1
// Scintilla source code edit control
3
** Lexer for C++, C, Java, and JavaScript.
4
** Further folding features and configuration properties added by "Udo Lechner" <dlchnr(at)gmx(dot)net>
6
// Copyright 1998-2005 by Neil Hodgson <neilh@scintilla.org>
7
// The License.txt file describes the conditions under which this software may be distributed.
17
#pragma warning(disable: 4786)
26
#include "Scintilla.h"
30
#include "LexAccessor.h"
32
#include "StyleContext.h"
33
#include "CharacterSet.h"
34
#include "LexerModule.h"
35
#include "OptionSet.h"
36
#include "SparseState.h"
39
using namespace Scintilla;
42
static bool IsSpaceEquiv(int state) {
43
return (state <= SCE_C_COMMENTDOC) ||
44
// including SCE_C_DEFAULT, SCE_C_COMMENT, SCE_C_COMMENTLINE
45
(state == SCE_C_COMMENTLINEDOC) || (state == SCE_C_COMMENTDOCKEYWORD) ||
46
(state == SCE_C_COMMENTDOCKEYWORDERROR);
49
// Preconditions: sc.currentPos points to a character after '+' or '-'.
50
// The test for pos reaching 0 should be redundant,
51
// and is in only for safety measures.
52
// Limitation: this code will give the incorrect answer for code like
54
// Putting a space between the '++' post-inc operator and the '+' binary op
55
// fixes this, and is highly recommended for readability anyway.
56
static bool FollowsPostfixOperator(StyleContext &sc, LexAccessor &styler) {
57
int pos = (int) sc.currentPos;
59
char ch = styler[pos];
60
if (ch == '+' || ch == '-') {
61
return styler[pos - 1] == ch;
67
static bool followsReturnKeyword(StyleContext &sc, LexAccessor &styler) {
68
// Don't look at styles, so no need to flush.
69
int pos = (int) sc.currentPos;
70
int currentLine = styler.GetLine(pos);
71
int lineStartPos = styler.LineStart(currentLine);
73
while (--pos > lineStartPos) {
74
ch = styler.SafeGetCharAt(pos);
75
if (ch != ' ' && ch != '\t') {
79
const char *retBack = "nruter";
80
const char *s = retBack;
82
&& pos >= lineStartPos
83
&& styler.SafeGetCharAt(pos) == *s) {
90
static std::string GetRestOfLine(LexAccessor &styler, int start, bool allowSpace) {
91
std::string restOfLine;
93
char ch = styler.SafeGetCharAt(start + i, '\n');
94
while ((ch != '\r') && (ch != '\n')) {
95
if (allowSpace || (ch != ' '))
98
ch = styler.SafeGetCharAt(start + i, '\n');
103
static bool IsStreamCommentStyle(int style) {
104
return style == SCE_C_COMMENT ||
105
style == SCE_C_COMMENTDOC ||
106
style == SCE_C_COMMENTDOCKEYWORD ||
107
style == SCE_C_COMMENTDOCKEYWORDERROR;
110
static std::vector<std::string> Tokenize(const std::string &s) {
111
// Break into space separated tokens
113
std::vector<std::string> tokens;
114
for (const char *cp = s.c_str(); *cp; cp++) {
115
if ((*cp == ' ') || (*cp == '\t')) {
117
tokens.push_back(word);
125
tokens.push_back(word);
130
struct PPDefinition {
134
PPDefinition(int line_, const std::string &key_, const std::string &value_) :
135
line(line_), key(key_), value(value_) {
143
bool ValidLevel() const {
144
return level >= 0 && level < 32;
146
int maskLevel() const {
150
LinePPState() : state(0), ifTaken(0), level(-1) {
152
bool IsInactive() const {
155
bool CurrentIfTaken() {
156
return (ifTaken & maskLevel()) != 0;
158
void StartSection(bool on) {
162
state &= ~maskLevel();
163
ifTaken |= maskLevel();
165
state |= maskLevel();
166
ifTaken &= ~maskLevel();
172
state &= ~maskLevel();
173
ifTaken &= ~maskLevel();
177
void InvertCurrentLevel() {
179
state ^= maskLevel();
180
ifTaken |= maskLevel();
185
// Hold the preprocessor state for each line seen.
186
// Currently one entry per line but could become sparse with just one entry per preprocessor line.
188
std::vector<LinePPState> vlls;
190
LinePPState ForLine(int line) {
191
if ((line > 0) && (vlls.size() > static_cast<size_t>(line))) {
194
return LinePPState();
197
void Add(int line, LinePPState lls) {
203
// An individual named option for use in an OptionSet
205
// Options used for LexerCPP
207
bool stylingWithinPreprocessor;
208
bool identifiersAllowDollars;
209
bool trackPreprocessor;
210
bool updatePreprocessor;
211
bool triplequotedStrings;
213
bool foldSyntaxBased;
215
bool foldCommentMultiline;
216
bool foldCommentExplicit;
217
std::string foldExplicitStart;
218
std::string foldExplicitEnd;
219
bool foldExplicitAnywhere;
220
bool foldPreprocessor;
224
stylingWithinPreprocessor = false;
225
identifiersAllowDollars = true;
226
trackPreprocessor = true;
227
updatePreprocessor = true;
228
triplequotedStrings = false;
230
foldSyntaxBased = true;
232
foldCommentMultiline = true;
233
foldCommentExplicit = true;
234
foldExplicitStart = "";
235
foldExplicitEnd = "";
236
foldExplicitAnywhere = false;
237
foldPreprocessor = false;
243
static const char *const cppWordLists[] = {
244
"Primary keywords and identifiers",
245
"Secondary keywords and identifiers",
246
"Documentation comment keywords",
247
"Global classes and typedefs",
248
"Preprocessor definitions",
252
struct OptionSetCPP : public OptionSet<OptionsCPP> {
254
DefineProperty("styling.within.preprocessor", &OptionsCPP::stylingWithinPreprocessor,
255
"For C++ code, determines whether all preprocessor code is styled in the "
256
"preprocessor style (0, the default) or only from the initial # to the end "
257
"of the command word(1).");
259
DefineProperty("lexer.cpp.allow.dollars", &OptionsCPP::identifiersAllowDollars,
260
"Set to 0 to disallow the '$' character in identifiers with the cpp lexer.");
262
DefineProperty("lexer.cpp.track.preprocessor", &OptionsCPP::trackPreprocessor,
263
"Set to 1 to interpret #if/#else/#endif to grey out code that is not active.");
265
DefineProperty("lexer.cpp.update.preprocessor", &OptionsCPP::updatePreprocessor,
266
"Set to 1 to update preprocessor definitions when #define found.");
268
DefineProperty("lexer.cpp.triplequoted.strings", &OptionsCPP::triplequotedStrings,
269
"Set to 1 to enable highlighting of triple-quoted strings.");
271
DefineProperty("fold", &OptionsCPP::fold);
273
DefineProperty("fold.cpp.syntax.based", &OptionsCPP::foldSyntaxBased,
274
"Set this property to 0 to disable syntax based folding.");
276
DefineProperty("fold.comment", &OptionsCPP::foldComment,
277
"This option enables folding multi-line comments and explicit fold points when using the C++ lexer. "
278
"Explicit fold points allows adding extra folding by placing a //{ comment at the start and a //} "
279
"at the end of a section that should fold.");
281
DefineProperty("fold.cpp.comment.multiline", &OptionsCPP::foldCommentMultiline,
282
"Set this property to 0 to disable folding multi-line comments when fold.comment=1.");
284
DefineProperty("fold.cpp.comment.explicit", &OptionsCPP::foldCommentExplicit,
285
"Set this property to 0 to disable folding explicit fold points when fold.comment=1.");
287
DefineProperty("fold.cpp.explicit.start", &OptionsCPP::foldExplicitStart,
288
"The string to use for explicit fold start points, replacing the standard //{.");
290
DefineProperty("fold.cpp.explicit.end", &OptionsCPP::foldExplicitEnd,
291
"The string to use for explicit fold end points, replacing the standard //}.");
293
DefineProperty("fold.cpp.explicit.anywhere", &OptionsCPP::foldExplicitAnywhere,
294
"Set this property to 1 to enable explicit fold points anywhere, not just in line comments.");
296
DefineProperty("fold.preprocessor", &OptionsCPP::foldPreprocessor,
297
"This option enables folding preprocessor directives when using the C++ lexer. "
298
"Includes C#'s explicit #region and #endregion folding directives.");
300
DefineProperty("fold.compact", &OptionsCPP::foldCompact);
302
DefineProperty("fold.at.else", &OptionsCPP::foldAtElse,
303
"This option enables C++ folding on a \"} else {\" line of an if statement.");
305
DefineWordListSets(cppWordLists);
309
class LexerCPP : public ILexer {
311
CharacterSet setWord;
312
CharacterSet setNegationOp;
313
CharacterSet setArithmethicOp;
314
CharacterSet setRelOp;
315
CharacterSet setLogicalOp;
317
std::vector<PPDefinition> ppDefineHistory;
322
WordList ppDefinitions;
323
std::map<std::string, std::string> preprocessorDefinitionsStart;
326
SparseState<std::string> rawStringTerminators;
328
LexerCPP(bool caseSensitive_) :
329
caseSensitive(caseSensitive_),
330
setWord(CharacterSet::setAlphaNum, "._", 0x80, true),
331
setNegationOp(CharacterSet::setNone, "!"),
332
setArithmethicOp(CharacterSet::setNone, "+-/*%"),
333
setRelOp(CharacterSet::setNone, "=!<>"),
334
setLogicalOp(CharacterSet::setNone, "|&") {
338
void SCI_METHOD Release() {
341
int SCI_METHOD Version() const {
344
const char * SCI_METHOD PropertyNames() {
345
return osCPP.PropertyNames();
347
int SCI_METHOD PropertyType(const char *name) {
348
return osCPP.PropertyType(name);
350
const char * SCI_METHOD DescribeProperty(const char *name) {
351
return osCPP.DescribeProperty(name);
353
int SCI_METHOD PropertySet(const char *key, const char *val);
354
const char * SCI_METHOD DescribeWordListSets() {
355
return osCPP.DescribeWordListSets();
357
int SCI_METHOD WordListSet(int n, const char *wl);
358
void SCI_METHOD Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
359
void SCI_METHOD Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
361
void * SCI_METHOD PrivateCall(int, void *) {
365
static ILexer *LexerFactoryCPP() {
366
return new LexerCPP(true);
368
static ILexer *LexerFactoryCPPInsensitive() {
369
return new LexerCPP(false);
372
void EvaluateTokens(std::vector<std::string> &tokens);
373
bool EvaluateExpression(const std::string &expr, const std::map<std::string, std::string> &preprocessorDefinitions);
376
int SCI_METHOD LexerCPP::PropertySet(const char *key, const char *val) {
377
if (osCPP.PropertySet(&options, key, val)) {
383
int SCI_METHOD LexerCPP::WordListSet(int n, const char *wl) {
384
WordList *wordListN = 0;
387
wordListN = &keywords;
390
wordListN = &keywords2;
393
wordListN = &keywords3;
396
wordListN = &keywords4;
399
wordListN = &ppDefinitions;
402
int firstModification = -1;
406
if (*wordListN != wlNew) {
408
firstModification = 0;
410
// Rebuild preprocessorDefinitions
411
preprocessorDefinitionsStart.clear();
412
for (int nDefinition = 0; nDefinition < ppDefinitions.len; nDefinition++) {
413
char *cpDefinition = ppDefinitions.words[nDefinition];
414
char *cpEquals = strchr(cpDefinition, '=');
416
std::string name(cpDefinition, cpEquals - cpDefinition);
417
std::string val(cpEquals+1);
418
preprocessorDefinitionsStart[name] = val;
420
std::string name(cpDefinition);
421
std::string val("1");
422
preprocessorDefinitionsStart[name] = val;
428
return firstModification;
431
// Functor used to truncate history
434
After(int line_) : line(line_) {}
435
bool operator() (PPDefinition &p) const {
436
return p.line > line;
440
void SCI_METHOD LexerCPP::Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
441
LexAccessor styler(pAccess);
443
CharacterSet setOKBeforeRE(CharacterSet::setNone, "([{=,:;!%^&*|?~+-");
444
CharacterSet setCouldBePostOp(CharacterSet::setNone, "+-");
446
CharacterSet setDoxygen(CharacterSet::setAlpha, "$@\\&<>#{}[]");
448
CharacterSet setWordStart(CharacterSet::setAlpha, "_", 0x80, true);
450
if (options.identifiersAllowDollars) {
451
setWordStart.Add('$');
455
int chPrevNonWhite = ' ';
456
int visibleChars = 0;
457
bool lastWordWasUUID = false;
458
int styleBeforeDCKeyword = SCE_C_DEFAULT;
459
bool continuationLine = false;
460
bool isIncludePreprocessor = false;
462
int lineCurrent = styler.GetLine(startPos);
463
if ((initStyle == SCE_C_PREPROCESSOR) ||
464
(initStyle == SCE_C_COMMENTLINE) ||
465
(initStyle == SCE_C_COMMENTLINEDOC)) {
466
// Set continuationLine if last character of previous line is '\'
467
if (lineCurrent > 0) {
468
int chBack = styler.SafeGetCharAt(startPos-1, 0);
469
int chBack2 = styler.SafeGetCharAt(startPos-2, 0);
470
int lineEndChar = '!';
471
if (chBack2 == '\r' && chBack == '\n') {
472
lineEndChar = styler.SafeGetCharAt(startPos-3, 0);
473
} else if (chBack == '\n' || chBack == '\r') {
474
lineEndChar = chBack2;
476
continuationLine = lineEndChar == '\\';
480
// look back to set chPrevNonWhite properly for better regex colouring
483
while (--back && IsSpaceEquiv(styler.StyleAt(back)))
485
if (styler.StyleAt(back) == SCE_C_OPERATOR) {
486
chPrevNonWhite = styler.SafeGetCharAt(back);
490
StyleContext sc(startPos, length, initStyle, styler, 0x7f);
491
LinePPState preproc = vlls.ForLine(lineCurrent);
493
bool definitionsChanged = false;
495
// Truncate ppDefineHistory before current line
497
if (!options.updatePreprocessor)
498
ppDefineHistory.clear();
500
std::vector<PPDefinition>::iterator itInvalid = std::find_if(ppDefineHistory.begin(), ppDefineHistory.end(), After(lineCurrent-1));
501
if (itInvalid != ppDefineHistory.end()) {
502
ppDefineHistory.erase(itInvalid, ppDefineHistory.end());
503
definitionsChanged = true;
506
std::map<std::string, std::string> preprocessorDefinitions = preprocessorDefinitionsStart;
507
for (std::vector<PPDefinition>::iterator itDef = ppDefineHistory.begin(); itDef != ppDefineHistory.end(); ++itDef) {
508
preprocessorDefinitions[itDef->key] = itDef->value;
511
const int maskActivity = 0x3F;
512
std::string rawStringTerminator = rawStringTerminators.ValueAt(lineCurrent-1);
513
SparseState<std::string> rawSTNew(lineCurrent);
515
int activitySet = preproc.IsInactive() ? 0x40 : 0;
517
for (; sc.More(); sc.Forward()) {
519
if (sc.atLineStart) {
520
if ((sc.state == SCE_C_STRING) || (sc.state == SCE_C_CHARACTER)) {
521
// Prevent SCE_C_STRINGEOL from leaking back to previous line which
522
// ends with a line continuation by locking in the state upto this position.
523
sc.SetState(sc.state);
525
// Reset states to begining of colourise so no surprises
526
// if different sets of lines lexed.
528
lastWordWasUUID = false;
529
isIncludePreprocessor = false;
530
if (preproc.IsInactive()) {
532
sc.SetState(sc.state | activitySet);
536
if (sc.Match("#else") || sc.Match("#end") || sc.Match("#if")) {
545
vlls.Add(lineCurrent, preproc);
546
if (rawStringTerminator != "") {
547
rawSTNew.Set(lineCurrent-1, rawStringTerminator);
551
// Handle line continuation generically.
553
if (sc.chNext == '\n' || sc.chNext == '\r') {
555
if (sc.ch == '\r' && sc.chNext == '\n') {
558
continuationLine = true;
563
const bool atLineEndBeforeSwitch = sc.atLineEnd;
565
// Determine if the current state should terminate.
566
switch (sc.state & maskActivity) {
568
sc.SetState(SCE_C_DEFAULT|activitySet);
571
// We accept almost anything because of hex. and number suffixes
572
if (!(setWord.Contains(sc.ch) || ((sc.ch == '+' || sc.ch == '-') && (sc.chPrev == 'e' || sc.chPrev == 'E')))) {
573
sc.SetState(SCE_C_DEFAULT|activitySet);
576
case SCE_C_IDENTIFIER:
577
if (!setWord.Contains(sc.ch) || (sc.ch == '.')) {
580
sc.GetCurrent(s, sizeof(s));
582
sc.GetCurrentLowered(s, sizeof(s));
584
if (keywords.InList(s)) {
585
lastWordWasUUID = strcmp(s, "uuid") == 0;
586
sc.ChangeState(SCE_C_WORD|activitySet);
587
} else if (keywords2.InList(s)) {
588
sc.ChangeState(SCE_C_WORD2|activitySet);
589
} else if (keywords4.InList(s)) {
590
sc.ChangeState(SCE_C_GLOBALCLASS|activitySet);
592
const bool literalString = sc.ch == '\"';
593
if (literalString || sc.ch == '\'') {
594
size_t lenS = strlen(s);
595
const bool raw = literalString && sc.chPrev == 'R';
600
((lenS == 1) && ((s[0] == 'L') || (s[0] == 'u') || (s[0] == 'U'))) ||
601
((lenS == 2) && literalString && (s[0] == 'u') && (s[1] == '8'));
604
sc.ChangeState((raw ? SCE_C_STRINGRAW : SCE_C_STRING)|activitySet);
606
sc.ChangeState(SCE_C_CHARACTER|activitySet);
609
sc.SetState(SCE_C_DEFAULT|activitySet);
612
case SCE_C_PREPROCESSOR:
613
if (sc.atLineStart && !continuationLine) {
614
sc.SetState(SCE_C_DEFAULT|activitySet);
615
} else if (options.stylingWithinPreprocessor) {
616
if (IsASpace(sc.ch)) {
617
sc.SetState(SCE_C_DEFAULT|activitySet);
620
if (sc.Match('/', '*') || sc.Match('/', '/')) {
621
sc.SetState(SCE_C_DEFAULT|activitySet);
626
if (sc.Match('*', '/')) {
628
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
631
case SCE_C_COMMENTDOC:
632
if (sc.Match('*', '/')) {
634
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
635
} else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
636
// Verify that we have the conditions to mark a comment-doc-keyword
637
if ((IsASpace(sc.chPrev) || sc.chPrev == '*') && (!IsASpace(sc.chNext))) {
638
styleBeforeDCKeyword = SCE_C_COMMENTDOC;
639
sc.SetState(SCE_C_COMMENTDOCKEYWORD|activitySet);
643
case SCE_C_COMMENTLINE:
644
if (sc.atLineStart && !continuationLine) {
645
sc.SetState(SCE_C_DEFAULT|activitySet);
648
case SCE_C_COMMENTLINEDOC:
649
if (sc.atLineStart && !continuationLine) {
650
sc.SetState(SCE_C_DEFAULT|activitySet);
651
} else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
652
// Verify that we have the conditions to mark a comment-doc-keyword
653
if ((IsASpace(sc.chPrev) || sc.chPrev == '/' || sc.chPrev == '!') && (!IsASpace(sc.chNext))) {
654
styleBeforeDCKeyword = SCE_C_COMMENTLINEDOC;
655
sc.SetState(SCE_C_COMMENTDOCKEYWORD|activitySet);
659
case SCE_C_COMMENTDOCKEYWORD:
660
if ((styleBeforeDCKeyword == SCE_C_COMMENTDOC) && sc.Match('*', '/')) {
661
sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR);
663
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
664
} else if (!setDoxygen.Contains(sc.ch)) {
667
sc.GetCurrent(s, sizeof(s));
669
sc.GetCurrentLowered(s, sizeof(s));
671
if (!IsASpace(sc.ch) || !keywords3.InList(s + 1)) {
672
sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR|activitySet);
674
sc.SetState(styleBeforeDCKeyword);
679
sc.ChangeState(SCE_C_STRINGEOL|activitySet);
680
} else if (isIncludePreprocessor) {
682
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
683
isIncludePreprocessor = false;
685
} else if (sc.ch == '\\') {
686
if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
689
} else if (sc.ch == '\"') {
690
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
693
case SCE_C_STRINGRAW:
694
if (sc.Match(rawStringTerminator.c_str())) {
695
for (size_t termPos=rawStringTerminator.size(); termPos; termPos--)
697
sc.SetState(SCE_C_DEFAULT|activitySet);
698
rawStringTerminator = "";
701
case SCE_C_CHARACTER:
703
sc.ChangeState(SCE_C_STRINGEOL|activitySet);
704
} else if (sc.ch == '\\') {
705
if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
708
} else if (sc.ch == '\'') {
709
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
713
if (sc.atLineStart) {
714
sc.SetState(SCE_C_DEFAULT|activitySet);
715
} else if (sc.ch == '/') {
717
while ((sc.ch < 0x80) && islower(sc.ch))
718
sc.Forward(); // gobble regex flags
719
sc.SetState(SCE_C_DEFAULT|activitySet);
720
} else if (sc.ch == '\\') {
721
// Gobble up the quoted character
722
if (sc.chNext == '\\' || sc.chNext == '/') {
727
case SCE_C_STRINGEOL:
728
if (sc.atLineStart) {
729
sc.SetState(SCE_C_DEFAULT|activitySet);
734
if (sc.chNext == '\"') {
737
sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
741
case SCE_C_TRIPLEVERBATIM:
742
if (sc.Match ("\"\"\"")) {
743
while (sc.Match('"')) {
746
sc.SetState(SCE_C_DEFAULT|activitySet);
750
if (sc.ch == '\r' || sc.ch == '\n' || sc.ch == ')') {
751
sc.SetState(SCE_C_DEFAULT|activitySet);
755
if (sc.atLineEnd && !atLineEndBeforeSwitch) {
756
// State exit processing consumed characters up to end of line.
758
vlls.Add(lineCurrent, preproc);
761
// Determine if a new state should be entered.
762
if ((sc.state & maskActivity) == SCE_C_DEFAULT) {
763
if (sc.Match('@', '\"')) {
764
sc.SetState(SCE_C_VERBATIM|activitySet);
766
} else if (options.triplequotedStrings && sc.Match("\"\"\"")) {
767
sc.SetState(SCE_C_TRIPLEVERBATIM|activitySet);
769
} else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
770
if (lastWordWasUUID) {
771
sc.SetState(SCE_C_UUID|activitySet);
772
lastWordWasUUID = false;
774
sc.SetState(SCE_C_NUMBER|activitySet);
776
} else if (setWordStart.Contains(sc.ch) || (sc.ch == '@')) {
777
if (lastWordWasUUID) {
778
sc.SetState(SCE_C_UUID|activitySet);
779
lastWordWasUUID = false;
781
sc.SetState(SCE_C_IDENTIFIER|activitySet);
783
} else if (sc.Match('/', '*')) {
784
if (sc.Match("/**") || sc.Match("/*!")) { // Support of Qt/Doxygen doc. style
785
sc.SetState(SCE_C_COMMENTDOC|activitySet);
787
sc.SetState(SCE_C_COMMENT|activitySet);
789
sc.Forward(); // Eat the * so it isn't used for the end of the comment
790
} else if (sc.Match('/', '/')) {
791
if ((sc.Match("///") && !sc.Match("////")) || sc.Match("//!"))
792
// Support of Qt/Doxygen doc. style
793
sc.SetState(SCE_C_COMMENTLINEDOC|activitySet);
795
sc.SetState(SCE_C_COMMENTLINE|activitySet);
796
} else if (sc.ch == '/'
797
&& (setOKBeforeRE.Contains(chPrevNonWhite)
798
|| followsReturnKeyword(sc, styler))
799
&& (!setCouldBePostOp.Contains(chPrevNonWhite)
800
|| !FollowsPostfixOperator(sc, styler))) {
801
sc.SetState(SCE_C_REGEX|activitySet); // JavaScript's RegEx
802
} else if (sc.ch == '\"') {
803
if (sc.chPrev == 'R') {
804
sc.SetState(SCE_C_STRINGRAW|activitySet);
805
rawStringTerminator = ")";
806
for (int termPos = sc.currentPos + 1;;termPos++) {
807
char chTerminator = styler.SafeGetCharAt(termPos, '(');
808
if (chTerminator == '(')
810
rawStringTerminator += chTerminator;
812
rawStringTerminator += '\"';
814
sc.SetState(SCE_C_STRING|activitySet);
816
isIncludePreprocessor = false; // ensure that '>' won't end the string
817
} else if (isIncludePreprocessor && sc.ch == '<') {
818
sc.SetState(SCE_C_STRING|activitySet);
819
} else if (sc.ch == '\'') {
820
sc.SetState(SCE_C_CHARACTER|activitySet);
821
} else if (sc.ch == '#' && visibleChars == 0) {
822
// Preprocessor commands are alone on their line
823
sc.SetState(SCE_C_PREPROCESSOR|activitySet);
824
// Skip whitespace between # and preprocessor word
827
} while ((sc.ch == ' ' || sc.ch == '\t') && sc.More());
829
sc.SetState(SCE_C_DEFAULT|activitySet);
830
} else if (sc.Match("include")) {
831
isIncludePreprocessor = true;
833
if (options.trackPreprocessor) {
834
if (sc.Match("ifdef") || sc.Match("ifndef")) {
835
bool isIfDef = sc.Match("ifdef");
836
int i = isIfDef ? 5 : 6;
837
std::string restOfLine = GetRestOfLine(styler, sc.currentPos + i + 1, false);
838
bool foundDef = preprocessorDefinitions.find(restOfLine) != preprocessorDefinitions.end();
839
preproc.StartSection(isIfDef == foundDef);
840
} else if (sc.Match("if")) {
841
std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 2, true);
842
bool ifGood = EvaluateExpression(restOfLine, preprocessorDefinitions);
843
preproc.StartSection(ifGood);
844
} else if (sc.Match("else")) {
845
if (!preproc.CurrentIfTaken()) {
846
preproc.InvertCurrentLevel();
847
activitySet = preproc.IsInactive() ? 0x40 : 0;
849
sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
850
} else if (!preproc.IsInactive()) {
851
preproc.InvertCurrentLevel();
852
activitySet = preproc.IsInactive() ? 0x40 : 0;
854
sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
856
} else if (sc.Match("elif")) {
857
// Ensure only one chosen out of #if .. #elif .. #elif .. #else .. #endif
858
if (!preproc.CurrentIfTaken()) {
860
std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 2, true);
861
bool ifGood = EvaluateExpression(restOfLine, preprocessorDefinitions);
863
preproc.InvertCurrentLevel();
864
activitySet = preproc.IsInactive() ? 0x40 : 0;
866
sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
868
} else if (!preproc.IsInactive()) {
869
preproc.InvertCurrentLevel();
870
activitySet = preproc.IsInactive() ? 0x40 : 0;
872
sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
874
} else if (sc.Match("endif")) {
875
preproc.EndSection();
876
activitySet = preproc.IsInactive() ? 0x40 : 0;
877
sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
878
} else if (sc.Match("define")) {
879
if (options.updatePreprocessor && !preproc.IsInactive()) {
880
std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 6, true);
881
if (restOfLine.find(")") == std::string::npos) { // Don't handle macros with arguments
882
std::vector<std::string> tokens = Tokenize(restOfLine);
884
std::string value("1");
885
if (tokens.size() >= 1) {
887
if (tokens.size() >= 2) {
890
preprocessorDefinitions[key] = value;
891
ppDefineHistory.push_back(PPDefinition(lineCurrent, key, value));
892
definitionsChanged = true;
899
} else if (isoperator(static_cast<char>(sc.ch))) {
900
sc.SetState(SCE_C_OPERATOR|activitySet);
904
if (!IsASpace(sc.ch) && !IsSpaceEquiv(sc.state)) {
905
chPrevNonWhite = sc.ch;
908
continuationLine = false;
910
const bool rawStringsChanged = rawStringTerminators.Merge(rawSTNew, lineCurrent);
911
if (definitionsChanged || rawStringsChanged)
912
styler.ChangeLexerState(startPos, startPos + length);
916
// Store both the current line's fold level and the next lines in the
917
// level store to make it easy to pick up with each increment
918
// and to make it possible to fiddle the current level for "} else {".
920
void SCI_METHOD LexerCPP::Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
925
LexAccessor styler(pAccess);
927
unsigned int endPos = startPos + length;
928
int visibleChars = 0;
929
int lineCurrent = styler.GetLine(startPos);
930
int levelCurrent = SC_FOLDLEVELBASE;
932
levelCurrent = styler.LevelAt(lineCurrent-1) >> 16;
933
int levelMinCurrent = levelCurrent;
934
int levelNext = levelCurrent;
935
char chNext = styler[startPos];
936
int styleNext = styler.StyleAt(startPos);
937
int style = initStyle;
938
const bool userDefinedFoldMarkers = !options.foldExplicitStart.empty() && !options.foldExplicitEnd.empty();
939
for (unsigned int i = startPos; i < endPos; i++) {
941
chNext = styler.SafeGetCharAt(i + 1);
942
int stylePrev = style;
944
styleNext = styler.StyleAt(i + 1);
945
bool atEOL = (ch == '\r' && chNext != '\n') || (ch == '\n');
946
if (options.foldComment && options.foldCommentMultiline && IsStreamCommentStyle(style)) {
947
if (!IsStreamCommentStyle(stylePrev) && (stylePrev != SCE_C_COMMENTLINEDOC)) {
949
} else if (!IsStreamCommentStyle(styleNext) && (styleNext != SCE_C_COMMENTLINEDOC) && !atEOL) {
950
// Comments don't end at end of line and the next character may be unstyled.
954
if (options.foldComment && options.foldCommentExplicit && ((style == SCE_C_COMMENTLINE) || options.foldExplicitAnywhere)) {
955
if (userDefinedFoldMarkers) {
956
if (styler.Match(i, options.foldExplicitStart.c_str())) {
958
} else if (styler.Match(i, options.foldExplicitEnd.c_str())) {
962
if ((ch == '/') && (chNext == '/')) {
963
char chNext2 = styler.SafeGetCharAt(i + 2);
964
if (chNext2 == '{') {
966
} else if (chNext2 == '}') {
972
if (options.foldPreprocessor && (style == SCE_C_PREPROCESSOR)) {
974
unsigned int j = i + 1;
975
while ((j < endPos) && IsASpaceOrTab(styler.SafeGetCharAt(j))) {
978
if (styler.Match(j, "region") || styler.Match(j, "if")) {
980
} else if (styler.Match(j, "end")) {
985
if (options.foldSyntaxBased && (style == SCE_C_OPERATOR)) {
987
// Measure the minimum before a '{' to allow
988
// folding on "} else {"
989
if (levelMinCurrent > levelNext) {
990
levelMinCurrent = levelNext;
993
} else if (ch == '}') {
999
if (atEOL || (i == endPos-1)) {
1000
int levelUse = levelCurrent;
1001
if (options.foldSyntaxBased && options.foldAtElse) {
1002
levelUse = levelMinCurrent;
1004
int lev = levelUse | levelNext << 16;
1005
if (visibleChars == 0 && options.foldCompact)
1006
lev |= SC_FOLDLEVELWHITEFLAG;
1007
if (levelUse < levelNext)
1008
lev |= SC_FOLDLEVELHEADERFLAG;
1009
if (lev != styler.LevelAt(lineCurrent)) {
1010
styler.SetLevel(lineCurrent, lev);
1013
levelCurrent = levelNext;
1014
levelMinCurrent = levelCurrent;
1015
if (atEOL && (i == static_cast<unsigned int>(styler.Length()-1))) {
1016
// There is an empty line at end of file so give it same level and empty
1017
styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG);
1024
void LexerCPP::EvaluateTokens(std::vector<std::string> &tokens) {
1026
// Evaluate defined() statements to either 0 or 1
1027
for (size_t i=0; (i+2)<tokens.size();) {
1028
if ((tokens[i] == "defined") && (tokens[i+1] == "(")) {
1029
const char *val = "0";
1030
if (tokens[i+2] == ")") {
1032
tokens.erase(tokens.begin() + i + 1, tokens.begin() + i + 3);
1033
} else if (((i+2)<tokens.size()) && (tokens[i+3] == ")")) {
1035
tokens.erase(tokens.begin() + i + 1, tokens.begin() + i + 4);
1044
// Find bracketed subexpressions and recurse on them
1045
std::vector<std::string>::iterator itBracket = std::find(tokens.begin(), tokens.end(), "(");
1046
std::vector<std::string>::iterator itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1047
while ((itBracket != tokens.end()) && (itEndBracket != tokens.end()) && (itEndBracket > itBracket)) {
1048
std::vector<std::string> inBracket(itBracket + 1, itEndBracket);
1049
EvaluateTokens(inBracket);
1051
// The insertion is done before the removal because there were failures with the opposite approach
1052
tokens.insert(itBracket, inBracket.begin(), inBracket.end());
1053
itBracket = std::find(tokens.begin(), tokens.end(), "(");
1054
itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1055
tokens.erase(itBracket, itEndBracket + 1);
1057
itBracket = std::find(tokens.begin(), tokens.end(), "(");
1058
itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1061
// Evaluate logical negations
1062
for (size_t j=0; (j+1)<tokens.size();) {
1063
if (setNegationOp.Contains(tokens[j][0])) {
1064
int isTrue = atoi(tokens[j+1].c_str());
1065
if (tokens[j] == "!")
1067
std::vector<std::string>::iterator itInsert =
1068
tokens.erase(tokens.begin() + j, tokens.begin() + j + 2);
1069
tokens.insert(itInsert, isTrue ? "1" : "0");
1075
// Evaluate expressions in precedence order
1076
enum precedence { precArithmetic, precRelative, precLogical };
1077
for (int prec=precArithmetic; prec <= precLogical; prec++) {
1078
// Looking at 3 tokens at a time so end at 2 before end
1079
for (size_t k=0; (k+2)<tokens.size();) {
1080
char chOp = tokens[k+1][0];
1082
((prec==precArithmetic) && setArithmethicOp.Contains(chOp)) ||
1083
((prec==precRelative) && setRelOp.Contains(chOp)) ||
1084
((prec==precLogical) && setLogicalOp.Contains(chOp))
1086
int valA = atoi(tokens[k].c_str());
1087
int valB = atoi(tokens[k+2].c_str());
1089
if (tokens[k+1] == "+")
1090
result = valA + valB;
1091
else if (tokens[k+1] == "-")
1092
result = valA - valB;
1093
else if (tokens[k+1] == "*")
1094
result = valA * valB;
1095
else if (tokens[k+1] == "/")
1096
result = valA / (valB ? valB : 1);
1097
else if (tokens[k+1] == "%")
1098
result = valA % (valB ? valB : 1);
1099
else if (tokens[k+1] == "<")
1100
result = valA < valB;
1101
else if (tokens[k+1] == "<=")
1102
result = valA <= valB;
1103
else if (tokens[k+1] == ">")
1104
result = valA > valB;
1105
else if (tokens[k+1] == ">=")
1106
result = valA >= valB;
1107
else if (tokens[k+1] == "==")
1108
result = valA == valB;
1109
else if (tokens[k+1] == "!=")
1110
result = valA != valB;
1111
else if (tokens[k+1] == "||")
1112
result = valA || valB;
1113
else if (tokens[k+1] == "&&")
1114
result = valA && valB;
1116
sprintf(sResult, "%d", result);
1117
std::vector<std::string>::iterator itInsert =
1118
tokens.erase(tokens.begin() + k, tokens.begin() + k + 3);
1119
tokens.insert(itInsert, sResult);
1127
bool LexerCPP::EvaluateExpression(const std::string &expr, const std::map<std::string, std::string> &preprocessorDefinitions) {
1128
// Break into tokens, replacing with definitions
1130
std::vector<std::string> tokens;
1131
const char *cp = expr.c_str();
1133
if (setWord.Contains(*cp)) {
1136
std::map<std::string, std::string>::const_iterator it = preprocessorDefinitions.find(word);
1137
if (it != preprocessorDefinitions.end()) {
1138
tokens.push_back(it->second);
1139
} else if (!word.empty() && ((word[0] >= '0' && word[0] <= '9') || (word == "defined"))) {
1140
tokens.push_back(word);
1146
if ((*cp != ' ') && (*cp != '\t')) {
1147
std::string op(cp, 1);
1148
if (setRelOp.Contains(*cp)) {
1149
if (setRelOp.Contains(cp[1])) {
1153
} else if (setLogicalOp.Contains(*cp)) {
1154
if (setLogicalOp.Contains(cp[1])) {
1159
tokens.push_back(op);
1165
EvaluateTokens(tokens);
1167
// "0" or "" -> false else true
1168
bool isFalse = tokens.empty() ||
1169
((tokens.size() == 1) && ((tokens[0] == "") || tokens[0] == "0"));
1173
LexerModule lmCPP(SCLEX_CPP, LexerCPP::LexerFactoryCPP, "cpp", cppWordLists);
1174
LexerModule lmCPPNoCase(SCLEX_CPPNOCASE, LexerCPP::LexerFactoryCPPInsensitive, "cppnocase", cppWordLists);
added
removed
scintilla/lexers/LexCPP.cxx
