scintilla: Update scintilla with changeset 3662:1d1c06df8a2f using gtk+3
[anjuta-extras.git] / plugins / scintilla / scintilla / LexCPP.cxx
blob22477c32f367baf06e4d7c1c1eb3730cf142f78b
1 // Scintilla source code edit control
2 /** @file LexCPP.cxx
3 ** Lexer for C++, C, Java, and JavaScript.
4 ** Further folding features and configuration properties added by "Udo Lechner" <dlchnr(at)gmx(dot)net>
5 **/
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.
9 #include <stdlib.h>
10 #include <string.h>
11 #include <ctype.h>
12 #include <stdio.h>
13 #include <stdarg.h>
14 #include <assert.h>
16 #ifdef _MSC_VER
17 #pragma warning(disable: 4786)
18 #endif
20 #include <string>
21 #include <vector>
22 #include <map>
23 #include <algorithm>
25 #include "ILexer.h"
26 #include "Scintilla.h"
27 #include "SciLexer.h"
29 #include "WordList.h"
30 #include "LexAccessor.h"
31 #include "Accessor.h"
32 #include "StyleContext.h"
33 #include "CharacterSet.h"
34 #include "LexerModule.h"
35 #include "OptionSet.h"
36 #include "SparseState.h"
38 #ifdef SCI_NAMESPACE
39 using namespace Scintilla;
40 #endif
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
53 // a = b+++/ptn/...
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;
58 while (--pos > 0) {
59 char ch = styler[pos];
60 if (ch == '+' || ch == '-') {
61 return styler[pos - 1] == ch;
64 return false;
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);
72 char ch;
73 while (--pos > lineStartPos) {
74 ch = styler.SafeGetCharAt(pos);
75 if (ch != ' ' && ch != '\t') {
76 break;
79 const char *retBack = "nruter";
80 const char *s = retBack;
81 while (*s
82 && pos >= lineStartPos
83 && styler.SafeGetCharAt(pos) == *s) {
84 s++;
85 pos--;
87 return !*s;
90 static std::string GetRestOfLine(LexAccessor &styler, int start, bool allowSpace) {
91 std::string restOfLine;
92 int i =0;
93 char ch = styler.SafeGetCharAt(start + i, '\n');
94 while ((ch != '\r') && (ch != '\n')) {
95 if (allowSpace || (ch != ' '))
96 restOfLine += ch;
97 i++;
98 ch = styler.SafeGetCharAt(start + i, '\n');
100 return restOfLine;
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
112 std::string word;
113 std::vector<std::string> tokens;
114 for (const char *cp = s.c_str(); *cp; cp++) {
115 if ((*cp == ' ') || (*cp == '\t')) {
116 if (!word.empty()) {
117 tokens.push_back(word);
118 word = "";
120 } else {
121 word += *cp;
124 if (!word.empty()) {
125 tokens.push_back(word);
127 return tokens;
130 struct PPDefinition {
131 int line;
132 std::string key;
133 std::string value;
134 PPDefinition(int line_, const std::string &key_, const std::string &value_) :
135 line(line_), key(key_), value(value_) {
139 class LinePPState {
140 int state;
141 int ifTaken;
142 int level;
143 bool ValidLevel() const {
144 return level >= 0 && level < 32;
146 int maskLevel() const {
147 return 1 << level;
149 public:
150 LinePPState() : state(0), ifTaken(0), level(-1) {
152 bool IsInactive() const {
153 return state != 0;
155 bool CurrentIfTaken() {
156 return (ifTaken & maskLevel()) != 0;
158 void StartSection(bool on) {
159 level++;
160 if (ValidLevel()) {
161 if (on) {
162 state &= ~maskLevel();
163 ifTaken |= maskLevel();
164 } else {
165 state |= maskLevel();
166 ifTaken &= ~maskLevel();
170 void EndSection() {
171 if (ValidLevel()) {
172 state &= ~maskLevel();
173 ifTaken &= ~maskLevel();
175 level--;
177 void InvertCurrentLevel() {
178 if (ValidLevel()) {
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.
187 class PPStates {
188 std::vector<LinePPState> vlls;
189 public:
190 LinePPState ForLine(int line) {
191 if ((line > 0) && (vlls.size() > static_cast<size_t>(line))) {
192 return vlls[line];
193 } else {
194 return LinePPState();
197 void Add(int line, LinePPState lls) {
198 vlls.resize(line+1);
199 vlls[line] = lls;
203 // An individual named option for use in an OptionSet
205 // Options used for LexerCPP
206 struct OptionsCPP {
207 bool stylingWithinPreprocessor;
208 bool identifiersAllowDollars;
209 bool trackPreprocessor;
210 bool updatePreprocessor;
211 bool triplequotedStrings;
212 bool fold;
213 bool foldSyntaxBased;
214 bool foldComment;
215 bool foldCommentMultiline;
216 bool foldCommentExplicit;
217 std::string foldExplicitStart;
218 std::string foldExplicitEnd;
219 bool foldExplicitAnywhere;
220 bool foldPreprocessor;
221 bool foldCompact;
222 bool foldAtElse;
223 OptionsCPP() {
224 stylingWithinPreprocessor = false;
225 identifiersAllowDollars = true;
226 trackPreprocessor = true;
227 updatePreprocessor = true;
228 triplequotedStrings = false;
229 fold = false;
230 foldSyntaxBased = true;
231 foldComment = false;
232 foldCommentMultiline = true;
233 foldCommentExplicit = true;
234 foldExplicitStart = "";
235 foldExplicitEnd = "";
236 foldExplicitAnywhere = false;
237 foldPreprocessor = false;
238 foldCompact = false;
239 foldAtElse = 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> {
253 OptionSetCPP() {
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 {
310 bool caseSensitive;
311 CharacterSet setWord;
312 CharacterSet setNegationOp;
313 CharacterSet setArithmethicOp;
314 CharacterSet setRelOp;
315 CharacterSet setLogicalOp;
316 PPStates vlls;
317 std::vector<PPDefinition> ppDefineHistory;
318 WordList keywords;
319 WordList keywords2;
320 WordList keywords3;
321 WordList keywords4;
322 WordList ppDefinitions;
323 std::map<std::string, std::string> preprocessorDefinitionsStart;
324 OptionsCPP options;
325 OptionSetCPP osCPP;
326 SparseState<std::string> rawStringTerminators;
327 enum { activeFlag = 0x40 };
328 public:
329 LexerCPP(bool caseSensitive_) :
330 caseSensitive(caseSensitive_),
331 setWord(CharacterSet::setAlphaNum, "._", 0x80, true),
332 setNegationOp(CharacterSet::setNone, "!"),
333 setArithmethicOp(CharacterSet::setNone, "+-/*%"),
334 setRelOp(CharacterSet::setNone, "=!<>"),
335 setLogicalOp(CharacterSet::setNone, "|&") {
337 ~LexerCPP() {
339 void SCI_METHOD Release() {
340 delete this;
342 int SCI_METHOD Version() const {
343 return lvOriginal;
345 const char * SCI_METHOD PropertyNames() {
346 return osCPP.PropertyNames();
348 int SCI_METHOD PropertyType(const char *name) {
349 return osCPP.PropertyType(name);
351 const char * SCI_METHOD DescribeProperty(const char *name) {
352 return osCPP.DescribeProperty(name);
354 int SCI_METHOD PropertySet(const char *key, const char *val);
355 const char * SCI_METHOD DescribeWordListSets() {
356 return osCPP.DescribeWordListSets();
358 int SCI_METHOD WordListSet(int n, const char *wl);
359 void SCI_METHOD Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
360 void SCI_METHOD Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
362 void * SCI_METHOD PrivateCall(int, void *) {
363 return 0;
366 static ILexer *LexerFactoryCPP() {
367 return new LexerCPP(true);
369 static ILexer *LexerFactoryCPPInsensitive() {
370 return new LexerCPP(false);
372 static int MaskActive(int style) {
373 return style & ~activeFlag;
375 void EvaluateTokens(std::vector<std::string> &tokens);
376 bool EvaluateExpression(const std::string &expr, const std::map<std::string, std::string> &preprocessorDefinitions);
379 int SCI_METHOD LexerCPP::PropertySet(const char *key, const char *val) {
380 if (osCPP.PropertySet(&options, key, val)) {
381 return 0;
383 return -1;
386 int SCI_METHOD LexerCPP::WordListSet(int n, const char *wl) {
387 WordList *wordListN = 0;
388 switch (n) {
389 case 0:
390 wordListN = &keywords;
391 break;
392 case 1:
393 wordListN = &keywords2;
394 break;
395 case 2:
396 wordListN = &keywords3;
397 break;
398 case 3:
399 wordListN = &keywords4;
400 break;
401 case 4:
402 wordListN = &ppDefinitions;
403 break;
405 int firstModification = -1;
406 if (wordListN) {
407 WordList wlNew;
408 wlNew.Set(wl);
409 if (*wordListN != wlNew) {
410 wordListN->Set(wl);
411 firstModification = 0;
412 if (n == 4) {
413 // Rebuild preprocessorDefinitions
414 preprocessorDefinitionsStart.clear();
415 for (int nDefinition = 0; nDefinition < ppDefinitions.len; nDefinition++) {
416 char *cpDefinition = ppDefinitions.words[nDefinition];
417 char *cpEquals = strchr(cpDefinition, '=');
418 if (cpEquals) {
419 std::string name(cpDefinition, cpEquals - cpDefinition);
420 std::string val(cpEquals+1);
421 preprocessorDefinitionsStart[name] = val;
422 } else {
423 std::string name(cpDefinition);
424 std::string val("1");
425 preprocessorDefinitionsStart[name] = val;
431 return firstModification;
434 // Functor used to truncate history
435 struct After {
436 int line;
437 After(int line_) : line(line_) {}
438 bool operator()(PPDefinition &p) const {
439 return p.line > line;
443 void SCI_METHOD LexerCPP::Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
444 LexAccessor styler(pAccess);
446 CharacterSet setOKBeforeRE(CharacterSet::setNone, "([{=,:;!%^&*|?~+-");
447 CharacterSet setCouldBePostOp(CharacterSet::setNone, "+-");
449 CharacterSet setDoxygen(CharacterSet::setAlpha, "$@\\&<>#{}[]");
451 CharacterSet setWordStart(CharacterSet::setAlpha, "_", 0x80, true);
453 if (options.identifiersAllowDollars) {
454 setWordStart.Add('$');
455 setWord.Add('$');
458 int chPrevNonWhite = ' ';
459 int visibleChars = 0;
460 bool lastWordWasUUID = false;
461 int styleBeforeDCKeyword = SCE_C_DEFAULT;
462 bool continuationLine = false;
463 bool isIncludePreprocessor = false;
465 int lineCurrent = styler.GetLine(startPos);
466 if ((initStyle == SCE_C_PREPROCESSOR) ||
467 (initStyle == SCE_C_COMMENTLINE) ||
468 (initStyle == SCE_C_COMMENTLINEDOC)) {
469 // Set continuationLine if last character of previous line is '\'
470 if (lineCurrent > 0) {
471 int chBack = styler.SafeGetCharAt(startPos-1, 0);
472 int chBack2 = styler.SafeGetCharAt(startPos-2, 0);
473 int lineEndChar = '!';
474 if (chBack2 == '\r' && chBack == '\n') {
475 lineEndChar = styler.SafeGetCharAt(startPos-3, 0);
476 } else if (chBack == '\n' || chBack == '\r') {
477 lineEndChar = chBack2;
479 continuationLine = lineEndChar == '\\';
483 // look back to set chPrevNonWhite properly for better regex colouring
484 if (startPos > 0) {
485 int back = startPos;
486 while (--back && IsSpaceEquiv(styler.StyleAt(back)))
488 if (styler.StyleAt(back) == SCE_C_OPERATOR) {
489 chPrevNonWhite = styler.SafeGetCharAt(back);
493 StyleContext sc(startPos, length, initStyle, styler, 0x7f);
494 LinePPState preproc = vlls.ForLine(lineCurrent);
496 bool definitionsChanged = false;
498 // Truncate ppDefineHistory before current line
500 if (!options.updatePreprocessor)
501 ppDefineHistory.clear();
503 std::vector<PPDefinition>::iterator itInvalid = std::find_if(ppDefineHistory.begin(), ppDefineHistory.end(), After(lineCurrent-1));
504 if (itInvalid != ppDefineHistory.end()) {
505 ppDefineHistory.erase(itInvalid, ppDefineHistory.end());
506 definitionsChanged = true;
509 std::map<std::string, std::string> preprocessorDefinitions = preprocessorDefinitionsStart;
510 for (std::vector<PPDefinition>::iterator itDef = ppDefineHistory.begin(); itDef != ppDefineHistory.end(); ++itDef) {
511 preprocessorDefinitions[itDef->key] = itDef->value;
514 std::string rawStringTerminator = rawStringTerminators.ValueAt(lineCurrent-1);
515 SparseState<std::string> rawSTNew(lineCurrent);
517 int activitySet = preproc.IsInactive() ? activeFlag : 0;
519 for (; sc.More(); sc.Forward()) {
521 if (sc.atLineStart) {
522 if ((sc.state == SCE_C_STRING) || (sc.state == SCE_C_CHARACTER)) {
523 // Prevent SCE_C_STRINGEOL from leaking back to previous line which
524 // ends with a line continuation by locking in the state upto this position.
525 sc.SetState(sc.state);
527 // Reset states to begining of colourise so no surprises
528 // if different sets of lines lexed.
529 visibleChars = 0;
530 lastWordWasUUID = false;
531 isIncludePreprocessor = false;
532 if (preproc.IsInactive()) {
533 activitySet = activeFlag;
534 sc.SetState(sc.state | activitySet);
536 if (activitySet) {
537 if (sc.ch == '#') {
538 if (sc.Match("#else") || sc.Match("#end") || sc.Match("#if")) {
539 //activitySet = 0;
545 if (sc.atLineEnd) {
546 lineCurrent++;
547 vlls.Add(lineCurrent, preproc);
548 if (rawStringTerminator != "") {
549 rawSTNew.Set(lineCurrent-1, rawStringTerminator);
553 // Handle line continuation generically.
554 if (sc.ch == '\\') {
555 if (sc.chNext == '\n' || sc.chNext == '\r') {
556 sc.Forward();
557 if (sc.ch == '\r' && sc.chNext == '\n') {
558 sc.Forward();
560 continuationLine = true;
561 continue;
565 const bool atLineEndBeforeSwitch = sc.atLineEnd;
567 // Determine if the current state should terminate.
568 switch (MaskActive(sc.state)) {
569 case SCE_C_OPERATOR:
570 sc.SetState(SCE_C_DEFAULT|activitySet);
571 break;
572 case SCE_C_NUMBER:
573 // We accept almost anything because of hex. and number suffixes
574 if (!(setWord.Contains(sc.ch) || ((sc.ch == '+' || sc.ch == '-') && (sc.chPrev == 'e' || sc.chPrev == 'E')))) {
575 sc.SetState(SCE_C_DEFAULT|activitySet);
577 break;
578 case SCE_C_IDENTIFIER:
579 if (!setWord.Contains(sc.ch) || (sc.ch == '.')) {
580 char s[1000];
581 if (caseSensitive) {
582 sc.GetCurrent(s, sizeof(s));
583 } else {
584 sc.GetCurrentLowered(s, sizeof(s));
586 if (keywords.InList(s)) {
587 lastWordWasUUID = strcmp(s, "uuid") == 0;
588 sc.ChangeState(SCE_C_WORD|activitySet);
589 } else if (keywords2.InList(s)) {
590 sc.ChangeState(SCE_C_WORD2|activitySet);
591 } else if (keywords4.InList(s)) {
592 sc.ChangeState(SCE_C_GLOBALCLASS|activitySet);
594 const bool literalString = sc.ch == '\"';
595 if (literalString || sc.ch == '\'') {
596 size_t lenS = strlen(s);
597 const bool raw = literalString && sc.chPrev == 'R';
598 if (raw)
599 s[lenS--] = '\0';
600 bool valid =
601 (lenS == 0) ||
602 ((lenS == 1) && ((s[0] == 'L') || (s[0] == 'u') || (s[0] == 'U'))) ||
603 ((lenS == 2) && literalString && (s[0] == 'u') && (s[1] == '8'));
604 if (valid) {
605 if (literalString)
606 sc.ChangeState((raw ? SCE_C_STRINGRAW : SCE_C_STRING)|activitySet);
607 else
608 sc.ChangeState(SCE_C_CHARACTER|activitySet);
611 sc.SetState(SCE_C_DEFAULT|activitySet);
613 break;
614 case SCE_C_PREPROCESSOR:
615 if (sc.atLineStart && !continuationLine) {
616 sc.SetState(SCE_C_DEFAULT|activitySet);
617 } else if (options.stylingWithinPreprocessor) {
618 if (IsASpace(sc.ch)) {
619 sc.SetState(SCE_C_DEFAULT|activitySet);
621 } else {
622 if (sc.Match('/', '*') || sc.Match('/', '/')) {
623 sc.SetState(SCE_C_DEFAULT|activitySet);
626 break;
627 case SCE_C_COMMENT:
628 if (sc.Match('*', '/')) {
629 sc.Forward();
630 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
632 break;
633 case SCE_C_COMMENTDOC:
634 if (sc.Match('*', '/')) {
635 sc.Forward();
636 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
637 } else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
638 // Verify that we have the conditions to mark a comment-doc-keyword
639 if ((IsASpace(sc.chPrev) || sc.chPrev == '*') && (!IsASpace(sc.chNext))) {
640 styleBeforeDCKeyword = SCE_C_COMMENTDOC;
641 sc.SetState(SCE_C_COMMENTDOCKEYWORD|activitySet);
644 break;
645 case SCE_C_COMMENTLINE:
646 if (sc.atLineStart && !continuationLine) {
647 sc.SetState(SCE_C_DEFAULT|activitySet);
649 break;
650 case SCE_C_COMMENTLINEDOC:
651 if (sc.atLineStart && !continuationLine) {
652 sc.SetState(SCE_C_DEFAULT|activitySet);
653 } else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
654 // Verify that we have the conditions to mark a comment-doc-keyword
655 if ((IsASpace(sc.chPrev) || sc.chPrev == '/' || sc.chPrev == '!') && (!IsASpace(sc.chNext))) {
656 styleBeforeDCKeyword = SCE_C_COMMENTLINEDOC;
657 sc.SetState(SCE_C_COMMENTDOCKEYWORD|activitySet);
660 break;
661 case SCE_C_COMMENTDOCKEYWORD:
662 if ((styleBeforeDCKeyword == SCE_C_COMMENTDOC) && sc.Match('*', '/')) {
663 sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR);
664 sc.Forward();
665 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
666 } else if (!setDoxygen.Contains(sc.ch)) {
667 char s[100];
668 if (caseSensitive) {
669 sc.GetCurrent(s, sizeof(s));
670 } else {
671 sc.GetCurrentLowered(s, sizeof(s));
673 if (!IsASpace(sc.ch) || !keywords3.InList(s + 1)) {
674 sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR|activitySet);
676 sc.SetState(styleBeforeDCKeyword);
678 break;
679 case SCE_C_STRING:
680 if (sc.atLineEnd) {
681 sc.ChangeState(SCE_C_STRINGEOL|activitySet);
682 } else if (isIncludePreprocessor) {
683 if (sc.ch == '>') {
684 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
685 isIncludePreprocessor = false;
687 } else if (sc.ch == '\\') {
688 if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
689 sc.Forward();
691 } else if (sc.ch == '\"') {
692 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
694 break;
695 case SCE_C_STRINGRAW:
696 if (sc.Match(rawStringTerminator.c_str())) {
697 for (size_t termPos=rawStringTerminator.size(); termPos; termPos--)
698 sc.Forward();
699 sc.SetState(SCE_C_DEFAULT|activitySet);
700 rawStringTerminator = "";
702 break;
703 case SCE_C_CHARACTER:
704 if (sc.atLineEnd) {
705 sc.ChangeState(SCE_C_STRINGEOL|activitySet);
706 } else if (sc.ch == '\\') {
707 if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
708 sc.Forward();
710 } else if (sc.ch == '\'') {
711 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
713 break;
714 case SCE_C_REGEX:
715 if (sc.atLineStart) {
716 sc.SetState(SCE_C_DEFAULT|activitySet);
717 } else if (sc.ch == '/') {
718 sc.Forward();
719 while ((sc.ch < 0x80) && islower(sc.ch))
720 sc.Forward(); // gobble regex flags
721 sc.SetState(SCE_C_DEFAULT|activitySet);
722 } else if (sc.ch == '\\') {
723 // Gobble up the quoted character
724 if (sc.chNext == '\\' || sc.chNext == '/') {
725 sc.Forward();
728 break;
729 case SCE_C_STRINGEOL:
730 if (sc.atLineStart) {
731 sc.SetState(SCE_C_DEFAULT|activitySet);
733 break;
734 case SCE_C_VERBATIM:
735 if (sc.ch == '\"') {
736 if (sc.chNext == '\"') {
737 sc.Forward();
738 } else {
739 sc.ForwardSetState(SCE_C_DEFAULT|activitySet);
742 break;
743 case SCE_C_TRIPLEVERBATIM:
744 if (sc.Match("\"\"\"")) {
745 while (sc.Match('"')) {
746 sc.Forward();
748 sc.SetState(SCE_C_DEFAULT|activitySet);
750 break;
751 case SCE_C_UUID:
752 if (sc.ch == '\r' || sc.ch == '\n' || sc.ch == ')') {
753 sc.SetState(SCE_C_DEFAULT|activitySet);
757 if (sc.atLineEnd && !atLineEndBeforeSwitch) {
758 // State exit processing consumed characters up to end of line.
759 lineCurrent++;
760 vlls.Add(lineCurrent, preproc);
763 // Determine if a new state should be entered.
764 if (MaskActive(sc.state) == SCE_C_DEFAULT) {
765 if (sc.Match('@', '\"')) {
766 sc.SetState(SCE_C_VERBATIM|activitySet);
767 sc.Forward();
768 } else if (options.triplequotedStrings && sc.Match("\"\"\"")) {
769 sc.SetState(SCE_C_TRIPLEVERBATIM|activitySet);
770 sc.Forward(2);
771 } else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
772 if (lastWordWasUUID) {
773 sc.SetState(SCE_C_UUID|activitySet);
774 lastWordWasUUID = false;
775 } else {
776 sc.SetState(SCE_C_NUMBER|activitySet);
778 } else if (setWordStart.Contains(sc.ch) || (sc.ch == '@')) {
779 if (lastWordWasUUID) {
780 sc.SetState(SCE_C_UUID|activitySet);
781 lastWordWasUUID = false;
782 } else {
783 sc.SetState(SCE_C_IDENTIFIER|activitySet);
785 } else if (sc.Match('/', '*')) {
786 if (sc.Match("/**") || sc.Match("/*!")) { // Support of Qt/Doxygen doc. style
787 sc.SetState(SCE_C_COMMENTDOC|activitySet);
788 } else {
789 sc.SetState(SCE_C_COMMENT|activitySet);
791 sc.Forward(); // Eat the * so it isn't used for the end of the comment
792 } else if (sc.Match('/', '/')) {
793 if ((sc.Match("///") && !sc.Match("////")) || sc.Match("//!"))
794 // Support of Qt/Doxygen doc. style
795 sc.SetState(SCE_C_COMMENTLINEDOC|activitySet);
796 else
797 sc.SetState(SCE_C_COMMENTLINE|activitySet);
798 } else if (sc.ch == '/'
799 && (setOKBeforeRE.Contains(chPrevNonWhite)
800 || followsReturnKeyword(sc, styler))
801 && (!setCouldBePostOp.Contains(chPrevNonWhite)
802 || !FollowsPostfixOperator(sc, styler))) {
803 sc.SetState(SCE_C_REGEX|activitySet); // JavaScript's RegEx
804 } else if (sc.ch == '\"') {
805 if (sc.chPrev == 'R') {
806 sc.SetState(SCE_C_STRINGRAW|activitySet);
807 rawStringTerminator = ")";
808 for (int termPos = sc.currentPos + 1;; termPos++) {
809 char chTerminator = styler.SafeGetCharAt(termPos, '(');
810 if (chTerminator == '(')
811 break;
812 rawStringTerminator += chTerminator;
814 rawStringTerminator += '\"';
815 } else {
816 sc.SetState(SCE_C_STRING|activitySet);
818 isIncludePreprocessor = false; // ensure that '>' won't end the string
819 } else if (isIncludePreprocessor && sc.ch == '<') {
820 sc.SetState(SCE_C_STRING|activitySet);
821 } else if (sc.ch == '\'') {
822 sc.SetState(SCE_C_CHARACTER|activitySet);
823 } else if (sc.ch == '#' && visibleChars == 0) {
824 // Preprocessor commands are alone on their line
825 sc.SetState(SCE_C_PREPROCESSOR|activitySet);
826 // Skip whitespace between # and preprocessor word
827 do {
828 sc.Forward();
829 } while ((sc.ch == ' ' || sc.ch == '\t') && sc.More());
830 if (sc.atLineEnd) {
831 sc.SetState(SCE_C_DEFAULT|activitySet);
832 } else if (sc.Match("include")) {
833 isIncludePreprocessor = true;
834 } else {
835 if (options.trackPreprocessor) {
836 if (sc.Match("ifdef") || sc.Match("ifndef")) {
837 bool isIfDef = sc.Match("ifdef");
838 int i = isIfDef ? 5 : 6;
839 std::string restOfLine = GetRestOfLine(styler, sc.currentPos + i + 1, false);
840 bool foundDef = preprocessorDefinitions.find(restOfLine) != preprocessorDefinitions.end();
841 preproc.StartSection(isIfDef == foundDef);
842 } else if (sc.Match("if")) {
843 std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 2, true);
844 bool ifGood = EvaluateExpression(restOfLine, preprocessorDefinitions);
845 preproc.StartSection(ifGood);
846 } else if (sc.Match("else")) {
847 if (!preproc.CurrentIfTaken()) {
848 preproc.InvertCurrentLevel();
849 activitySet = preproc.IsInactive() ? activeFlag : 0;
850 if (!activitySet)
851 sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
852 } else if (!preproc.IsInactive()) {
853 preproc.InvertCurrentLevel();
854 activitySet = preproc.IsInactive() ? activeFlag : 0;
855 if (!activitySet)
856 sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
858 } else if (sc.Match("elif")) {
859 // Ensure only one chosen out of #if .. #elif .. #elif .. #else .. #endif
860 if (!preproc.CurrentIfTaken()) {
861 // Similar to #if
862 std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 2, true);
863 bool ifGood = EvaluateExpression(restOfLine, preprocessorDefinitions);
864 if (ifGood) {
865 preproc.InvertCurrentLevel();
866 activitySet = preproc.IsInactive() ? activeFlag : 0;
867 if (!activitySet)
868 sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
870 } else if (!preproc.IsInactive()) {
871 preproc.InvertCurrentLevel();
872 activitySet = preproc.IsInactive() ? activeFlag : 0;
873 if (!activitySet)
874 sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
876 } else if (sc.Match("endif")) {
877 preproc.EndSection();
878 activitySet = preproc.IsInactive() ? activeFlag : 0;
879 sc.ChangeState(SCE_C_PREPROCESSOR|activitySet);
880 } else if (sc.Match("define")) {
881 if (options.updatePreprocessor && !preproc.IsInactive()) {
882 std::string restOfLine = GetRestOfLine(styler, sc.currentPos + 6, true);
883 if (restOfLine.find(")") == std::string::npos) { // Don't handle macros with arguments
884 std::vector<std::string> tokens = Tokenize(restOfLine);
885 std::string key;
886 std::string value("1");
887 if (tokens.size() >= 1) {
888 key = tokens[0];
889 if (tokens.size() >= 2) {
890 value = tokens[1];
892 preprocessorDefinitions[key] = value;
893 ppDefineHistory.push_back(PPDefinition(lineCurrent, key, value));
894 definitionsChanged = true;
901 } else if (isoperator(static_cast<char>(sc.ch))) {
902 sc.SetState(SCE_C_OPERATOR|activitySet);
906 if (!IsASpace(sc.ch) && !IsSpaceEquiv(sc.state)) {
907 chPrevNonWhite = sc.ch;
908 visibleChars++;
910 continuationLine = false;
912 const bool rawStringsChanged = rawStringTerminators.Merge(rawSTNew, lineCurrent);
913 if (definitionsChanged || rawStringsChanged)
914 styler.ChangeLexerState(startPos, startPos + length);
915 sc.Complete();
918 // Store both the current line's fold level and the next lines in the
919 // level store to make it easy to pick up with each increment
920 // and to make it possible to fiddle the current level for "} else {".
922 void SCI_METHOD LexerCPP::Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
924 if (!options.fold)
925 return;
927 LexAccessor styler(pAccess);
929 unsigned int endPos = startPos + length;
930 int visibleChars = 0;
931 int lineCurrent = styler.GetLine(startPos);
932 int levelCurrent = SC_FOLDLEVELBASE;
933 if (lineCurrent > 0)
934 levelCurrent = styler.LevelAt(lineCurrent-1) >> 16;
935 int levelMinCurrent = levelCurrent;
936 int levelNext = levelCurrent;
937 char chNext = styler[startPos];
938 int styleNext = MaskActive(styler.StyleAt(startPos));
939 int style = MaskActive(initStyle);
940 const bool userDefinedFoldMarkers = !options.foldExplicitStart.empty() && !options.foldExplicitEnd.empty();
941 for (unsigned int i = startPos; i < endPos; i++) {
942 char ch = chNext;
943 chNext = styler.SafeGetCharAt(i + 1);
944 int stylePrev = style;
945 style = styleNext;
946 styleNext = MaskActive(styler.StyleAt(i + 1));
947 bool atEOL = (ch == '\r' && chNext != '\n') || (ch == '\n');
948 if (options.foldComment && options.foldCommentMultiline && IsStreamCommentStyle(style)) {
949 if (!IsStreamCommentStyle(stylePrev) && (stylePrev != SCE_C_COMMENTLINEDOC)) {
950 levelNext++;
951 } else if (!IsStreamCommentStyle(styleNext) && (styleNext != SCE_C_COMMENTLINEDOC) && !atEOL) {
952 // Comments don't end at end of line and the next character may be unstyled.
953 levelNext--;
956 if (options.foldComment && options.foldCommentExplicit && ((style == SCE_C_COMMENTLINE) || options.foldExplicitAnywhere)) {
957 if (userDefinedFoldMarkers) {
958 if (styler.Match(i, options.foldExplicitStart.c_str())) {
959 levelNext++;
960 } else if (styler.Match(i, options.foldExplicitEnd.c_str())) {
961 levelNext--;
963 } else {
964 if ((ch == '/') && (chNext == '/')) {
965 char chNext2 = styler.SafeGetCharAt(i + 2);
966 if (chNext2 == '{') {
967 levelNext++;
968 } else if (chNext2 == '}') {
969 levelNext--;
974 if (options.foldPreprocessor && (style == SCE_C_PREPROCESSOR)) {
975 if (ch == '#') {
976 unsigned int j = i + 1;
977 while ((j < endPos) && IsASpaceOrTab(styler.SafeGetCharAt(j))) {
978 j++;
980 if (styler.Match(j, "region") || styler.Match(j, "if")) {
981 levelNext++;
982 } else if (styler.Match(j, "end")) {
983 levelNext--;
987 if (options.foldSyntaxBased && (style == SCE_C_OPERATOR)) {
988 if (ch == '{') {
989 // Measure the minimum before a '{' to allow
990 // folding on "} else {"
991 if (levelMinCurrent > levelNext) {
992 levelMinCurrent = levelNext;
994 levelNext++;
995 } else if (ch == '}') {
996 levelNext--;
999 if (!IsASpace(ch))
1000 visibleChars++;
1001 if (atEOL || (i == endPos-1)) {
1002 int levelUse = levelCurrent;
1003 if (options.foldSyntaxBased && options.foldAtElse) {
1004 levelUse = levelMinCurrent;
1006 int lev = levelUse | levelNext << 16;
1007 if (visibleChars == 0 && options.foldCompact)
1008 lev |= SC_FOLDLEVELWHITEFLAG;
1009 if (levelUse < levelNext)
1010 lev |= SC_FOLDLEVELHEADERFLAG;
1011 if (lev != styler.LevelAt(lineCurrent)) {
1012 styler.SetLevel(lineCurrent, lev);
1014 lineCurrent++;
1015 levelCurrent = levelNext;
1016 levelMinCurrent = levelCurrent;
1017 if (atEOL && (i == static_cast<unsigned int>(styler.Length()-1))) {
1018 // There is an empty line at end of file so give it same level and empty
1019 styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG);
1021 visibleChars = 0;
1026 void LexerCPP::EvaluateTokens(std::vector<std::string> &tokens) {
1028 // Evaluate defined() statements to either 0 or 1
1029 for (size_t i=0; (i+2)<tokens.size();) {
1030 if ((tokens[i] == "defined") && (tokens[i+1] == "(")) {
1031 const char *val = "0";
1032 if (tokens[i+2] == ")") {
1033 // defined()
1034 tokens.erase(tokens.begin() + i + 1, tokens.begin() + i + 3);
1035 } else if (((i+2)<tokens.size()) && (tokens[i+3] == ")")) {
1036 // defined(<int>)
1037 tokens.erase(tokens.begin() + i + 1, tokens.begin() + i + 4);
1038 val = "1";
1040 tokens[i] = val;
1041 } else {
1042 i++;
1046 // Find bracketed subexpressions and recurse on them
1047 std::vector<std::string>::iterator itBracket = std::find(tokens.begin(), tokens.end(), "(");
1048 std::vector<std::string>::iterator itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1049 while ((itBracket != tokens.end()) && (itEndBracket != tokens.end()) && (itEndBracket > itBracket)) {
1050 std::vector<std::string> inBracket(itBracket + 1, itEndBracket);
1051 EvaluateTokens(inBracket);
1053 // The insertion is done before the removal because there were failures with the opposite approach
1054 tokens.insert(itBracket, inBracket.begin(), inBracket.end());
1055 itBracket = std::find(tokens.begin(), tokens.end(), "(");
1056 itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1057 tokens.erase(itBracket, itEndBracket + 1);
1059 itBracket = std::find(tokens.begin(), tokens.end(), "(");
1060 itEndBracket = std::find(tokens.begin(), tokens.end(), ")");
1063 // Evaluate logical negations
1064 for (size_t j=0; (j+1)<tokens.size();) {
1065 if (setNegationOp.Contains(tokens[j][0])) {
1066 int isTrue = atoi(tokens[j+1].c_str());
1067 if (tokens[j] == "!")
1068 isTrue = !isTrue;
1069 std::vector<std::string>::iterator itInsert =
1070 tokens.erase(tokens.begin() + j, tokens.begin() + j + 2);
1071 tokens.insert(itInsert, isTrue ? "1" : "0");
1072 } else {
1073 j++;
1077 // Evaluate expressions in precedence order
1078 enum precedence { precArithmetic, precRelative, precLogical };
1079 for (int prec=precArithmetic; prec <= precLogical; prec++) {
1080 // Looking at 3 tokens at a time so end at 2 before end
1081 for (size_t k=0; (k+2)<tokens.size();) {
1082 char chOp = tokens[k+1][0];
1083 if (
1084 ((prec==precArithmetic) && setArithmethicOp.Contains(chOp)) ||
1085 ((prec==precRelative) && setRelOp.Contains(chOp)) ||
1086 ((prec==precLogical) && setLogicalOp.Contains(chOp))
1088 int valA = atoi(tokens[k].c_str());
1089 int valB = atoi(tokens[k+2].c_str());
1090 int result = 0;
1091 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;
1097 else if (tokens[k+1] == "/")
1098 result = valA / (valB ? valB : 1);
1099 else if (tokens[k+1] == "%")
1100 result = valA % (valB ? valB : 1);
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;
1115 else if (tokens[k+1] == "&&")
1116 result = valA && valB;
1117 char sResult[30];
1118 sprintf(sResult, "%d", result);
1119 std::vector<std::string>::iterator itInsert =
1120 tokens.erase(tokens.begin() + k, tokens.begin() + k + 3);
1121 tokens.insert(itInsert, sResult);
1122 } else {
1123 k++;
1129 bool LexerCPP::EvaluateExpression(const std::string &expr, const std::map<std::string, std::string> &preprocessorDefinitions) {
1130 // Break into tokens, replacing with definitions
1131 std::string word;
1132 std::vector<std::string> tokens;
1133 const char *cp = expr.c_str();
1134 for (;;) {
1135 if (setWord.Contains(*cp)) {
1136 word += *cp;
1137 } else {
1138 std::map<std::string, std::string>::const_iterator it = preprocessorDefinitions.find(word);
1139 if (it != preprocessorDefinitions.end()) {
1140 tokens.push_back(it->second);
1141 } else if (!word.empty() && ((word[0] >= '0' && word[0] <= '9') || (word == "defined"))) {
1142 tokens.push_back(word);
1144 word = "";
1145 if (!*cp) {
1146 break;
1148 if ((*cp != ' ') && (*cp != '\t')) {
1149 std::string op(cp, 1);
1150 if (setRelOp.Contains(*cp)) {
1151 if (setRelOp.Contains(cp[1])) {
1152 op += cp[1];
1153 cp++;
1155 } else if (setLogicalOp.Contains(*cp)) {
1156 if (setLogicalOp.Contains(cp[1])) {
1157 op += cp[1];
1158 cp++;
1161 tokens.push_back(op);
1164 cp++;
1167 EvaluateTokens(tokens);
1169 // "0" or "" -> false else true
1170 bool isFalse = tokens.empty() ||
1171 ((tokens.size() == 1) && ((tokens[0] == "") || tokens[0] == "0"));
1172 return !isFalse;
1175 LexerModule lmCPP(SCLEX_CPP, LexerCPP::LexerFactoryCPP, "cpp", cppWordLists);
1176 LexerModule lmCPPNoCase(SCLEX_CPPNOCASE, LexerCPP::LexerFactoryCPPInsensitive, "cppnocase", cppWordLists);