92bad0a60a
Fix #5822
547 lines
16 KiB
C++
547 lines
16 KiB
C++
// Scintilla source code edit control
|
|
/** @file LexLaTeX.cxx
|
|
** Lexer for LaTeX2e.
|
|
**/
|
|
// Copyright 1998-2001 by Neil Hodgson <neilh@scintilla.org>
|
|
// The License.txt file describes the conditions under which this software may be distributed.
|
|
|
|
// Modified by G. HU in 2013. Added folding, syntax highting inside math environments, and changed some minor behaviors.
|
|
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
#include <stdio.h>
|
|
#include <stdarg.h>
|
|
#include <assert.h>
|
|
#include <ctype.h>
|
|
#include <vector>
|
|
|
|
#include "ILexer.h"
|
|
#include "Scintilla.h"
|
|
#include "SciLexer.h"
|
|
|
|
#include "PropSetSimple.h"
|
|
#include "WordList.h"
|
|
#include "LexAccessor.h"
|
|
#include "Accessor.h"
|
|
#include "StyleContext.h"
|
|
#include "CharacterSet.h"
|
|
#include "LexerModule.h"
|
|
#include "DefaultLexer.h"
|
|
#include "LexerBase.h"
|
|
|
|
using namespace Scintilla;
|
|
|
|
using namespace std;
|
|
|
|
struct latexFoldSave {
|
|
latexFoldSave() : structLev(0) {
|
|
for (int i = 0; i < 8; ++i) openBegins[i] = 0;
|
|
}
|
|
latexFoldSave(const latexFoldSave &save) : structLev(save.structLev) {
|
|
for (int i = 0; i < 8; ++i) openBegins[i] = save.openBegins[i];
|
|
}
|
|
latexFoldSave &operator=(const latexFoldSave &save) {
|
|
if (this != &save) {
|
|
structLev = save.structLev;
|
|
for (int i = 0; i < 8; ++i) openBegins[i] = save.openBegins[i];
|
|
}
|
|
return *this;
|
|
}
|
|
int openBegins[8];
|
|
Sci_Position structLev;
|
|
};
|
|
|
|
class LexerLaTeX : public LexerBase {
|
|
private:
|
|
vector<int> modes;
|
|
void setMode(Sci_Position line, int mode) {
|
|
if (line >= static_cast<Sci_Position>(modes.size())) modes.resize(line + 1, 0);
|
|
modes[line] = mode;
|
|
}
|
|
int getMode(Sci_Position line) {
|
|
if (line >= 0 && line < static_cast<Sci_Position>(modes.size())) return modes[line];
|
|
return 0;
|
|
}
|
|
void truncModes(Sci_Position numLines) {
|
|
if (static_cast<Sci_Position>(modes.size()) > numLines * 2 + 256)
|
|
modes.resize(numLines + 128);
|
|
}
|
|
|
|
vector<latexFoldSave> saves;
|
|
void setSave(Sci_Position line, const latexFoldSave &save) {
|
|
if (line >= static_cast<Sci_Position>(saves.size())) saves.resize(line + 1);
|
|
saves[line] = save;
|
|
}
|
|
void getSave(Sci_Position line, latexFoldSave &save) {
|
|
if (line >= 0 && line < static_cast<Sci_Position>(saves.size())) save = saves[line];
|
|
else {
|
|
save.structLev = 0;
|
|
for (int i = 0; i < 8; ++i) save.openBegins[i] = 0;
|
|
}
|
|
}
|
|
void truncSaves(Sci_Position numLines) {
|
|
if (static_cast<Sci_Position>(saves.size()) > numLines * 2 + 256)
|
|
saves.resize(numLines + 128);
|
|
}
|
|
public:
|
|
static ILexer4 *LexerFactoryLaTeX() {
|
|
return new LexerLaTeX();
|
|
}
|
|
void SCI_METHOD Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override;
|
|
void SCI_METHOD Fold(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override;
|
|
};
|
|
|
|
static bool latexIsSpecial(int ch) {
|
|
return (ch == '#') || (ch == '$') || (ch == '%') || (ch == '&') || (ch == '_') ||
|
|
(ch == '{') || (ch == '}') || (ch == ' ');
|
|
}
|
|
|
|
static bool latexIsBlank(int ch) {
|
|
return (ch == ' ') || (ch == '\t');
|
|
}
|
|
|
|
static bool latexIsBlankAndNL(int ch) {
|
|
return (ch == ' ') || (ch == '\t') || (ch == '\r') || (ch == '\n');
|
|
}
|
|
|
|
static bool latexIsLetter(int ch) {
|
|
return IsASCII(ch) && isalpha(ch);
|
|
}
|
|
|
|
static bool latexIsTagValid(Sci_Position &i, Sci_Position l, Accessor &styler) {
|
|
while (i < l) {
|
|
if (styler.SafeGetCharAt(i) == '{') {
|
|
while (i < l) {
|
|
i++;
|
|
if (styler.SafeGetCharAt(i) == '}') {
|
|
return true;
|
|
} else if (!latexIsLetter(styler.SafeGetCharAt(i)) &&
|
|
styler.SafeGetCharAt(i)!='*') {
|
|
return false;
|
|
}
|
|
}
|
|
} else if (!latexIsBlank(styler.SafeGetCharAt(i))) {
|
|
return false;
|
|
}
|
|
i++;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
static bool latexNextNotBlankIs(Sci_Position i, Accessor &styler, char needle) {
|
|
char ch;
|
|
while (i < styler.Length()) {
|
|
ch = styler.SafeGetCharAt(i);
|
|
if (!latexIsBlankAndNL(ch) && ch != '*') {
|
|
if (ch == needle)
|
|
return true;
|
|
else
|
|
return false;
|
|
}
|
|
i++;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
static bool latexLastWordIs(Sci_Position start, Accessor &styler, const char *needle) {
|
|
Sci_PositionU i = 0;
|
|
Sci_PositionU l = static_cast<Sci_PositionU>(strlen(needle));
|
|
Sci_Position ini = start-l+1;
|
|
char s[32];
|
|
|
|
while (i < l && i < 31) {
|
|
s[i] = styler.SafeGetCharAt(ini + i);
|
|
i++;
|
|
}
|
|
s[i] = '\0';
|
|
|
|
return (strcmp(s, needle) == 0);
|
|
}
|
|
|
|
static bool latexLastWordIsMathEnv(Sci_Position pos, Accessor &styler) {
|
|
Sci_Position i, j;
|
|
char s[32];
|
|
const char *mathEnvs[] = { "align", "alignat", "flalign", "gather",
|
|
"multiline", "displaymath", "eqnarray", "equation" };
|
|
if (styler.SafeGetCharAt(pos) != '}') return false;
|
|
for (i = pos - 1; i >= 0; --i) {
|
|
if (styler.SafeGetCharAt(i) == '{') break;
|
|
if (pos - i >= 20) return false;
|
|
}
|
|
if (i < 0 || i == pos - 1) return false;
|
|
++i;
|
|
for (j = 0; i + j < pos; ++j)
|
|
s[j] = styler.SafeGetCharAt(i + j);
|
|
s[j] = '\0';
|
|
if (j == 0) return false;
|
|
if (s[j - 1] == '*') s[--j] = '\0';
|
|
for (i = 0; i < static_cast<int>(sizeof(mathEnvs) / sizeof(const char *)); ++i)
|
|
if (strcmp(s, mathEnvs[i]) == 0) return true;
|
|
return false;
|
|
}
|
|
|
|
static inline void latexStateReset(int &mode, int &state) {
|
|
switch (mode) {
|
|
case 1: state = SCE_L_MATH; break;
|
|
case 2: state = SCE_L_MATH2; break;
|
|
default: state = SCE_L_DEFAULT; break;
|
|
}
|
|
}
|
|
|
|
// There are cases not handled correctly, like $abcd\textrm{what is $x+y$}z+w$.
|
|
// But I think it's already good enough.
|
|
void SCI_METHOD LexerLaTeX::Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) {
|
|
// startPos is assumed to be the first character of a line
|
|
Accessor styler(pAccess, &props);
|
|
styler.StartAt(startPos);
|
|
int mode = getMode(styler.GetLine(startPos) - 1);
|
|
int state = initStyle;
|
|
if (state == SCE_L_ERROR || state == SCE_L_SHORTCMD || state == SCE_L_SPECIAL) // should not happen
|
|
latexStateReset(mode, state);
|
|
|
|
char chNext = styler.SafeGetCharAt(startPos);
|
|
char chVerbatimDelim = '\0';
|
|
styler.StartSegment(startPos);
|
|
Sci_Position lengthDoc = startPos + length;
|
|
|
|
for (Sci_Position i = startPos; i < lengthDoc; i++) {
|
|
char ch = chNext;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
|
|
if (styler.IsLeadByte(ch)) {
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
continue;
|
|
}
|
|
|
|
if (ch == '\r' || ch == '\n')
|
|
setMode(styler.GetLine(i), mode);
|
|
|
|
switch (state) {
|
|
case SCE_L_DEFAULT :
|
|
switch (ch) {
|
|
case '\\' :
|
|
styler.ColourTo(i - 1, state);
|
|
if (latexIsLetter(chNext)) {
|
|
state = SCE_L_COMMAND;
|
|
} else if (latexIsSpecial(chNext)) {
|
|
styler.ColourTo(i + 1, SCE_L_SPECIAL);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
} else if (chNext == '\r' || chNext == '\n') {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
} else if (IsASCII(chNext)) {
|
|
styler.ColourTo(i + 1, SCE_L_SHORTCMD);
|
|
if (chNext == '(') {
|
|
mode = 1;
|
|
state = SCE_L_MATH;
|
|
} else if (chNext == '[') {
|
|
mode = 2;
|
|
state = SCE_L_MATH2;
|
|
}
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
}
|
|
break;
|
|
case '$' :
|
|
styler.ColourTo(i - 1, state);
|
|
if (chNext == '$') {
|
|
styler.ColourTo(i + 1, SCE_L_SHORTCMD);
|
|
mode = 2;
|
|
state = SCE_L_MATH2;
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
} else {
|
|
styler.ColourTo(i, SCE_L_SHORTCMD);
|
|
mode = 1;
|
|
state = SCE_L_MATH;
|
|
}
|
|
break;
|
|
case '%' :
|
|
styler.ColourTo(i - 1, state);
|
|
state = SCE_L_COMMENT;
|
|
break;
|
|
}
|
|
break;
|
|
// These 3 will never be reached.
|
|
case SCE_L_ERROR:
|
|
case SCE_L_SPECIAL:
|
|
case SCE_L_SHORTCMD:
|
|
break;
|
|
case SCE_L_COMMAND :
|
|
if (!latexIsLetter(chNext)) {
|
|
styler.ColourTo(i, state);
|
|
if (latexNextNotBlankIs(i + 1, styler, '[' )) {
|
|
state = SCE_L_CMDOPT;
|
|
} else if (latexLastWordIs(i, styler, "\\begin")) {
|
|
state = SCE_L_TAG;
|
|
} else if (latexLastWordIs(i, styler, "\\end")) {
|
|
state = SCE_L_TAG2;
|
|
} else if (latexLastWordIs(i, styler, "\\verb") && chNext != '*' && chNext != ' ') {
|
|
chVerbatimDelim = chNext;
|
|
state = SCE_L_VERBATIM;
|
|
} else {
|
|
latexStateReset(mode, state);
|
|
}
|
|
}
|
|
break;
|
|
case SCE_L_CMDOPT :
|
|
if (ch == ']') {
|
|
styler.ColourTo(i, state);
|
|
latexStateReset(mode, state);
|
|
}
|
|
break;
|
|
case SCE_L_TAG :
|
|
if (latexIsTagValid(i, lengthDoc, styler)) {
|
|
styler.ColourTo(i, state);
|
|
latexStateReset(mode, state);
|
|
if (latexLastWordIs(i, styler, "{verbatim}")) {
|
|
state = SCE_L_VERBATIM;
|
|
} else if (latexLastWordIs(i, styler, "{comment}")) {
|
|
state = SCE_L_COMMENT2;
|
|
} else if (latexLastWordIs(i, styler, "{math}") && mode == 0) {
|
|
mode = 1;
|
|
state = SCE_L_MATH;
|
|
} else if (latexLastWordIsMathEnv(i, styler) && mode == 0) {
|
|
mode = 2;
|
|
state = SCE_L_MATH2;
|
|
}
|
|
} else {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
latexStateReset(mode, state);
|
|
ch = styler.SafeGetCharAt(i);
|
|
if (ch == '\r' || ch == '\n') setMode(styler.GetLine(i), mode);
|
|
}
|
|
chNext = styler.SafeGetCharAt(i+1);
|
|
break;
|
|
case SCE_L_TAG2 :
|
|
if (latexIsTagValid(i, lengthDoc, styler)) {
|
|
styler.ColourTo(i, state);
|
|
latexStateReset(mode, state);
|
|
} else {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
latexStateReset(mode, state);
|
|
ch = styler.SafeGetCharAt(i);
|
|
if (ch == '\r' || ch == '\n') setMode(styler.GetLine(i), mode);
|
|
}
|
|
chNext = styler.SafeGetCharAt(i+1);
|
|
break;
|
|
case SCE_L_MATH :
|
|
switch (ch) {
|
|
case '\\' :
|
|
styler.ColourTo(i - 1, state);
|
|
if (latexIsLetter(chNext)) {
|
|
Sci_Position match = i + 3;
|
|
if (latexLastWordIs(match, styler, "\\end")) {
|
|
match++;
|
|
if (latexIsTagValid(match, lengthDoc, styler)) {
|
|
if (latexLastWordIs(match, styler, "{math}"))
|
|
mode = 0;
|
|
}
|
|
}
|
|
state = SCE_L_COMMAND;
|
|
} else if (latexIsSpecial(chNext)) {
|
|
styler.ColourTo(i + 1, SCE_L_SPECIAL);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
} else if (chNext == '\r' || chNext == '\n') {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
} else if (IsASCII(chNext)) {
|
|
if (chNext == ')') {
|
|
mode = 0;
|
|
state = SCE_L_DEFAULT;
|
|
}
|
|
styler.ColourTo(i + 1, SCE_L_SHORTCMD);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
}
|
|
break;
|
|
case '$' :
|
|
styler.ColourTo(i - 1, state);
|
|
styler.ColourTo(i, SCE_L_SHORTCMD);
|
|
mode = 0;
|
|
state = SCE_L_DEFAULT;
|
|
break;
|
|
case '%' :
|
|
styler.ColourTo(i - 1, state);
|
|
state = SCE_L_COMMENT;
|
|
break;
|
|
}
|
|
break;
|
|
case SCE_L_MATH2 :
|
|
switch (ch) {
|
|
case '\\' :
|
|
styler.ColourTo(i - 1, state);
|
|
if (latexIsLetter(chNext)) {
|
|
Sci_Position match = i + 3;
|
|
if (latexLastWordIs(match, styler, "\\end")) {
|
|
match++;
|
|
if (latexIsTagValid(match, lengthDoc, styler)) {
|
|
if (latexLastWordIsMathEnv(match, styler))
|
|
mode = 0;
|
|
}
|
|
}
|
|
state = SCE_L_COMMAND;
|
|
} else if (latexIsSpecial(chNext)) {
|
|
styler.ColourTo(i + 1, SCE_L_SPECIAL);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
} else if (chNext == '\r' || chNext == '\n') {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
} else if (IsASCII(chNext)) {
|
|
if (chNext == ']') {
|
|
mode = 0;
|
|
state = SCE_L_DEFAULT;
|
|
}
|
|
styler.ColourTo(i + 1, SCE_L_SHORTCMD);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
}
|
|
break;
|
|
case '$' :
|
|
styler.ColourTo(i - 1, state);
|
|
if (chNext == '$') {
|
|
styler.ColourTo(i + 1, SCE_L_SHORTCMD);
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
mode = 0;
|
|
state = SCE_L_DEFAULT;
|
|
} else { // This may not be an error, e.g. \begin{equation}\text{$a$}\end{equation}
|
|
styler.ColourTo(i, SCE_L_SHORTCMD);
|
|
}
|
|
break;
|
|
case '%' :
|
|
styler.ColourTo(i - 1, state);
|
|
state = SCE_L_COMMENT;
|
|
break;
|
|
}
|
|
break;
|
|
case SCE_L_COMMENT :
|
|
if (ch == '\r' || ch == '\n') {
|
|
styler.ColourTo(i - 1, state);
|
|
latexStateReset(mode, state);
|
|
}
|
|
break;
|
|
case SCE_L_COMMENT2 :
|
|
if (ch == '\\') {
|
|
Sci_Position match = i + 3;
|
|
if (latexLastWordIs(match, styler, "\\end")) {
|
|
match++;
|
|
if (latexIsTagValid(match, lengthDoc, styler)) {
|
|
if (latexLastWordIs(match, styler, "{comment}")) {
|
|
styler.ColourTo(i - 1, state);
|
|
state = SCE_L_COMMAND;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
break;
|
|
case SCE_L_VERBATIM :
|
|
if (ch == '\\') {
|
|
Sci_Position match = i + 3;
|
|
if (latexLastWordIs(match, styler, "\\end")) {
|
|
match++;
|
|
if (latexIsTagValid(match, lengthDoc, styler)) {
|
|
if (latexLastWordIs(match, styler, "{verbatim}")) {
|
|
styler.ColourTo(i - 1, state);
|
|
state = SCE_L_COMMAND;
|
|
}
|
|
}
|
|
}
|
|
} else if (chNext == chVerbatimDelim) {
|
|
styler.ColourTo(i + 1, state);
|
|
latexStateReset(mode, state);
|
|
chVerbatimDelim = '\0';
|
|
i++;
|
|
chNext = styler.SafeGetCharAt(i + 1);
|
|
} else if (chVerbatimDelim != '\0' && (ch == '\n' || ch == '\r')) {
|
|
styler.ColourTo(i, SCE_L_ERROR);
|
|
latexStateReset(mode, state);
|
|
chVerbatimDelim = '\0';
|
|
}
|
|
break;
|
|
}
|
|
}
|
|
if (lengthDoc == styler.Length()) truncModes(styler.GetLine(lengthDoc - 1));
|
|
styler.ColourTo(lengthDoc - 1, state);
|
|
styler.Flush();
|
|
}
|
|
|
|
static int latexFoldSaveToInt(const latexFoldSave &save) {
|
|
int sum = 0;
|
|
for (int i = 0; i <= save.structLev; ++i)
|
|
sum += save.openBegins[i];
|
|
return ((sum + save.structLev + SC_FOLDLEVELBASE) & SC_FOLDLEVELNUMBERMASK);
|
|
}
|
|
|
|
// Change folding state while processing a line
|
|
// Return the level before the first relevant command
|
|
void SCI_METHOD LexerLaTeX::Fold(Sci_PositionU startPos, Sci_Position length, int, IDocument *pAccess) {
|
|
const char *structWords[7] = {"part", "chapter", "section", "subsection",
|
|
"subsubsection", "paragraph", "subparagraph"};
|
|
Accessor styler(pAccess, &props);
|
|
Sci_PositionU endPos = startPos + length;
|
|
Sci_Position curLine = styler.GetLine(startPos);
|
|
latexFoldSave save;
|
|
getSave(curLine - 1, save);
|
|
do {
|
|
char ch, buf[16];
|
|
Sci_Position i, j;
|
|
int lev = -1;
|
|
bool needFold = false;
|
|
for (i = static_cast<Sci_Position>(startPos); i < static_cast<Sci_Position>(endPos); ++i) {
|
|
ch = styler.SafeGetCharAt(i);
|
|
if (ch == '\r' || ch == '\n') break;
|
|
if (ch != '\\' || styler.StyleAt(i) != SCE_L_COMMAND) continue;
|
|
for (j = 0; j < 15 && i + 1 < static_cast<Sci_Position>(endPos); ++j, ++i) {
|
|
buf[j] = styler.SafeGetCharAt(i + 1);
|
|
if (!latexIsLetter(buf[j])) break;
|
|
}
|
|
buf[j] = '\0';
|
|
if (strcmp(buf, "begin") == 0) {
|
|
if (lev < 0) lev = latexFoldSaveToInt(save);
|
|
++save.openBegins[save.structLev];
|
|
needFold = true;
|
|
}
|
|
else if (strcmp(buf, "end") == 0) {
|
|
while (save.structLev > 0 && save.openBegins[save.structLev] == 0)
|
|
--save.structLev;
|
|
if (lev < 0) lev = latexFoldSaveToInt(save);
|
|
if (save.openBegins[save.structLev] > 0) --save.openBegins[save.structLev];
|
|
}
|
|
else {
|
|
for (j = 0; j < 7; ++j)
|
|
if (strcmp(buf, structWords[j]) == 0) break;
|
|
if (j >= 7) continue;
|
|
save.structLev = j; // level before the command
|
|
for (j = save.structLev + 1; j < 8; ++j) {
|
|
save.openBegins[save.structLev] += save.openBegins[j];
|
|
save.openBegins[j] = 0;
|
|
}
|
|
if (lev < 0) lev = latexFoldSaveToInt(save);
|
|
++save.structLev; // level after the command
|
|
needFold = true;
|
|
}
|
|
}
|
|
if (lev < 0) lev = latexFoldSaveToInt(save);
|
|
if (needFold) lev |= SC_FOLDLEVELHEADERFLAG;
|
|
styler.SetLevel(curLine, lev);
|
|
setSave(curLine, save);
|
|
++curLine;
|
|
startPos = styler.LineStart(curLine);
|
|
if (static_cast<Sci_Position>(startPos) == styler.Length()) {
|
|
lev = latexFoldSaveToInt(save);
|
|
styler.SetLevel(curLine, lev);
|
|
setSave(curLine, save);
|
|
truncSaves(curLine);
|
|
}
|
|
} while (startPos < endPos);
|
|
styler.Flush();
|
|
}
|
|
|
|
static const char *const emptyWordListDesc[] = {
|
|
0
|
|
};
|
|
|
|
LexerModule lmLatex(SCLEX_LATEX, LexerLaTeX::LexerFactoryLaTeX, "latex", emptyWordListDesc);
|