copying and image conversion
runparams.dryrun = true;
runparams.for_search = true;
// @TODO Support \item[text]
// Kornel: Added textsl, textsf, textit, texttt and noun
// + allow to search for colored text too
- while (regex_replace(t, t, REGEX_BOS "\\\\(((emph|noun|minisec|text(bf|sl|sf|it|tt))|((textcolor|foreignlanguage)\\{[a-z]+\\})|(u|uu)line|(s|x)out|uwave)|((sub)?(((sub)?section)|paragraph)|part|chapter)\\*?)\\{", "")
+ while (regex_replace(t, t, REGEX_BOS "\\\\(((footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge|emph|noun|minisec|text(bf|md|sl|sf|it|tt))|((textcolor|foreignlanguage)\\{[a-z]+\\})|(u|uu)line|(s|x)out|uwave)|((sub)?(((sub)?section)|paragraph)|part|chapter)\\*?)\\{", "")
|| regex_replace(t, t, REGEX_BOS "\\$", "")
|| regex_replace(t, t, REGEX_BOS "\\\\\\[ ", "")
- || regex_replace(t, t, REGEX_BOS "\\\\item ", "")
+ || regex_replace(t, t, REGEX_BOS " ?\\\\item\\{[a-z]+\\}", "")
|| regex_replace(t, t, REGEX_BOS "\\\\begin\\{[a-zA-Z_]*\\*?\\} ", ""))
;
LYXERR(Debug::FIND, " after removing leading $, \\[ , \\emph{, \\textbf{, etc.: '" << t << "'");
static Features identifyFeatures(string const & s)
{
static regex const feature("\\\\(([a-z]+(\\{([a-z]+)\\}|\\*)?))\\{");
- static regex const valid("^(((emph|noun|text(bf|sl|sf|it|tt)|(textcolor|foreignlanguage)\\{[a-z]+\\})|item |(u|uu)line|(s|x)out|uwave)|((sub)?(((sub)?section)|paragraph)|part|chapter)\\*?)$");
+ static regex const valid("^(((footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge|emph|noun|text(bf|md|sl|sf|it|tt)|(textcolor|foreignlanguage|item)\\{[a-z]+\\})|(u|uu)line|(s|x)out|uwave)|((sub)?(((sub)?section)|paragraph)|part|chapter)\\*?)$");
smatch sub;
bool displ = true;
Features info;
class KeyInfo {
public:
enum KeyType {
+ /* Char type with content discarded
+ * like \hspace{1cm} */
+ noContent,
+ /* Char, like \backslash */
isChar,
+ /* \part, \section*, ... */
isSectioning,
- isMain, /* for \\foreignlanguage */
+ /* \foreignlanguage{ngerman}, ... */
+ isMain,
+ /* inside \code{} or \footnote{}
+ * to discard language in content */
+ noMain,
isRegex,
+ /* \begin{eqnarray}...\end{eqnarray}, ... $...$ */
isMath,
+ /* fonts, colors, markups, ... */
isStandard,
+ /* footnotesize, ... large, ...
+ * Ignore all of them */
isSize,
invalid,
+ /* inputencoding, shortcut, ...
+ * Discard also content, because they do not help in search */
doRemove,
- isIgnored /* to be ignored by creating infos */
+ /* twocolumns, ...
+ * like remove, but also all arguments */
+ removeWithArg,
+ /* item, listitem */
+ isList,
+ /* tex, latex, ... like isChar */
+ isIgnored,
+ /* like \lettrine[lines=5]{}{} */
+ cleanToStart,
+ /* End of arguments marker for lettrine,
+ * so that they can be ignored */
+ endArguments
};
KeyInfo()
: keytype(invalid),
head(""),
+ _tokensize(-1),
+ _tokenstart(-1),
+ _dataStart(-1),
+ _dataEnd(-1),
parenthesiscount(1),
- disabled(false)
+ disabled(false),
+ used(false)
{};
KeyInfo(KeyType type, int parcount, bool disable)
: keytype(type),
+ _tokensize(-1),
+ _tokenstart(-1),
+ _dataStart(-1),
+ _dataEnd(-1),
parenthesiscount(parcount),
- disabled(disable) {};
+ disabled(disable),
+ used(false) {};
KeyType keytype;
string head;
int _tokensize;
int _dataEnd;
int parenthesiscount;
bool disabled;
+ bool used; /* by pattern */
};
class Border {
#define MAXOPENED 30
class Intervall {
+ bool isPatternString;
public:
- Intervall() : ignoreidx(-1), actualdeptindex(0) {};
+ explicit Intervall(bool isPattern) :
+ isPatternString(isPattern),
+ ignoreidx(-1),
+ actualdeptindex(0) { depts[0] = 0; closes[0] = 0;};
string par;
int ignoreidx;
int depts[MAXOPENED];
int closes[MAXOPENED];
int actualdeptindex;
Border borders[2*MAXOPENED];
- // int previousNotIgnored(int);
+ int previousNotIgnored(int);
int nextNotIgnored(int);
void handleOpenP(int i);
void handleCloseP(int i, bool closingAllowed);
void addIntervall(int upper);
void addIntervall(int low, int upper); /* if explicit */
void setForDefaultLang(int upTo);
- int findclosing(int start, int end);
+ int findclosing(int start, int end, char up, char down, int repeat);
void handleParentheses(int lastpos, bool closingAllowed);
void output(ostringstream &os, int lastpos);
// string show(int lastpos);
closes[1] = -1;
}
-#if 0
int Intervall::previousNotIgnored(int start)
{
int idx = 0; /* int intervalls */
for (idx = ignoreidx; idx >= 0; --idx) {
if (start > borders[idx].upper)
- return(start);
+ return start;
if (start >= borders[idx].low)
start = borders[idx].low-1;
}
return start;
}
-#endif
int Intervall::nextNotIgnored(int start)
{
int idx = 0; /* int intervalls */
for (idx = 0; idx <= ignoreidx; idx++) {
if (start < borders[idx].low)
- return(start);
+ return start;
if (start < borders[idx].upper)
start = borders[idx].upper;
}
private:
int entidx;
Entries entries;
- KeyInfo analyze(string key);
Intervall interval;
- void buildKeys();
- void buildEntries();
- void makeKey(const string &, KeyInfo);
+ void buildKeys(bool);
+ void buildEntries(bool);
+ void makeKey(const string &, KeyInfo, bool isPatternString);
void processRegion(int start, int region_end); /* remove {} parts */
void removeHead(KeyInfo&, int count=0);
+
public:
- LatexInfo(string par) {
+ LatexInfo(string par, bool isPatternString) : entidx(-1), interval(isPatternString) {
interval.par = par;
- buildKeys();
+ buildKeys(isPatternString);
entries = vector<KeyInfo>();
- buildEntries();
+ buildEntries(isPatternString);
};
int getFirstKey() {
entidx = 0;
return true;
}
else
- return(false);
+ return false;
+ };
+ int find(int start, KeyInfo::KeyType keytype) {
+ if (start < 0)
+ return (-1);
+ int tmpIdx = start;
+ while (tmpIdx < int(entries.size())) {
+ if (entries[tmpIdx].keytype == keytype)
+ return tmpIdx;
+ tmpIdx++;
+ }
+ return(-1);
};
int process(ostringstream &os, KeyInfo &actual);
int dispatch(ostringstream &os, int previousStart, KeyInfo &actual);
return entries[keyinfo];
};
void setForDefaultLang(int upTo) {interval.setForDefaultLang(upTo);};
-
+ void addIntervall(int low, int up) { interval.addIntervall(low, up); };
};
-int Intervall::findclosing(int start, int end)
+int Intervall::findclosing(int start, int end, char up = '{', char down = '}', int repeat = 1)
{
int skip = 0;
int depth = 0;
+ repeat--;
for (int i = start; i < end; i += 1 + skip) {
char c;
c = par[i];
skip = 0;
if (c == '\\') skip = 1;
- else if (c == '{') {
+ else if (c == up) {
depth++;
}
- else if (c == '}') {
- if (depth == 0) return(i);
+ else if (c == down) {
+ if (depth == 0) {
+ if ((repeat <= 0) || (par[i+1] != up))
+ return i;
+ }
--depth;
}
}
- return(end);
+ return end;
}
-void LatexInfo::buildEntries()
+class MathInfo {
+ class MathEntry {
+ public:
+ string wait;
+ size_t mathEnd;
+ size_t mathStart;
+ size_t mathSize;
+ };
+ size_t actualIdx;
+ vector<MathEntry> entries;
+ public:
+ MathInfo() {
+ actualIdx = 0;
+ }
+ void insert(string wait, size_t start, size_t end) {
+ MathEntry m = MathEntry();
+ m.wait = wait;
+ m.mathStart = start;
+ m.mathEnd = end;
+ m.mathSize = end - start;
+ entries.push_back(m);
+ }
+ bool empty() { return entries.empty(); };
+ size_t getEndPos() {
+ if (entries.empty() || (actualIdx >= entries.size())) {
+ return 0;
+ }
+ return entries[actualIdx].mathEnd;
+ }
+ size_t getStartPos() {
+ if (entries.empty() || (actualIdx >= entries.size())) {
+ return 100000; /* definitely enough? */
+ }
+ return entries[actualIdx].mathStart;
+ }
+ size_t getFirstPos() {
+ actualIdx = 0;
+ return getStartPos();
+ }
+ size_t getSize() {
+ if (entries.empty() || (actualIdx >= entries.size())) {
+ return size_t(0);
+ }
+ return entries[actualIdx].mathSize;
+ }
+ void incrEntry() { actualIdx++; };
+};
+
+void LatexInfo::buildEntries(bool isPatternString)
{
- static regex const rmath("\\\\(begin|end)\\{((eqnarray|equation|flalign|gather|multiline|align)\\*?)\\}");
- static regex const rkeys("\\\\((([a-zA-Z]+\\*?)(\\{([a-z]+\\*?)\\}|=[0-9]+[a-z]+)?))");
+ static regex const rmath("\\$|\\\\\\[|\\\\\\]|\\\\(begin|end)\\{((eqnarray|equation|flalign|gather|multline|align|alignat)\\*?)\\}");
+ static regex const rkeys("\\$|\\\\\\[|\\\\\\]|\\\\((([a-zA-Z]+\\*?)(\\{([a-z]+\\*?)\\}|=[0-9]+[a-z]+)?))");
+ static bool disableLanguageOverride = false;
smatch sub, submath;
bool evaluatingRegexp = false;
+ MathInfo mi;
+ bool evaluatingMath = false;
+ bool evaluatingCode = false;
+ size_t codeEnd = 0;
+ bool evaluatingOptional = false;
+ size_t optionalEnd = 0;
+ int codeStart = -1;
KeyInfo found;
bool math_end_waiting = false;
size_t math_pos = 10000;
- int math_size = 0;
- int math_end_pos = -1;
string math_end;
for (sregex_iterator itmath(interval.par.begin(), interval.par.end(), rmath), end; itmath != end; ++itmath) {
submath = *itmath;
if (math_end_waiting) {
- if ((submath.str(1).compare("end") == 0) &&
+ size_t pos = submath.position(size_t(0));
+ if ((math_end == "$") &&
+ (submath.str(0) == "$") &&
+ (interval.par[pos-1] != '\\')) {
+ mi.insert("$", math_pos, pos + 1);
+ math_end_waiting = false;
+ }
+ else if ((math_end == "\\]") &&
+ (submath.str(0) == "\\]")) {
+ mi.insert("\\]", math_pos, pos + 2);
+ math_end_waiting = false;
+ }
+ else if ((submath.str(1).compare("end") == 0) &&
(submath.str(2).compare(math_end) == 0)) {
- math_size = submath.position(0) + submath.str(0).length() - math_pos;
+ mi.insert(math_end, math_pos, pos + submath.str(0).length());
math_end_waiting = false;
}
+ else
+ continue;
}
else {
if (submath.str(1).compare("begin") == 0) {
math_end_waiting = true;
math_end = submath.str(2);
- math_pos = submath.position(0);
+ math_pos = submath.position(size_t(0));
+ }
+ else if (submath.str(0).compare("\\[") == 0) {
+ math_end_waiting = true;
+ math_end = "\\]";
+ math_pos = submath.position(size_t(0));
+ }
+ else if (submath.str(0) == "$") {
+ size_t pos = submath.position(size_t(0));
+ if ((pos == 0) || (interval.par[pos-1] != '\\')) {
+ math_end_waiting = true;
+ math_end = "$";
+ math_pos = pos;
+ }
}
}
}
+ // Ignore language if there is math somewhere in pattern-string
+ if (isPatternString) {
+ if (! mi.empty()) {
+ // Disable language
+ keys["foreignlanguage"].disabled = true;
+ disableLanguageOverride = true;
+ }
+ else
+ disableLanguageOverride = false;
+ }
+ else {
+ if (disableLanguageOverride) {
+ keys["foreignlanguage"].disabled = true;
+ }
+ }
+ math_pos = mi.getFirstPos();
for (sregex_iterator it(interval.par.begin(), interval.par.end(), rkeys), end; it != end; ++it) {
sub = *it;
+ string key = sub.str(3);
+ if (key == "") {
+ if (sub.str(0)[0] == '\\')
+ key = sub.str(0)[1];
+ else {
+ key = sub.str(0);
+ if (key == "$") {
+ size_t k_pos = sub.position(size_t(0));
+ if ((k_pos > 0) && (interval.par[k_pos - 1] == '\\')) {
+ // Escaped '$', ignoring
+ continue;
+ }
+ }
+ }
+ };
if (evaluatingRegexp) {
if (sub.str(1).compare("endregexp") == 0) {
evaluatingRegexp = false;
// found._tokenstart already set
- found._dataEnd = sub.position(0) + 13;
+ found._dataEnd = sub.position(size_t(0)) + 13;
found._dataStart = found._dataEnd;
found._tokensize = found._dataEnd - found._tokenstart;
found.parenthesiscount = 0;
}
}
else {
- if (keys.find(sub.str(3)) == keys.end()) {
- LYXERR(Debug::FIND, "Found unknown key " << sub.str(0));
- continue;
+ if (evaluatingMath) {
+ if (size_t(sub.position(size_t(0))) < mi.getEndPos())
+ continue;
+ evaluatingMath = false;
+ mi.incrEntry();
+ math_pos = mi.getStartPos();
}
- found = keys[sub.str(3)];
- if (sub.str(3).compare("regexp") == 0) {
+ if (keys.find(key) == keys.end()) {
+ found = KeyInfo(KeyInfo::isStandard, 0, true);
+ if (isPatternString) {
+ found.keytype = KeyInfo::isChar;
+ found.disabled = false;
+ found.used = true;
+ }
+ keys[key] = found;
+ }
+ else
+ found = keys[key];
+ if (key.compare("regexp") == 0) {
evaluatingRegexp = true;
- found._tokenstart = sub.position(0);
+ found._tokenstart = sub.position(size_t(0));
found._tokensize = 0;
continue;
}
if (found.keytype == KeyInfo::isIgnored)
continue;
else if (found.keytype == KeyInfo::isMath) {
- if (size_t(sub.position(0)) == math_pos) {
- found = keys[sub.str(3)];
- found._tokenstart = sub.position(0);
- found._tokensize = math_size;
+ if (size_t(sub.position(size_t(0))) == math_pos) {
+ found = keys[key];
+ found._tokenstart = sub.position(size_t(0));
+ found._tokensize = mi.getSize();
found._dataEnd = found._tokenstart + found._tokensize;
found._dataStart = found._dataEnd;
found.parenthesiscount = 0;
- math_end_pos = found._dataEnd;
+ evaluatingMath = true;
+ }
+ else {
+ // begin|end of unknown env, discard
+ // First handle tables
+ // longtable|tabular
+ bool discardComment;
+ found = keys[key];
+ found.keytype = KeyInfo::doRemove;
+ if ((sub.str(5).compare("longtable") == 0) ||
+ (sub.str(5).compare("tabular") == 0)) {
+ discardComment = true; /* '%' */
+ }
+ else {
+ discardComment = false;
+ static regex const removeArgs("^(multicols|multipar|sectionbox|subsectionbox|tcolorbox)$");
+ smatch sub2;
+ string token = sub.str(5);
+ if (regex_match(token, sub2, removeArgs)) {
+ found.keytype = KeyInfo::removeWithArg;
+ }
+ }
+ // discard spaces before pos(0)
+ int pos = sub.position(size_t(0));
+ int count;
+ for (count = 0; pos - count > 0; count++) {
+ char c = interval.par[pos-count-1];
+ if (discardComment) {
+ if ((c != ' ') && (c != '%'))
+ break;
+ }
+ else if (c != ' ')
+ break;
+ }
+ found._tokenstart = pos - count;
+ if (sub.str(1).compare(0, 5, "begin") == 0) {
+ size_t pos1 = pos + sub.str(0).length();
+ if (sub.str(5).compare("cjk") == 0) {
+ pos1 = interval.findclosing(pos1+1, interval.par.length()) + 1;
+ if ((interval.par[pos1] == '{') && (interval.par[pos1+1] == '}'))
+ pos1 += 2;
+ found.keytype = KeyInfo::isMain;
+ found._dataStart = pos1;
+ found._dataEnd = interval.par.length();
+ found.disabled = keys["foreignlanguage"].disabled;
+ found.used = keys["foreignlanguage"].used;
+ found._tokensize = pos1 - found._tokenstart;
+ found.head = interval.par.substr(found._tokenstart, found._tokensize);
+ }
+ else {
+ while (interval.par[pos1] == '[') {
+ pos1 = interval.findclosing(pos1+1, interval.par.length(), '[', ']')+1;
+ }
+ if (interval.par[pos1] == '{') {
+ found._dataEnd = interval.findclosing(pos1+1, interval.par.length()) + 1;
+ }
+ else {
+ found._dataEnd = pos1;
+ }
+ found._dataStart = found._dataEnd;
+ found._tokensize = count + found._dataEnd - pos;
+ found.parenthesiscount = 0;
+ found.disabled = true;
+ }
+ }
+ else {
+ // Handle "\end{...}"
+ found._dataStart = pos + sub.str(0).length();
+ found._dataEnd = found._dataStart;
+ found._tokensize = count + found._dataEnd - pos;
+ found.parenthesiscount = 0;
+ found.disabled = true;
+ }
}
- else
- continue;
}
else if (found.keytype != KeyInfo::isRegex) {
- found._tokenstart = sub.position(0);
- if (found._tokenstart < math_end_pos) {
- // Ignore if we are inside math equation
- continue;
- }
+ found._tokenstart = sub.position(size_t(0));
if (found.parenthesiscount == 0) {
// Probably to be discarded
- char following = interval.par[sub.position(0) + sub.str(3).length() + 1];
+ size_t following_pos = sub.position(size_t(0)) + sub.str(3).length() + 1;
+ char following = interval.par[following_pos];
if (following == ' ')
found.head = "\\" + sub.str(3) + " ";
else if (following == '=') {
found.head = sub.str(0);
}
else
- found.head = "\\" + sub.str(3);
+ found.head = "\\" + key;
found._tokensize = found.head.length();
found._dataEnd = found._tokenstart + found._tokensize;
found._dataStart = found._dataEnd;
}
else {
+ int params = found._tokenstart + key.length() + 1;
+ if (evaluatingOptional) {
+ if (size_t(found._tokenstart) > optionalEnd) {
+ evaluatingOptional = false;
+ }
+ else {
+ found.disabled = true;
+ }
+ }
+ int optend = params;
+ while (interval.par[optend] == '[') {
+ // discard optional parameters
+ optend = interval.findclosing(optend+1, interval.par.length(), '[', ']') + 1;
+ }
+ if (optend > params) {
+ key += interval.par.substr(params, optend-params);
+ evaluatingOptional = true;
+ optionalEnd = optend;
+ }
+ string token = sub.str(5);
+ int closings = found.parenthesiscount;
if (found.parenthesiscount == 1) {
- found.head = "\\" + sub.str(3) + "{";
+ found.head = "\\" + key + "{";
}
- else if (found.parenthesiscount == 2) {
- found.head = sub.str(0) + "{";
- found._tokensize = found.head.length();
+ else if (found.parenthesiscount > 1) {
+ if (token != "") {
+ found.head = sub.str(0) + "{";
+ closings = found.parenthesiscount - 1;
+ }
+ else {
+ found.head = "\\" + key + "{";
+ }
}
- found._tokensize = found.head.length();
found._dataStart = found._tokenstart + found.head.length();
- found._dataEnd = interval.findclosing(found._dataStart, interval.par.length());
+ if (interval.par.substr(found._dataStart-1, 15).compare("\\endarguments{}") == 0) {
+ found._dataStart += 15;
+ }
+ size_t endpos = interval.findclosing(found._dataStart, interval.par.length(), '{', '}', closings);
+ if (found.keytype == KeyInfo::isList) {
+ // Check if it really is list env
+ static regex const listre("^([a-z]+)$");
+ smatch sub2;
+ if (!regex_match(token, sub2, listre)) {
+ // Change the key of this entry. It is not in a list/item environment
+ found.keytype = KeyInfo::endArguments;
+ }
+ }
+ if (found.keytype == KeyInfo::noMain) {
+ evaluatingCode = true;
+ codeEnd = endpos;
+ codeStart = found._dataStart;
+ }
+ else if (evaluatingCode) {
+ if (size_t(found._dataStart) > codeEnd)
+ evaluatingCode = false;
+ else if (found.keytype == KeyInfo::isMain) {
+ // Disable this key, treate it as standard
+ found.keytype = KeyInfo::isStandard;
+ found.disabled = true;
+ if ((codeEnd == interval.par.length()) &&
+ (found._tokenstart == codeStart)) {
+ // trickery, because the code inset starts
+ // with \selectlanguage ...
+ codeEnd = endpos;
+ if (entries.size() > 1) {
+ entries[entries.size()-1]._dataEnd = codeEnd;
+ }
+ }
+ }
+ }
+ if ((endpos == interval.par.length()) &&
+ (found.keytype == KeyInfo::doRemove)) {
+ // Missing closing => error in latex-input?
+ // therefore do not delete remaining data
+ found._dataStart -= 1;
+ found._dataEnd = found._dataStart;
+ }
+ else
+ found._dataEnd = endpos;
+ }
+ if (isPatternString) {
+ keys[key].used = true;
}
}
entries.push_back(found);
}
}
-void LatexInfo::makeKey(const string &keysstring, KeyInfo keyI)
+void LatexInfo::makeKey(const string &keysstring, KeyInfo keyI, bool isPatternString)
{
stringstream s(keysstring);
string key;
- KeyInfo keyII(keyI);
const char delim = '|';
while (getline(s, key, delim)) {
+ KeyInfo keyII(keyI);
+ if (isPatternString) {
+ keyII.used = false;
+ }
+ else if ( !keys[key].used)
+ keyII.disabled = true;
keys[key] = keyII;
}
}
-void LatexInfo::buildKeys()
+void LatexInfo::buildKeys(bool isPatternString)
{
- static bool keysBuilt = false;
- static bool ignoreFamily = false;
- static bool ignoreSeries = false;
- static bool ignoreShape = false;
- static bool ignoreUnderline = false;
- static bool ignoreMarkUp = false;
- static bool ignoreStrikeOut = false;
- static bool ignoreSectioning = false;
- static bool ignoreFrontMatter= true;
- static bool ignoreColor = false;
- static bool ignoreLanguage = false;
- if (keysBuilt) return;
+ static bool keysBuilt = false;
+ if (keysBuilt && !isPatternString) return;
- // Know statdard keys with 1 parameter.
+ // Known standard keys with 1 parameter.
// Split is done, if not at start of region
- makeKey("textsf|textss|texttt", KeyInfo(KeyInfo::isStandard, 1, ignoreFamily));
- makeKey("textbf", KeyInfo(KeyInfo::isStandard, 1, ignoreSeries));
- makeKey("textit|textsc|textsl", KeyInfo(KeyInfo::isStandard, 1, ignoreShape));
- makeKey("uuline|uline|uwave", KeyInfo(KeyInfo::isStandard, 1, ignoreUnderline));
- makeKey("emph|noun", KeyInfo(KeyInfo::isStandard, 1, ignoreMarkUp));
- makeKey("sout|xout", KeyInfo(KeyInfo::isStandard, 1, ignoreStrikeOut));
-
+ makeKey("textsf|textss|texttt", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getFamily()), isPatternString);
+ makeKey("textbf", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getSeries()), isPatternString);
+ makeKey("textit|textsc|textsl", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getShape()), isPatternString);
+ makeKey("uuline|uline|uwave", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getUnderline()), isPatternString);
+ makeKey("emph|noun", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getMarkUp()), isPatternString);
+ makeKey("sout|xout", KeyInfo(KeyInfo::isStandard, 1, ignoreFormats.getStrikeOut()), isPatternString);
makeKey("section|subsection|subsubsection|paragraph|subparagraph|minisec",
- KeyInfo(KeyInfo::isSectioning, 1, ignoreSectioning));
+ KeyInfo(KeyInfo::isSectioning, 1, ignoreFormats.getSectioning()), isPatternString);
makeKey("section*|subsection*|subsubsection*|paragraph*",
- KeyInfo(KeyInfo::isSectioning, 1, ignoreSectioning));
- makeKey("part|part*|chapter|chapter*", KeyInfo(KeyInfo::isSectioning, 1, ignoreSectioning));
- makeKey("title|subtitle|author|subject|publishers|dedication|uppertitleback|lowertitleback|extratitle|lyxaddress|lyxrightaddress", KeyInfo(KeyInfo::isSectioning, 1, ignoreFrontMatter));
+ KeyInfo(KeyInfo::isSectioning, 1, ignoreFormats.getSectioning()), isPatternString);
+ makeKey("part|part*|chapter|chapter*", KeyInfo(KeyInfo::isSectioning, 1, ignoreFormats.getSectioning()), isPatternString);
+ makeKey("title|subtitle|author|subject|publishers|dedication|uppertitleback|lowertitleback|extratitle|lyxaddress|lyxrightaddress", KeyInfo(KeyInfo::isSectioning, 1, ignoreFormats.getFrontMatter()), isPatternString);
// Regex
- makeKey("regexp", KeyInfo(KeyInfo::isRegex, 1, false));
+ makeKey("regexp", KeyInfo(KeyInfo::isRegex, 1, false), isPatternString);
// Split is done, if not at start of region
- makeKey("textcolor", KeyInfo(KeyInfo::isStandard, 2, ignoreColor));
+ makeKey("textcolor", KeyInfo(KeyInfo::isStandard, 2, ignoreFormats.getColor()), isPatternString);
// Split is done always.
- makeKey("foreignlanguage", KeyInfo(KeyInfo::isMain, 2, ignoreLanguage));
+ makeKey("foreignlanguage", KeyInfo(KeyInfo::isMain, 2, ignoreFormats.getLanguage()), isPatternString);
// Know charaters
// No split
- makeKey("backslash|textbackslash", KeyInfo(KeyInfo::isChar, 1, false));
-
+ makeKey("backslash|textbackslash|slash", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("textasciicircum|textasciitilde", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("textasciiacute|texemdash", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("dots|ldots", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ // Spaces
+ makeKey("quad|qquad|hfill|dotfill", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("textvisiblespace|nobreakspace", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("negthickspace|negmedspace|negthinspace", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ // Skip
+ // makeKey("enskip|smallskip|medskip|bigskip|vfill", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ // Custom space/skip, remove the content (== length value)
+ makeKey("vspace|hspace|mspace", KeyInfo(KeyInfo::noContent, 1, false), isPatternString);
+ // Found in fr/UserGuide.lyx
+ makeKey("og|fg", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ // quotes
+ makeKey("textquotedbl|quotesinglbase|lyxarrow", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("textquotedblleft|textquotedblright", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
// Known macros to remove (including their parameter)
// No split
- makeKey("inputencoding|shortcut", KeyInfo(KeyInfo::doRemove, 1, false));
+ makeKey("inputencoding|shortcut|label|ref|index|bibitem", KeyInfo(KeyInfo::doRemove, 1, false), isPatternString);
+ makeKey("addtocounter|setlength", KeyInfo(KeyInfo::noContent, 2, true), isPatternString);
+ // handle like standard keys with 1 parameter.
+ makeKey("url|href|vref|thanks", KeyInfo(KeyInfo::isStandard, 1, false), isPatternString);
// Macros to remove, but let the parameter survive
// No split
- makeKey("url|href|menuitem|footnote|code", KeyInfo(KeyInfo::isStandard, 1, true));
+ makeKey("menuitem|textmd|textrm", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+
+ // Remove language spec from content of these insets
+ makeKey("code", KeyInfo(KeyInfo::noMain, 1, false), isPatternString);
// Same effect as previous, parameter will survive (because there is no one anyway)
// No split
- makeKey("noindent", KeyInfo(KeyInfo::isStandard, 0, true));
- // like ('tiny{}' or '\tiny ' ... }
- makeKey("footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge", KeyInfo(KeyInfo::isSize, 0, true));
+ makeKey("noindent|textcompwordmark", KeyInfo(KeyInfo::isStandard, 0, true), isPatternString);
+ // Remove table decorations
+ makeKey("hline|tabularnewline|toprule|bottomrule|midrule", KeyInfo(KeyInfo::doRemove, 0, true), isPatternString);
+ // Discard shape-header
+ makeKey("circlepar|diamondpar|heartpar|nutpar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+ makeKey("trianglerightpar|hexagonpar|starpar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+ makeKey("triangleuppar|triangledownpar|droppar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+ makeKey("triangleleftpar|shapepar|dropuppar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+ makeKey("hphantom|footnote|includegraphics", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
+ // like ('tiny{}' or '\tiny ' ... )
+ makeKey("footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge", KeyInfo(KeyInfo::isSize, 0, false), isPatternString);
// Survives, like known character
- makeKey("lyx", KeyInfo(KeyInfo::isIgnored, 0, false));
-
- makeKey("begin", KeyInfo(KeyInfo::isMath, 1, false));
- makeKey("end", KeyInfo(KeyInfo::isMath, 1, false));
-
- makeKey("par|uldepth|ULdepth", KeyInfo(KeyInfo::doRemove, 0, true));
-
- keysBuilt = true;
+ makeKey("lyx|LyX|latex|LaTeX|latexe|LaTeXe|tex|TeX", KeyInfo(KeyInfo::isChar, 0, false), isPatternString);
+ makeKey("item|listitem", KeyInfo(KeyInfo::isList, 1, false), isPatternString);
+
+ makeKey("begin|end", KeyInfo(KeyInfo::isMath, 1, false), isPatternString);
+ makeKey("[|]", KeyInfo(KeyInfo::isMath, 1, false), isPatternString);
+ makeKey("$", KeyInfo(KeyInfo::isMath, 1, false), isPatternString);
+
+ makeKey("par|uldepth|ULdepth|protect|nobreakdash|medskip|relax", KeyInfo(KeyInfo::isStandard, 0, true), isPatternString);
+ // Remove RTL/LTR marker
+ makeKey("l|r|textlr|textfr|textar|beginl|endl", KeyInfo(KeyInfo::isStandard, 0, true), isPatternString);
+ makeKey("lettrine", KeyInfo(KeyInfo::cleanToStart, 0, true), isPatternString);
+ makeKey("lyxslide", KeyInfo(KeyInfo::isSectioning, 1, true), isPatternString);
+ makeKey("endarguments", KeyInfo(KeyInfo::endArguments, 0, true), isPatternString);
+ makeKey("twocolumn", KeyInfo(KeyInfo::removeWithArg, 2, true), isPatternString);
+ makeKey("tnotetext|ead|fntext|cortext|address", KeyInfo(KeyInfo::removeWithArg, 0, true), isPatternString);
+ makeKey("lyxend", KeyInfo(KeyInfo::isStandard, 0, true), isPatternString);
+ if (isPatternString) {
+ // Allow the first searched string to rebuild the keys too
+ keysBuilt = false;
+ }
+ else {
+ // no need to rebuild again
+ keysBuilt = true;
+ }
}
/*
string Intervall::show(int lastpos)
{
int idx = 0; /* int intervalls */
- int count = 0;
string s;
int i = 0;
for (idx = 0; idx <= ignoreidx; idx++) {
int i = 0;
for (idx = 0; idx <= ignoreidx; idx++) {
if (i < lastpos) {
- int printsize;
if (i <= borders[idx].low) {
+ int printsize;
if (borders[idx].low > lastpos)
printsize = lastpos - i;
else
for (int i = actualdeptindex; i > 0; --i) {
os << "}";
}
+ if (! isPatternString)
+ os << "\n";
handleParentheses(lastpos, true); /* extra closings '}' allowed here */
}
void LatexInfo::processRegion(int start, int region_end)
{
- while (start < region_end) {
- if (interval.par[start] == '{') {
+ while (start < region_end) { /* Let {[} and {]} survive */
+ if ((interval.par[start] == '{') &&
+ (interval.par[start+1] != ']') &&
+ (interval.par[start+1] != '[')) {
// Closing is allowed past the region
int closing = interval.findclosing(start+1, interval.par.length());
interval.addIntervall(start, start+1);
}
else {
// Remove header hull, that is "\url{abcd}" ==> "abcd"
- interval.addIntervall(actual._tokenstart, actual._dataStart);
+ interval.addIntervall(actual._tokenstart - count, actual._dataStart);
interval.addIntervall(actual._dataEnd, actual._dataEnd+1);
}
}
int LatexInfo::dispatch(ostringstream &os, int previousStart, KeyInfo &actual)
{
- int nextKeyIdx;
+ int nextKeyIdx = 0;
switch (actual.keytype)
- {
+ {
+ case KeyInfo::cleanToStart: {
+ actual._dataEnd = actual._dataStart;
+ nextKeyIdx = getNextKey();
+ // Search for end of arguments
+ int tmpIdx = find(nextKeyIdx, KeyInfo::endArguments);
+ if (tmpIdx > 0) {
+ for (int i = nextKeyIdx; i <= tmpIdx; i++) {
+ entries[i].disabled = true;
+ }
+ actual._dataEnd = entries[tmpIdx]._dataEnd;
+ }
+ while (interval.par[actual._dataEnd] == ' ')
+ actual._dataEnd++;
+ interval.addIntervall(0, actual._dataEnd+1);
+ interval.actualdeptindex = 0;
+ interval.depts[0] = actual._dataEnd+1;
+ interval.closes[0] = -1;
+ break;
+ }
+ case KeyInfo::noContent: { /* char like "\hspace{2cm}" */
+ if (actual.disabled)
+ interval.addIntervall(actual._tokenstart, actual._dataEnd);
+ else
+ interval.addIntervall(actual._dataStart, actual._dataEnd);
+ }
+ // fall through
case KeyInfo::isChar: {
nextKeyIdx = getNextKey();
break;
}
case KeyInfo::isSize: {
- if (actual.disabled) {
- // Allways disabled
+ if (actual.disabled || (interval.par[actual._dataStart] != '{') || (interval.par[actual._dataStart-1] == ' ')) {
processRegion(actual._dataEnd, actual._dataEnd+1); /* remove possibly following {} */
interval.addIntervall(actual._tokenstart, actual._dataEnd+1);
nextKeyIdx = getNextKey();
} else {
+ // Here _dataStart points to '{', so correct it
+ actual._dataStart += 1;
+ actual._tokensize += 1;
+ actual.parenthesiscount = 1;
+ if (interval.par[actual._dataStart] == '}') {
+ // Determine the end if used like '{\tiny{}...}'
+ actual._dataEnd = interval.findclosing(actual._dataStart+1, interval.par.length()) + 1;
+ interval.addIntervall(actual._dataStart, actual._dataStart+1);
+ }
+ else {
+ // Determine the end if used like '\tiny{...}'
+ actual._dataEnd = interval.findclosing(actual._dataStart, interval.par.length()) + 1;
+ }
// Split on this key if not at start
int start = interval.nextNotIgnored(previousStart);
if (start < actual._tokenstart) {
}
break;
}
+ case KeyInfo::endArguments:
+ // Remove trailing '{}' too
+ actual._dataStart += 2;
+ actual._dataEnd += 2;
+ interval.addIntervall(actual._tokenstart, actual._dataEnd);
+ nextKeyIdx = getNextKey();
+ break;
+ case KeyInfo::noMain:
+ // fall through
case KeyInfo::isStandard: {
if (actual.disabled) {
removeHead(actual);
processRegion(actual._dataStart, actual._dataStart+1);
nextKeyIdx = getNextKey();
} else {
- // Split on this key if not at start
+ // Split on this key if not at datastart of calling entry
int start = interval.nextNotIgnored(previousStart);
if (start < actual._tokenstart) {
interval.output(os, actual._tokenstart);
}
break;
}
- case KeyInfo::doRemove: {
- // Remove the key with all parameters
+ case KeyInfo::removeWithArg: {
+ nextKeyIdx = getNextKey();
+ // Search for end of arguments
+ int tmpIdx = find(nextKeyIdx, KeyInfo::endArguments);
+ if (tmpIdx > 0) {
+ for (int i = nextKeyIdx; i <= tmpIdx; i++) {
+ entries[i].disabled = true;
+ }
+ actual._dataEnd = entries[tmpIdx]._dataEnd;
+ }
interval.addIntervall(actual._tokenstart, actual._dataEnd+1);
+ break;
+ }
+ case KeyInfo::doRemove: {
+ // Remove the key with all parameters and following spaces
+ size_t pos;
+ for (pos = actual._dataEnd+1; pos < interval.par.length(); pos++) {
+ if ((interval.par[pos] != ' ') && (interval.par[pos] != '%'))
+ break;
+ }
+ interval.addIntervall(actual._tokenstart, pos);
nextKeyIdx = getNextKey();
break;
}
- case KeyInfo::isSectioning: {
+ case KeyInfo::isList: {
// Discard space before _tokenstart
int count;
for (count = 0; count < actual._tokenstart; count++) {
if (interval.par[actual._tokenstart-count-1] != ' ')
break;
}
+ nextKeyIdx = getNextKey();
+ int tmpIdx = find(nextKeyIdx, KeyInfo::endArguments);
+ if (tmpIdx > 0) {
+ // Special case: \item is not a list, but a command (like in Style Author_Biography in maa-monthly.layout)
+ // with arguments
+ // How else can we catch this one?
+ for (int i = nextKeyIdx; i <= tmpIdx; i++) {
+ entries[i].disabled = true;
+ }
+ actual._dataEnd = entries[tmpIdx]._dataEnd;
+ }
+ else if (nextKeyIdx > 0) {
+ // Ignore any lang entries inside data region
+ for (int i = nextKeyIdx; i < int(entries.size()) && entries[i]._tokenstart < actual._dataEnd; i++) {
+ if (entries[i].keytype == KeyInfo::isMain)
+ entries[i].disabled = true;
+ }
+ }
+ if (actual.disabled) {
+ interval.addIntervall(actual._tokenstart-count, actual._dataEnd+1);
+ }
+ else {
+ interval.addIntervall(actual._tokenstart-count, actual._tokenstart);
+ }
+ if (interval.par[actual._dataEnd+1] == '[') {
+ int posdown = interval.findclosing(actual._dataEnd+2, interval.par.length(), '[', ']');
+ if ((interval.par[actual._dataEnd+2] == '{') &&
+ (interval.par[posdown-1] == '}')) {
+ interval.addIntervall(actual._dataEnd+1,actual._dataEnd+3);
+ interval.addIntervall(posdown-1, posdown+1);
+ }
+ else {
+ interval.addIntervall(actual._dataEnd+1, actual._dataEnd+2);
+ interval.addIntervall(posdown, posdown+1);
+ }
+ int blk = interval.nextNotIgnored(actual._dataEnd+1);
+ if (blk > posdown) {
+ // Discard at most 1 space after empty item
+ int count;
+ for (count = 0; count < 1; count++) {
+ if (interval.par[blk+count] != ' ')
+ break;
+ }
+ if (count > 0)
+ interval.addIntervall(blk, blk+count);
+ }
+ }
+ break;
+ }
+ case KeyInfo::isSectioning: {
+ // Discard spaces before _tokenstart
+ int count;
+ int val = actual._tokenstart;
+ for (count = 0; count < actual._tokenstart;) {
+ val = interval.previousNotIgnored(val-1);
+ if (interval.par[val] != ' ')
+ break;
+ else {
+ count = actual._tokenstart - val;
+ }
+ }
if (actual.disabled) {
removeHead(actual, count);
nextKeyIdx = getNextKey();
break;
}
case KeyInfo::isMain: {
+ if (interval.par.substr(actual._dataStart, 2) == "% ")
+ interval.addIntervall(actual._dataStart, actual._dataStart+2);
if (actual.disabled) {
removeHead(actual);
- interval.resetOpenedP(actual._dataStart-1);
+ if ((interval.par.substr(actual._dataStart, 3) == " \\[") ||
+ (interval.par.substr(actual._dataStart, 8) == " \\begin{")) {
+ // Discard also the space before math-equation
+ interval.addIntervall(actual._dataStart, actual._dataStart+1);
+ }
+ nextKeyIdx = getNextKey();
+ // interval.resetOpenedP(actual._dataStart-1);
}
else {
if (actual._tokenstart == 0) {
nextKeyIdx = getNextKey();
break;
}
- }
- return(nextKeyIdx);
+ }
+ return nextKeyIdx;
}
int LatexInfo::process(ostringstream &os, KeyInfo &actual )
}
KeyInfo &nextKey = getKeyInfo(nextKeyIdx);
- if (nextKey.keytype == KeyInfo::isMain) {
+ if ((nextKey.keytype == KeyInfo::isMain) && !nextKey.disabled) {
(void) dispatch(os, actual._dataStart, nextKey);
end = nextKey._tokenstart;
break;
output_end = interval.nextNotIgnored(actual._dataEnd);
else
output_end = interval.nextNotIgnored(end);
- if (interval.nextNotIgnored(actual._dataStart) < output_end)
+ if ((actual.keytype == KeyInfo::isMain) && actual.disabled) {
+ interval.addIntervall(actual._tokenstart, actual._tokenstart+actual._tokensize);
+ }
+ // Remove possible empty data
+ int dstart = interval.nextNotIgnored(actual._dataStart);
+ while ((dstart < output_end) && (interval.par[dstart] == '{')) {
+ interval.addIntervall(dstart, dstart+1);
+ int dend = interval.findclosing(dstart+1, output_end);
+ interval.addIntervall(dend, dend+1);
+ dstart = interval.nextNotIgnored(dstart+1);
+ }
+ if (dstart < output_end)
interval.output(os, output_end);
interval.addIntervall(actual._tokenstart, end);
return nextKeyIdx;
}
-string splitOnKnownMacros(string par) {
+string splitOnKnownMacros(string par, bool isPatternString)
+{
ostringstream os;
- LatexInfo li(par);
+ LatexInfo li(par, isPatternString);
+ // LYXERR0("Berfore split: " << par);
KeyInfo DummyKey = KeyInfo(KeyInfo::KeyType::isMain, 2, true);
DummyKey.head = "";
DummyKey._tokensize = 0;
KeyInfo firstKey = li.getKeyInfo(firstkeyIdx);
int nextkeyIdx;
if ((firstKey.keytype != KeyInfo::isMain) || firstKey.disabled) {
- // Create dummy firstKey
+ // Use dummy firstKey
firstKey = DummyKey;
(void) li.setNextKey(firstkeyIdx);
}
+ else {
+ if (par.substr(firstKey._dataStart, 2) == "% ")
+ li.addIntervall(firstKey._dataStart, firstKey._dataStart+2);
+ }
nextkeyIdx = li.process(os, firstKey);
while (nextkeyIdx >= 0) {
// Check for a possible gap between the last
firstKey._dataStart = datastart;
firstKey._dataEnd = par.length();
(void) li.setNextKey(nextkeyIdx);
- if (firstKey._tokensize > 0)
+ if (firstKey._tokensize > 0) {
+ // Fake the last opened parenthesis
li.setForDefaultLang(firstKey._tokensize);
- // Fake the last opened parenthesis
+ }
nextkeyIdx = li.process(os, firstKey);
}
else {
// Handle the remaining
firstKey._dataStart = li.nextNotIgnored(firstKey._dataStart);
firstKey._dataEnd = par.length();
- if (firstKey._dataStart < firstKey._dataEnd)
+ // Check if ! empty
+ if ((firstKey._dataStart < firstKey._dataEnd) &&
+ (par[firstKey._dataStart] != '}')) {
+ if (firstKey._tokensize > 0)
+ li.setForDefaultLang(firstKey._tokensize);
(void) li.process(os, firstKey);
+ }
s = os.str();
+ if (s.empty()) {
+ // return string definitelly impossible to match
+ s = "\\foreignlanguage{ignore}{ }";
+ }
}
else
- s = ""; /* found end */
+ s = par; /* no known macros found */
+ // LYXERR0("After split: " << s);
return s;
}
* Resulting modified string is set to "", if
* the searched tex does not contain all the features in the search pattern
*/
-static string correctlanguagesetting(string par, bool from_regex, bool withformat)
+static string correctlanguagesetting(string par, bool isPatternString, bool withformat)
{
static Features regex_f;
static int missed = 0;
while ((parlen > 0) && (par[parlen-1] == '\n')) {
parlen--;
}
+ if (isPatternString && (parlen > 0) && (par[parlen-1] == '~')) {
+ // Happens to be there in case of description or labeling environment
+ parlen--;
+ }
string result;
if (withformat) {
// Split the latex input into pieces which
// can be digested by our search engine
LYXERR(Debug::FIND, "input: \"" << par << "\"");
- result = splitOnKnownMacros(par);
+ result = splitOnKnownMacros(par.substr(0,parlen), isPatternString);
LYXERR(Debug::FIND, "After split: \"" << result << "\"");
}
else
result = par.substr(0, parlen);
- if (from_regex) {
+ if (isPatternString) {
missed = 0;
if (withformat) {
regex_f = identifyFeatures(result);
while ( regex_replace(par_as_string, par_as_string, "(.*)\\\\}$", "$1")) {
open_braces++;
}
+ /*
// save '\.'
regex_replace(par_as_string, par_as_string, "\\\\\\.", "_xxbdotxx_");
// handle '.' -> '[^]', replace later as '[^\}\{\\]'
regex_replace(par_as_string, par_as_string, "_xxbrrxx_", "]");
// restore '\.'
regex_replace(par_as_string, par_as_string, "_xxbdotxx_", "\\.");
+ */
}
LYXERR(Debug::FIND, "par_as_string now is '" << par_as_string << "'");
LYXERR(Debug::FIND, "Open braces: " << open_braces);
while (regex_replace(par_as_string, par_as_string, orig, dest));
}
regexp_str = "(" + lead_as_regexp + ")" + par_as_string;
- regexp2_str = "(" + lead_as_regexp + ").*" + par_as_string;
+ regexp2_str = "(" + lead_as_regexp + ").*?" + par_as_string;
}
LYXERR(Debug::FIND, "Setting regexp to : '" << regexp_str << "'");
regexp = lyx::regex(regexp_str);
}
+// Count number of characters in string
+// {]} ==> 1
+// \& ==> 1
+// --- ==> 1
+// \\[a-zA-Z]+ ==> 1
+static int computeSize(string s, int len)
+{
+ if (len == 0)
+ return 0;
+ int skip = 1;
+ int count = 0;
+ for (int i = 0; i < len; i += skip, count++) {
+ if (s[i] == '\\') {
+ skip = 2;
+ if (isalpha(s[i+1])) {
+ for (int j = 2; i+j < len; j++) {
+ if (! isalpha(s[i+j])) {
+ if (s[i+j] == ' ')
+ skip++;
+ else if ((s[i+j] == '{') && s[i+j+1] == '}')
+ skip += 2;
+ else if ((s[i+j] == '{') && (i + j + 1 >= len))
+ skip++;
+ break;
+ }
+ skip++;
+ }
+ }
+ }
+ else if (s[i] == '{') {
+ if (s[i+1] == '}')
+ skip = 2;
+ else
+ skip = 3;
+ }
+ else if (s[i] == '-') {
+ if (s[i+1] == '-') {
+ if (s[i+2] == '-')
+ skip = 3;
+ else
+ skip = 2;
+ }
+ else
+ skip = 1;
+ }
+ else {
+ skip = 1;
+ }
+ }
+ return count;
+}
+
int MatchStringAdv::findAux(DocIterator const & cur, int len, bool at_begin) const
{
if (at_begin &&
else
result = m[m.size() - close_wildcards].first - m[0].first;
+ size_t pos = m.position(size_t(0));
+ // Ignore last closing characters
+ while (result > 0) {
+ if (str[pos+result-1] == '}')
+ --result;
+ else
+ break;
+ }
if (result > leadingsize)
result -= leadingsize;
else
result = 0;
- return(result);
+ return computeSize(str.substr(pos+leadingsize,result), result);
}
// else !use_regexp: but all code paths above return
<< ", inTexted=" << cur.inTexted());
if (res == 0 || !at_begin || !opt.matchword || !cur.inTexted())
return res;
+ if ((len > 0) && (res < len))
+ return 0;
Paragraph const & par = cur.paragraph();
bool ws_left = (cur.pos() > 0)
? par.isWordSeparator(cur.pos() - 1)
: true;
- bool ws_right = (cur.pos() + res < par.size())
- ? par.isWordSeparator(cur.pos() + res)
+ bool ws_right = (cur.pos() + len < par.size())
+ ? par.isWordSeparator(cur.pos() + len)
: true;
LYXERR(Debug::FIND,
"cur.pos()=" << cur.pos() << ", res=" << res
<< ", separ: " << ws_left << ", " << ws_right
+ << ", len: " << len
<< endl);
- if (ws_left && ws_right)
- return res;
+ if (ws_left && ws_right) {
+ // Check for word separators inside the found 'word'
+ for (int i = 0; i < len; i++) {
+ if (par.isWordSeparator(cur.pos() + i))
+ return 0;
+ }
+ return res;
+ }
return 0;
}
while (!t.empty() && t[t.size() - 1] == '\n')
t = t.substr(0, t.size() - 1);
size_t pos;
- // Replace all other \n with spaces
- while ((pos = t.find("\n")) != string::npos)
- t.replace(pos, 1, " ");
+ // Handle all other '\n'
+ while ((pos = t.find("\n")) != string::npos) {
+ if (pos > 1 && t[pos-1] == '\\' && t[pos-2] == '\\' ) {
+ // Handle '\\\n'
+ if (std::isalnum(t[pos+1])) {
+ t.replace(pos-2, 3, " ");
+ }
+ else {
+ t.replace(pos-2, 3, "");
+ }
+ }
+ else if (!std::isalnum(t[pos+1]) || !std::isalnum(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, " ");
+ }
+ }
// Remove stale empty \emph{}, \textbf{} and similar blocks from latexify
// Kornel: Added textsl, textsf, textit, texttt and noun
// + allow to seach for colored text too
while (regex_replace(t, t, "\\\\((sub)?(((sub)?section)|paragraph)|part)\\*?(\\{(\\{\\})?\\})+", ""))
LYXERR(Debug::FIND, " further removing stale empty \\emph{}, \\textbf{} macros from: " << t);
- while (regex_replace(t, t, "\\\\(foreignlanguage|textcolor)\\{[a-z]+\\}(\\{(\\\\item |\\{\\})?\\})+", ""));
+ while (regex_replace(t, t, "\\\\(foreignlanguage|textcolor|item)\\{[a-z]+\\}(\\{(\\{\\})?\\})+", ""));
// FIXME - check what preceeds the brace
if (hack_braces) {
if (opt.ignoreformat)
AS_STR_INSETS | AS_STR_SKIPDELETE | AS_STR_PLAINTEXT,
&runparams);
} else if (cur.inMathed()) {
- docstring s;
CursorSlice cs = cur.top();
MathData md = cs.cell();
MathData::const_iterator it_end =
(( len == -1 || cs.pos() + len > int(md.size()))
? md.end()
: md.begin() + cs.pos() + len );
+ MathData md2;
for (MathData::const_iterator it = md.begin() + cs.pos();
it != it_end; ++it)
- s = s + asString(*it);
+ md2.push_back(*it);
+ docstring s = asString(md2);
LYXERR(Debug::FIND, "Stringified math: '" << s << "'");
return s;
}
((len == -1 || cs.pos() + len > int(md.size()))
? md.end()
: md.begin() + cs.pos() + len);
+ MathData md2;
for (MathData::const_iterator it = md.begin() + cs.pos();
it != it_end; ++it)
- ods << asString(*it);
+ md2.push_back(*it);
+ ods << asString(md2);
// Retrieve the math environment type, and add '$' or '$]'
// or others (\end{equation}) accordingly
for (int s = cur.depth() - 1; s >= 0; --s) {
cur.forwardPos();
} while (cur && cur.depth() > d && match(cur) > 0);
cur = old_cur;
- if (match(cur) <= 0) return 0;
+ int max_match = match(cur); /* match valid only if not searching whole words */
+ if (max_match <= 0) return 0;
LYXERR(Debug::FIND, "Ok");
// Compute the match length
- int len = 1;
+ int len = 1;
if (cur.pos() + len > cur.lastpos())
- return 0;
- LYXERR(Debug::FIND, "verifying unmatch with len = " << len);
- while (cur.pos() + len <= cur.lastpos() && match(cur, len) <= 0) {
- ++len;
- LYXERR(Debug::FIND, "verifying unmatch with len = " << len);
- }
- // Length of matched text (different from len param)
- int old_len = match(cur, len);
- if (old_len < 0) old_len = 0;
- int new_len;
- // Greedy behaviour while matching regexps
- while ((new_len = match(cur, len + 1)) > old_len) {
- ++len;
- old_len = new_len;
- LYXERR(Debug::FIND, "verifying match with len = " << len);
+ return 0;
+ if (match.opt.matchword) {
+ LYXERR(Debug::FIND, "verifying unmatch with len = " << len);
+ while (cur.pos() + len <= cur.lastpos() && match(cur, len) <= 0) {
+ ++len;
+ LYXERR(Debug::FIND, "verifying unmatch with len = " << len);
+ }
+ // Length of matched text (different from len param)
+ int old_match = match(cur, len);
+ if (old_match < 0)
+ old_match = 0;
+ int new_match;
+ // Greedy behaviour while matching regexps
+ while ((new_match = match(cur, len + 1)) > old_match) {
+ ++len;
+ old_match = new_match;
+ LYXERR(Debug::FIND, "verifying match with len = " << len);
+ }
+ if (old_match == 0)
+ len = 0;
+ }
+ else {
+ int minl = 1;
+ int maxl = cur.lastpos() - cur.pos();
+ // Greedy behaviour while matching regexps
+ while (maxl > minl) {
+ int actual_match = match(cur, len);
+ if (actual_match >= max_match) {
+ // actual_match > max_match _can_ happen,
+ // if the search area splits
+ // some following word so that the regex
+ // (e.g. 'r.*r\b' matches 'r' from the middle of the
+ // splitted word)
+ // This means, the len value is too big
+ maxl = len;
+ len = (int)((maxl + minl)/2);
+ }
+ else {
+ // (actual_match < max_match)
+ minl = len + 1;
+ len = (int)((maxl + minl)/2);
+ }
+ }
+ old_cur = cur;
+ // Search for real start of matched characters
+ while (len > 1) {
+ int actual_match;
+ do {
+ cur.forwardPos();
+ } while (cur.depth() > old_cur.depth()); /* Skip inner insets */
+ if (cur.depth() < old_cur.depth()) {
+ // Outer inset?
+ LYXERR0("cur.depth() < old_cur.depth(), this should never happen");
+ break;
+ }
+ if (cur.pos() != old_cur.pos()) {
+ // OK, forwarded 1 pos in actual inset
+ actual_match = match(cur, len-1);
+ if (actual_match == max_match) {
+ // Ha, got it! The shorter selection has the same match length
+ len--;
+ old_cur = cur;
+ }
+ else {
+ // OK, the shorter selection matches less chars, revert to previous value
+ cur = old_cur;
+ break;
+ }
+ }
+ else {
+ LYXERR0("cur.pos() == old_cur.pos(), this should never happen");
+ actual_match = match(cur, len);
+ if (actual_match == max_match)
+ old_cur = cur;
+ }
+ }
}
return len;
}
int match_len_zero_count = 0;
for (; !theApp()->longOperationCancelled() && cur; cur.forwardPos()) {
LYXERR(Debug::FIND, "Advancing cur: " << cur);
+ int match_len3 = match(cur, 1);
+ if (match_len3 < 0)
+ continue;
int match_len2 = match(cur);
LYXERR(Debug::FIND, "match_len2: " << match_len2);
if (match_len2 > 0) {