#include "ParIterator.h"
#include "TexRow.h"
#include "Text.h"
+#include "Encoding.h"
#include "frontends/Application.h"
#include "frontends/alert.h"
#include "support/lstrings.h"
#include "support/regex.h"
+#include "support/textutils.h"
#include <map>
using namespace std;
static docstring buffer_to_latex(Buffer & buffer)
{
- OutputParams runparams(&buffer.params().encoding());
+ //OutputParams runparams(&buffer.params().encoding());
+ OutputParams runparams(encodings.fromLyXName("utf8"));
odocstringstream ods;
otexstream os(ods);
runparams.nice = true;
- runparams.flavor = OutputParams::LATEX;
+ runparams.flavor = OutputParams::XETEX;
runparams.linelen = 10000; //lyxrc.plaintext_linelen;
// No side effect of file copying and image conversion
runparams.dryrun = true;
if (!opt.ignoreformat) {
str = buffer_to_latex(buffer);
} else {
- OutputParams runparams(&buffer.params().encoding());
+ // OutputParams runparams(&buffer.params().encoding());
+ OutputParams runparams(encodings.fromLyXName("utf8"));
runparams.nice = true;
- runparams.flavor = OutputParams::LATEX;
+ runparams.flavor = OutputParams::XETEX;
runparams.linelen = 10000; //lyxrc.plaintext_linelen;
runparams.dryrun = true;
runparams.for_search = true;
void resetOpenedP(int openPos);
void addIntervall(int upper);
void addIntervall(int low, int upper); /* if explicit */
+ void removeAccents();
void setForDefaultLang(KeyInfo &defLang);
int findclosing(int start, int end, char up, char down, int repeat);
void handleParentheses(int lastpos, bool closingAllowed);
}
}
+typedef map<string, string> AccentsMap;
+static AccentsMap accents = map<string, string>();
+
+static void buildaccent(string n, string param, string values)
+{
+ stringstream s(n);
+ string name;
+ const char delim = '|';
+ while (getline(s, name, delim)) {
+ size_t start = 0;
+ for (size_t i = 0; i < param.size(); i++) {
+ string key = name + "{" + param[i] + "}";
+ // get the corresponding utf8-value
+ if ((values[start] & 0xc0) != 0xc0) {
+ // should not happen, utf8 encoding starts at least with 11xxxxxx
+ // but value for '\dot{i}' is 'i', which is ascii
+ if ((values[start] & 0x80) == 0) {
+ // is ascii
+ accents[key] = values.substr(start, 1);
+ }
+ start++;
+ continue;
+ }
+ for (int j = 1; ;j++) {
+ if (start + j >= values.size()) {
+ accents[key] = values.substr(start, j);
+ start = values.size() - 1;
+ break;
+ }
+ else if ((values[start+j] & 0xc0) != 0x80) {
+ // This is the first byte of following utf8 char
+ accents[key] = values.substr(start, j);
+ start += j;
+ break;
+ }
+ }
+ }
+ }
+}
+
+static void buildAccentsMap()
+{
+ accents["imath"] = "ı";
+ accents["i"] = "ı";
+ accents["jmath"] = "ȷ";
+ accents["lyxmathsym{ß}"] = "ß";
+ accents["text{ß}"] = "ß";
+ accents["ddot{\\imath}"] = "ï";
+ buildaccent("ddot", "aAeEiIioOuUyY",
+ "äÄëËïÏïöÖüÜÿŸ"); // umlaut
+ buildaccent("dot|.", "cCeEGgIizZaAoObBdDfFyY",
+ "ċĊėĖĠġİİżŻȧȦȯȮḃḂḋḊḟḞẏẎ"); // dot{i} can only happen if ignoring case, but there is no lowercase of 'İ'
+ accents["acute{\\imath}"] = "í";
+ buildaccent("acute", "aAcCeElLoOnNrRsSuUyYzZiI",
+ "áÁćĆéÉĺĹóÓńŃŕŔśŚúÚýÝźŹíÍ");
+ buildaccent("dacute|H|h", "oOuU", "őŐűŰ"); // double acute
+ buildaccent("mathring|r", "aAuUwy",
+ "åÅůŮẘẙ"); // ring
+ accents["check{\\imath}"] = "ǐ";
+ accents["check{\\jmath}"] = "ǰ";
+ buildaccent("check|v", "cCdDaAeEiIoOuUgGkKhHlLnNrRsSTtzZ",
+ "čČďĎǎǍěĚǐǏǒǑǔǓǧǦǩǨȟȞľĽňŇřŘšŠŤťžŽ"); // caron
+ accents["hat{\\imath}"] = "î";
+ accents["hat{\\jmath}"] = "ĵ";
+ buildaccent("hat|^", "aAeEiIcCgGhHjJsSwWyYzZoOuU",
+ "âÂêÊîÎĉĈĝĜĥĤĵĴŝŜŵŴŷŶẑẐôÔûÛ"); // circ
+ accents["bar{\\imath}"] = "ī";
+ buildaccent("bar|=", "aAeEiIoOuUyY",
+ "āĀēĒīĪōŌūŪȳȲ"); // macron
+ accents["tilde{\\imath}"] = "ĩ";
+ buildaccent("tilde", "aAnNoOiIuU",
+ "ãÃñÑõÕĩĨũŨ"); // tilde
+ accents["breve{\\imath}"] = "ĭ";
+ buildaccent("breve|u", "aAeEgGiIoOuU",
+ "ăĂĕĔğĞĭĬŏŎŭŬ"); // breve
+ accents["grave{\\imath}"] = "ì";
+ buildaccent("grave|`", "aAeEiIoOuUnNwWyY",
+ "àÀèÈìÌòÒùÙǹǸẁẀỳỲ"); // grave
+ buildaccent("subdot|d", "BbDdHhKkLlMmNnRrSsTtVvWwZzAaEeIiOoUuYy",
+ "ḄḅḌḍḤḥḲḳḶḷṂṃṆṇṚṛṢṣṬṭṾṿẈẉẒẓẠạẸẹỊịỌọỤụỴỵ"); // dot below
+}
+
+/*
+ * Created accents in math or regexp environment
+ * are macros, but we need the utf8 equivalent
+ */
+void Intervall::removeAccents()
+{
+ if (accents.empty())
+ buildAccentsMap();
+ static regex const accre("\\\\((.|grave|breve|lyxmathsym|text|ddot|dot|acute|dacute|mathring|check|hat|bar|tilde|subdot)\\{[^\\{\\}]+\\}|(i|imath|jmath)(?![a-zA-Z]))");
+ smatch sub;
+ for (sregex_iterator itacc(par.begin(), par.end(), accre), end; itacc != end; ++itacc) {
+ sub = *itacc;
+ string key = sub.str(1);
+ if (accents.find(key) != accents.end()) {
+ string val = accents[key];
+ size_t pos = sub.position(0);
+ for (size_t i = 0; i < val.size(); i++) {
+ par[pos+i] = val[i];
+ }
+ addIntervall(pos+val.size(), pos + sub.str(0).size());
+ for (size_t i = pos+val.size(); i < pos + sub.str(0).size(); i++) {
+ // remove traces of any remaining chars
+ par[i] = ' ';
+ }
+ }
+ else {
+ LYXERR0("Not added accent for \"" << key << "\"");
+ }
+ }
+}
+
void Intervall::handleOpenP(int i)
{
actualdeptindex++;
size_t math_pos = 10000;
string math_end;
+ interval.removeAccents();
+
for (sregex_iterator itmath(interval.par.begin(), interval.par.end(), rmath), end; itmath != end; ++itmath) {
submath = *itmath;
if (math_end_waiting) {
found.parenthesiscount = 0;
found.head = interval.par.substr(found._tokenstart, found._tokensize);
}
- else
+ else {
continue;
+ }
}
else {
if (evaluatingMath) {
}
case KeyInfo::endArguments:
// Remove trailing '{}' too
- actual._dataStart += 2;
- actual._dataEnd += 2;
- interval.addIntervall(actual._tokenstart, actual._dataEnd);
+ actual._dataStart += 1;
+ actual._dataEnd += 1;
+ interval.addIntervall(actual._tokenstart, actual._dataEnd+1);
nextKeyIdx = getNextKey();
break;
case KeyInfo::noMain:
missed = 0;
if (withformat) {
regex_f = identifyFeatures(result);
- string features = "";
+ string features = "";
for (auto it = regex_f.cbegin(); it != regex_f.cend(); ++it) {
string a = it->first;
regex_with_format = true;
- features += " " + a;
+ features += " " + a;
// LYXERR0("Identified regex format:" << a);
}
- LYXERR(Debug::FIND, "Identified Features" << features);
+ LYXERR(Debug::FIND, "Identified Features" << features);
}
} else if (regex_with_format) {
while ((pos = t.find("\n")) != string::npos) {
if (pos > 1 && t[pos-1] == '\\' && t[pos-2] == '\\' ) {
// Handle '\\\n'
- if (std::isalnum(t[pos+1])) {
+ if (isAlnumASCII(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])) {
+ else if (!isAlnumASCII(t[pos+1]) || !isAlnumASCII(t[pos-1])) {
// '\n' adjacent to non-alpha-numerics, discard
t.replace(pos, 1, "");
}
// TODO Try adding a AS_STR_INSERTS as last arg
pos_type end = ( len == -1 || cur.pos() + len > int(par.size()) ) ?
int(par.size()) : cur.pos() + len;
- OutputParams runparams(&cur.buffer()->params().encoding());
+ // OutputParams runparams(&cur.buffer()->params().encoding());
+ OutputParams runparams(encodings.fromLyXName("utf8"));
runparams.nice = true;
- runparams.flavor = OutputParams::LATEX;
+ runparams.flavor = OutputParams::XETEX;
runparams.linelen = 10000; //lyxrc.plaintext_linelen;
// No side effect of file copying and image conversion
runparams.dryrun = true;
odocstringstream ods;
otexstream os(ods);
- OutputParams runparams(&buf.params().encoding());
+ //OutputParams runparams(&buf.params().encoding());
+ OutputParams runparams(encodings.fromLyXName("utf8"));
runparams.nice = false;
- runparams.flavor = OutputParams::LATEX;
+ runparams.flavor = OutputParams::XETEX;
runparams.linelen = 8000; //lyxrc.plaintext_linelen;
// No side effect of file copying and image conversion
runparams.dryrun = true;
DocIterator old_cur = cur;
for (int i = 0; i < increment && cur; cur.forwardPos(), i++) {
}
- if (! cur) {
+ if (! cur || (cur.pit() > old_cur.pit())) {
+ // Are we outside of the paragraph?
+ // This can happen if moving past some UTF8-encoded chars
cur = old_cur;
increment /= 2;
}
} else if (cur.inMathed()) {
odocstringstream ods;
otexstream os(ods);
- OutputParams runparams(&repl_buffer.params().encoding());
+ // OutputParams runparams(&repl_buffer.params().encoding());
+ OutputParams runparams(encodings.fromLyXName("utf8"));
runparams.nice = false;
- runparams.flavor = OutputParams::LATEX;
+ runparams.flavor = OutputParams::XETEX;
runparams.linelen = 8000; //lyxrc.plaintext_linelen;
runparams.dryrun = true;
TeXOnePar(repl_buffer, repl_buffer.text(), 0, os, runparams);