// @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|md|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\\{[a-z]+\\}", "")
static Features identifyFeatures(string const & s)
{
static regex const feature("\\\\(([a-z]+(\\{([a-z]+)\\}|\\*)?))\\{");
- static regex const valid("^(((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)\\*?)$");
+ static regex const valid("^(((footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge|emph|noun|text(bf|md|sl|sf|it|tt)|(textcolor|foreignlanguage|item)\\{[a-z]+\\})|(u|uu)line|(s|x)out|uwave)|((sub)?(((sub)?section)|paragraph)|part|chapter)\\*?)$");
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 */
- noMain, /* to discard language in content */
+ /* \foreignlanguage{ngerman}, ... */
+ isMain,
+ /* inside \code{} or \footnote{}
+ * to discard language in content */
+ noMain,
isRegex,
+ /* \begin{eqnarray}...\end{eqnarray}, ... $...$ */
isMath,
+ /* fonts, colors, markups, ... */
isStandard,
- noContent, /* discard content */
+ /* footnotesize, ... large, ...
+ * Ignore all of them */
isSize,
invalid,
+ /* inputencoding, shortcut, ...
+ * Discard also content, because they do not help in search */
doRemove,
+ /* item */
isList,
- isIgnored /* to be ignored by creating infos */
+ /* tex, latex, ... like isChar */
+ isIgnored
};
KeyInfo()
: keytype(invalid),
head(""),
+ _tokensize(-1),
+ _tokenstart(-1),
+ _dataStart(-1),
+ _dataEnd(-1),
parenthesiscount(1),
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),
used(false) {};
class Intervall {
bool isPatternString;
public:
- Intervall(bool isPattern) : isPatternString(isPattern), ignoreidx(-1), actualdeptindex(0) { depts[0] = 0;};
+ explicit Intervall(bool isPattern) :
+ isPatternString(isPattern),
+ ignoreidx(-1),
+ actualdeptindex(0) { depts[0] = 0; closes[0] = 0;};
string par;
int ignoreidx;
int depts[MAXOPENED];
void removeHead(KeyInfo&, int count=0);
public:
- LatexInfo(string par, bool isPatternString) : interval(isPatternString) {
+ LatexInfo(string par, bool isPatternString) : entidx(-1), interval(isPatternString) {
interval.par = par;
buildKeys(isPatternString);
entries = vector<KeyInfo>();
m.mathSize = end - start;
entries.push_back(m);
}
- bool evaluating(size_t pos) {
- while (actualIdx < entries.size()) {
- if (pos < entries[actualIdx].mathStart)
- return false;
- if (pos < entries[actualIdx].mathEnd)
- return true;
- actualIdx++;
- }
- return false;
- }
bool empty() { return entries.empty(); };
size_t getEndPos() {
if (entries.empty() || (actualIdx >= entries.size())) {
found._tokenstart = sub.position(size_t(0));
if (found.parenthesiscount == 0) {
// Probably to be discarded
- char following = interval.par[sub.position(size_t(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 == '=') {
}
else
found._dataEnd = endpos;
- if (isPatternString) {
- keys[key].used = true;
- }
+ }
+ if (isPatternString) {
+ keys[key].used = true;
}
}
entries.push_back(found);
makeKey("triangleuppar|triangledownpar|droppar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
makeKey("triangleleftpar|shapepar|dropuppar", KeyInfo(KeyInfo::isStandard, 1, true), isPatternString);
// like ('tiny{}' or '\tiny ' ... )
- makeKey("footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge", KeyInfo(KeyInfo::isSize, 0, true), isPatternString);
+ makeKey("footnotesize|tiny|scriptsize|small|large|Large|LARGE|huge|Huge", KeyInfo(KeyInfo::isSize, 0, false), isPatternString);
// Survives, like known character
makeKey("lyx|latex|latexe|tex", KeyInfo(KeyInfo::isIgnored, 0, false), isPatternString);
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
int nextKeyIdx = 0;
switch (actual.keytype)
{
+ case KeyInfo::noContent: { /* char like "\hspace{2cm}" */
+ 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::noContent: {
- interval.addIntervall(actual._dataStart, actual._dataEnd);
- }
- // fall through
case KeyInfo::noMain:
// fall through
case KeyInfo::isStandard: {
if ((actual.keytype == KeyInfo::isMain) && actual.disabled) {
interval.addIntervall(actual._tokenstart, actual._tokenstart+actual._tokensize);
}
- if (interval.nextNotIgnored(actual._dataStart) < output_end)
+ // 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;