1 // Scintilla source code edit control
2 /** @file LexCPP.cxx
3 ** Lexer for C++, C, Java, and JavaScript.
4 **/
5 // Copyright 1998-2005 by Neil Hodgson <neilh@scintilla.org>
6 // The License.txt file describes the conditions under which this software may be distributed.
7
8 #include <stdlib.h>
9 #include <string.h>
10 #include <ctype.h>
11 #include <stdio.h>
12 #include <stdarg.h>
13
14 #include "Platform.h"
15
16 #include "PropSet.h"
17 #include "Accessor.h"
18 #include "StyleContext.h"
19 #include "KeyWords.h"
20 #include "Scintilla.h"
21 #include "SciLexer.h"
22
23 #define KEYWORD_BOXHEADER 1
24 #define KEYWORD_FOLDCONTRACTED 2
25
IsOKBeforeRE(int ch)26 static bool IsOKBeforeRE(int ch) {
27 return (ch == '(') || (ch == '=') || (ch == ',');
28 }
29
IsAWordChar(int ch)30 static inline bool IsAWordChar(int ch) {
31 return (ch < 0x80) && (isalnum(ch) || ch == '.' || ch == '_');
32 }
33
IsAWordStart(int ch)34 static inline bool IsAWordStart(int ch) {
35 return (ch < 0x80) && (isalpha(ch) || ch == '_');
36 }
37
IsADoxygenChar(int ch)38 static inline bool IsADoxygenChar(int ch) {
39 return (ch < 0x80 && islower(ch)) || ch == '$' || ch == '@' ||
40 ch == '\\' || ch == '&' || ch == '<' ||
41 ch == '>' || ch == '#' || ch == '{' ||
42 ch == '}' || ch == '[' || ch == ']';
43 }
44
IsSpaceEquiv(int state)45 static bool IsSpaceEquiv(int state) {
46 return (state <= SCE_C_COMMENTDOC) ||
47 // including SCE_C_DEFAULT, SCE_C_COMMENT, SCE_C_COMMENTLINE
48 (state == SCE_C_COMMENTLINEDOC) || (state == SCE_C_COMMENTDOCKEYWORD) ||
49 (state == SCE_C_COMMENTDOCKEYWORDERROR);
50 }
51
ColouriseCppDoc(unsigned int startPos,int length,int initStyle,WordList * keywordlists[],Accessor & styler,bool caseSensitive)52 static void ColouriseCppDoc(unsigned int startPos, int length, int initStyle, WordList *keywordlists[],
53 Accessor &styler, bool caseSensitive) {
54
55 WordList &keywords = *keywordlists[0];
56 WordList &keywords2 = *keywordlists[1];
57 WordList &keywords3 = *keywordlists[2];
58 WordList &keywords4 = *keywordlists[3];
59
60 bool stylingWithinPreprocessor = styler.GetPropertyInt("styling.within.preprocessor") != 0;
61
62 int chPrevNonWhite = ' ';
63 int visibleChars = 0;
64 bool lastWordWasUUID = false;
65 int styleBeforeDCKeyword = SCE_C_DEFAULT;
66 bool continuationLine = false;
67
68 if (initStyle == SCE_C_PREPROCESSOR) {
69 // Set continuationLine if last character of previous line is '\'
70 int lineCurrent = styler.GetLine(startPos);
71 if (lineCurrent > 0) {
72 int chBack = styler.SafeGetCharAt(startPos-1, 0);
73 int chBack2 = styler.SafeGetCharAt(startPos-2, 0);
74 int lineEndChar = '!';
75 if (chBack2 == '\r' && chBack == '\n') {
76 lineEndChar = styler.SafeGetCharAt(startPos-3, 0);
77 } else if (chBack == '\n' || chBack == '\r') {
78 lineEndChar = chBack2;
79 }
80 continuationLine = lineEndChar == '\\';
81 }
82 }
83
84 // look back to set chPrevNonWhite properly for better regex colouring
85 if (startPos > 0) {
86 int back = startPos;
87 while (--back && IsSpaceEquiv(styler.StyleAt(back)))
88 ;
89 if (styler.StyleAt(back) == SCE_C_OPERATOR) {
90 chPrevNonWhite = styler.SafeGetCharAt(back);
91 }
92 }
93
94 StyleContext sc(startPos, length, initStyle, styler);
95
96 for (; sc.More(); sc.Forward()) {
97
98 if (sc.atLineStart) {
99 if (sc.state == SCE_C_STRING) {
100 // Prevent SCE_C_STRINGEOL from leaking back to previous line which
101 // ends with a line continuation by locking in the state upto this position.
102 sc.SetState(SCE_C_STRING);
103 }
104 // Reset states to begining of colourise so no surprises
105 // if different sets of lines lexed.
106 visibleChars = 0;
107 lastWordWasUUID = false;
108 }
109
110 // Handle line continuation generically.
111 if (sc.ch == '\\') {
112 if (sc.chNext == '\n' || sc.chNext == '\r') {
113 sc.Forward();
114 if (sc.ch == '\r' && sc.chNext == '\n') {
115 sc.Forward();
116 }
117 continuationLine = true;
118 continue;
119 }
120 }
121
122 // Determine if the current state should terminate.
123 switch (sc.state) {
124 case SCE_C_OPERATOR:
125 sc.SetState(SCE_C_DEFAULT);
126 break;
127 case SCE_C_NUMBER:
128 // We accept almost anything because of hex. and number suffixes
129 if (!IsAWordChar(sc.ch)) {
130 sc.SetState(SCE_C_DEFAULT);
131 }
132 break;
133 case SCE_C_IDENTIFIER:
134 if (!IsAWordChar(sc.ch) || (sc.ch == '.')) {
135 char s[1000];
136 if (caseSensitive) {
137 sc.GetCurrent(s, sizeof(s));
138 } else {
139 sc.GetCurrentLowered(s, sizeof(s));
140 }
141 if (keywords.InList(s)) {
142 lastWordWasUUID = strcmp(s, "uuid") == 0;
143 sc.ChangeState(SCE_C_WORD);
144 } else if (keywords2.InList(s)) {
145 sc.ChangeState(SCE_C_WORD2);
146 } else if (keywords4.InList(s)) {
147 sc.ChangeState(SCE_C_GLOBALCLASS);
148 }
149 sc.SetState(SCE_C_DEFAULT);
150 }
151 break;
152 case SCE_C_PREPROCESSOR:
153 if (sc.atLineStart && !continuationLine) {
154 sc.SetState(SCE_C_DEFAULT);
155 } else if (stylingWithinPreprocessor) {
156 if (IsASpace(sc.ch)) {
157 sc.SetState(SCE_C_DEFAULT);
158 }
159 } else {
160 if (sc.Match('/', '*') || sc.Match('/', '/')) {
161 sc.SetState(SCE_C_DEFAULT);
162 }
163 }
164 break;
165 case SCE_C_COMMENT:
166 if (sc.Match('*', '/')) {
167 sc.Forward();
168 sc.ForwardSetState(SCE_C_DEFAULT);
169 }
170 break;
171 case SCE_C_COMMENTDOC:
172 if (sc.Match('*', '/')) {
173 sc.Forward();
174 sc.ForwardSetState(SCE_C_DEFAULT);
175 } else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
176 // Verify that we have the conditions to mark a comment-doc-keyword
177 if ((IsASpace(sc.chPrev) || sc.chPrev == '*') && (!IsASpace(sc.chNext))) {
178 styleBeforeDCKeyword = SCE_C_COMMENTDOC;
179 sc.SetState(SCE_C_COMMENTDOCKEYWORD);
180 }
181 }
182 break;
183 case SCE_C_COMMENTLINE:
184 if (sc.atLineStart) {
185 sc.SetState(SCE_C_DEFAULT);
186 }
187 break;
188 case SCE_C_COMMENTLINEDOC:
189 if (sc.atLineStart) {
190 sc.SetState(SCE_C_DEFAULT);
191 } else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
192 // Verify that we have the conditions to mark a comment-doc-keyword
193 if ((IsASpace(sc.chPrev) || sc.chPrev == '/' || sc.chPrev == '!') && (!IsASpace(sc.chNext))) {
194 styleBeforeDCKeyword = SCE_C_COMMENTLINEDOC;
195 sc.SetState(SCE_C_COMMENTDOCKEYWORD);
196 }
197 }
198 break;
199 case SCE_C_COMMENTDOCKEYWORD:
200 if ((styleBeforeDCKeyword == SCE_C_COMMENTDOC) && sc.Match('*', '/')) {
201 sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR);
202 sc.Forward();
203 sc.ForwardSetState(SCE_C_DEFAULT);
204 } else if (!IsADoxygenChar(sc.ch)) {
205 char s[100];
206 if (caseSensitive) {
207 sc.GetCurrent(s, sizeof(s));
208 } else {
209 sc.GetCurrentLowered(s, sizeof(s));
210 }
211 if (!isspace(sc.ch) || !keywords3.InList(s + 1)) {
212 sc.ChangeState(SCE_C_COMMENTDOCKEYWORDERROR);
213 }
214 sc.SetState(styleBeforeDCKeyword);
215 }
216 break;
217 case SCE_C_STRING:
218 if (sc.atLineEnd) {
219 sc.ChangeState(SCE_C_STRINGEOL);
220 } else if (sc.ch == '\\') {
221 if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
222 sc.Forward();
223 }
224 } else if (sc.ch == '\"') {
225 sc.ForwardSetState(SCE_C_DEFAULT);
226 }
227 break;
228 case SCE_C_CHARACTER:
229 if (sc.atLineEnd) {
230 sc.ChangeState(SCE_C_STRINGEOL);
231 } else if (sc.ch == '\\') {
232 if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
233 sc.Forward();
234 }
235 } else if (sc.ch == '\'') {
236 sc.ForwardSetState(SCE_C_DEFAULT);
237 }
238 break;
239 case SCE_C_REGEX:
240 if (sc.atLineStart) {
241 sc.SetState(SCE_C_DEFAULT);
242 } else if (sc.ch == '/') {
243 sc.Forward();
244 while ((sc.ch < 0x80) && islower(sc.ch))
245 sc.Forward(); // gobble regex flags
246 sc.SetState(SCE_C_DEFAULT);
247 } else if (sc.ch == '\\') {
248 // Gobble up the quoted character
249 if (sc.chNext == '\\' || sc.chNext == '/') {
250 sc.Forward();
251 }
252 }
253 break;
254 case SCE_C_STRINGEOL:
255 if (sc.atLineStart) {
256 sc.SetState(SCE_C_DEFAULT);
257 }
258 break;
259 case SCE_C_VERBATIM:
260 if (sc.ch == '\"') {
261 if (sc.chNext == '\"') {
262 sc.Forward();
263 } else {
264 sc.ForwardSetState(SCE_C_DEFAULT);
265 }
266 }
267 break;
268 case SCE_C_UUID:
269 if (sc.ch == '\r' || sc.ch == '\n' || sc.ch == ')') {
270 sc.SetState(SCE_C_DEFAULT);
271 }
272 }
273
274 // Determine if a new state should be entered.
275 if (sc.state == SCE_C_DEFAULT) {
276 if (sc.Match('@', '\"')) {
277 sc.SetState(SCE_C_VERBATIM);
278 sc.Forward();
279 } else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
280 if (lastWordWasUUID) {
281 sc.SetState(SCE_C_UUID);
282 lastWordWasUUID = false;
283 } else {
284 sc.SetState(SCE_C_NUMBER);
285 }
286 } else if (IsAWordStart(sc.ch) || (sc.ch == '@')) {
287 if (lastWordWasUUID) {
288 sc.SetState(SCE_C_UUID);
289 lastWordWasUUID = false;
290 } else {
291 sc.SetState(SCE_C_IDENTIFIER);
292 }
293 } else if (sc.Match('/', '*')) {
294 if (sc.Match("/**") || sc.Match("/*!")) { // Support of Qt/Doxygen doc. style
295 sc.SetState(SCE_C_COMMENTDOC);
296 } else {
297 sc.SetState(SCE_C_COMMENT);
298 }
299 sc.Forward(); // Eat the * so it isn't used for the end of the comment
300 } else if (sc.Match('/', '/')) {
301 if ((sc.Match("///") && !sc.Match("////")) || sc.Match("//!"))
302 // Support of Qt/Doxygen doc. style
303 sc.SetState(SCE_C_COMMENTLINEDOC);
304 else
305 sc.SetState(SCE_C_COMMENTLINE);
306 } else if (sc.ch == '/' && IsOKBeforeRE(chPrevNonWhite)) {
307 sc.SetState(SCE_C_REGEX); // JavaScript's RegEx
308 } else if (sc.ch == '\"') {
309 sc.SetState(SCE_C_STRING);
310 } else if (sc.ch == '\'') {
311 sc.SetState(SCE_C_CHARACTER);
312 } else if (sc.ch == '#' && visibleChars == 0) {
313 // Preprocessor commands are alone on their line
314 sc.SetState(SCE_C_PREPROCESSOR);
315 // Skip whitespace between # and preprocessor word
316 do {
317 sc.Forward();
318 } while ((sc.ch == ' ' || sc.ch == '\t') && sc.More());
319 if (sc.atLineEnd) {
320 sc.SetState(SCE_C_DEFAULT);
321 }
322 } else if (isoperator(static_cast<char>(sc.ch))) {
323 sc.SetState(SCE_C_OPERATOR);
324 }
325 }
326
327 if (!IsASpace(sc.ch) && !IsSpaceEquiv(sc.state)) {
328 chPrevNonWhite = sc.ch;
329 visibleChars++;
330 }
331 continuationLine = false;
332 }
333 sc.Complete();
334 }
335
IsStreamCommentStyle(int style)336 static bool IsStreamCommentStyle(int style) {
337 return style == SCE_C_COMMENT ||
338 style == SCE_C_COMMENTDOC ||
339 style == SCE_C_COMMENTDOCKEYWORD ||
340 style == SCE_C_COMMENTDOCKEYWORDERROR;
341 }
342
343 // Store both the current line's fold level and the next lines in the
344 // level store to make it easy to pick up with each increment
345 // and to make it possible to fiddle the current level for "} else {".
FoldNoBoxCppDoc(unsigned int startPos,int length,int initStyle,Accessor & styler)346 static void FoldNoBoxCppDoc(unsigned int startPos, int length, int initStyle,
347 Accessor &styler) {
348 bool foldComment = styler.GetPropertyInt("fold.comment") != 0;
349 bool foldPreprocessor = styler.GetPropertyInt("fold.preprocessor") != 0;
350 bool foldCompact = styler.GetPropertyInt("fold.compact", 1) != 0;
351 bool foldAtElse = styler.GetPropertyInt("fold.at.else", 0) != 0;
352 unsigned int endPos = startPos + length;
353 int visibleChars = 0;
354 int lineCurrent = styler.GetLine(startPos);
355 int levelCurrent = SC_FOLDLEVELBASE;
356 if (lineCurrent > 0)
357 levelCurrent = styler.LevelAt(lineCurrent-1) >> 16;
358 int levelMinCurrent = levelCurrent;
359 int levelNext = levelCurrent;
360 char chNext = styler[startPos];
361 int styleNext = styler.StyleAt(startPos);
362 int style = initStyle;
363 for (unsigned int i = startPos; i < endPos; i++) {
364 char ch = chNext;
365 chNext = styler.SafeGetCharAt(i + 1);
366 int stylePrev = style;
367 style = styleNext;
368 styleNext = styler.StyleAt(i + 1);
369 bool atEOL = (ch == '\r' && chNext != '\n') || (ch == '\n');
370 if (foldComment && IsStreamCommentStyle(style)) {
371 if (!IsStreamCommentStyle(stylePrev)) {
372 levelNext++;
373 } else if (!IsStreamCommentStyle(styleNext) && !atEOL) {
374 // Comments don't end at end of line and the next character may be unstyled.
375 levelNext--;
376 }
377 }
378 if (foldComment && (style == SCE_C_COMMENTLINE)) {
379 if ((ch == '/') && (chNext == '/')) {
380 char chNext2 = styler.SafeGetCharAt(i + 2);
381 if (chNext2 == '{') {
382 levelNext++;
383 } else if (chNext2 == '}') {
384 levelNext--;
385 }
386 }
387 }
388 if (foldPreprocessor && (style == SCE_C_PREPROCESSOR)) {
389 if (ch == '#') {
390 unsigned int j = i + 1;
391 while ((j < endPos) && IsASpaceOrTab(styler.SafeGetCharAt(j))) {
392 j++;
393 }
394 if (styler.Match(j, "region") || styler.Match(j, "if")) {
395 levelNext++;
396 } else if (styler.Match(j, "end")) {
397 levelNext--;
398 }
399 }
400 }
401 if (style == SCE_C_OPERATOR) {
402 if (ch == '{') {
403 // Measure the minimum before a '{' to allow
404 // folding on "} else {"
405 if (levelMinCurrent > levelNext) {
406 levelMinCurrent = levelNext;
407 }
408 levelNext++;
409 } else if (ch == '}') {
410 levelNext--;
411 }
412 }
413 if (atEOL) {
414 int levelUse = levelCurrent;
415 if (foldAtElse) {
416 levelUse = levelMinCurrent;
417 }
418 int lev = levelUse | levelNext << 16;
419 if (visibleChars == 0 && foldCompact)
420 lev |= SC_FOLDLEVELWHITEFLAG;
421 if (levelUse < levelNext)
422 lev |= SC_FOLDLEVELHEADERFLAG;
423 if (lev != styler.LevelAt(lineCurrent)) {
424 styler.SetLevel(lineCurrent, lev);
425 }
426 lineCurrent++;
427 levelCurrent = levelNext;
428 levelMinCurrent = levelCurrent;
429 visibleChars = 0;
430 }
431 if (!isspacechar(ch))
432 visibleChars++;
433 }
434 }
435
FoldCppDoc(unsigned int startPos,int length,int initStyle,WordList * [],Accessor & styler)436 static void FoldCppDoc(unsigned int startPos, int length, int initStyle, WordList *[],
437 Accessor &styler) {
438 FoldNoBoxCppDoc(startPos, length, initStyle, styler);
439 }
440
441 static const char * const cppWordLists[] = {
442 "Primary keywords and identifiers",
443 "Secondary keywords and identifiers",
444 "Documentation comment keywords",
445 "Unused",
446 "Global classes and typedefs",
447 0,
448 };
449
ColouriseCppDocSensitive(unsigned int startPos,int length,int initStyle,WordList * keywordlists[],Accessor & styler)450 static void ColouriseCppDocSensitive(unsigned int startPos, int length, int initStyle, WordList *keywordlists[],
451 Accessor &styler) {
452 ColouriseCppDoc(startPos, length, initStyle, keywordlists, styler, true);
453 }
454
ColouriseCppDocInsensitive(unsigned int startPos,int length,int initStyle,WordList * keywordlists[],Accessor & styler)455 static void ColouriseCppDocInsensitive(unsigned int startPos, int length, int initStyle, WordList *keywordlists[],
456 Accessor &styler) {
457 ColouriseCppDoc(startPos, length, initStyle, keywordlists, styler, false);
458 }
459
460 LexerModule lmCPP(SCLEX_CPP, ColouriseCppDocSensitive, "cpp", FoldCppDoc, cppWordLists);
461 LexerModule lmCPPNoCase(SCLEX_CPPNOCASE, ColouriseCppDocInsensitive, "cppnocase", FoldCppDoc, cppWordLists);
462