1 /** @file LexRust.cxx
2 ** Lexer for Rust.
3 **
4 ** Copyright (c) 2013 by SiegeLord <slabode@aim.com>
5 ** Converted to lexer object and added further folding features/properties by "Udo Lechner" <dlchnr(at)gmx(dot)net>
6 **/
7 // Copyright 1998-2005 by Neil Hodgson <neilh@scintilla.org>
8 // The License.txt file describes the conditions under which this software may be distributed.
9
10 #include <stdlib.h>
11 #include <string.h>
12 #include <stdio.h>
13 #include <stdarg.h>
14 #include <assert.h>
15 #include <ctype.h>
16
17 #include <string>
18 #include <map>
19
20 #include "ILexer.h"
21 #include "Scintilla.h"
22 #include "SciLexer.h"
23
24 #include "PropSetSimple.h"
25 #include "WordList.h"
26 #include "LexAccessor.h"
27 #include "Accessor.h"
28 #include "StyleContext.h"
29 #include "CharacterSet.h"
30 #include "LexerModule.h"
31 #include "OptionSet.h"
32
33 #ifdef SCI_NAMESPACE
34 using namespace Scintilla;
35 #endif
36
37 static const int NUM_RUST_KEYWORD_LISTS = 7;
38 static const int MAX_RUST_IDENT_CHARS = 1023;
39
IsStreamCommentStyle(int style)40 static bool IsStreamCommentStyle(int style) {
41 return style == SCE_RUST_COMMENTBLOCK ||
42 style == SCE_RUST_COMMENTBLOCKDOC;
43 }
44
45 // Options used for LexerRust
46 struct OptionsRust {
47 bool fold;
48 bool foldSyntaxBased;
49 bool foldComment;
50 bool foldCommentMultiline;
51 bool foldCommentExplicit;
52 std::string foldExplicitStart;
53 std::string foldExplicitEnd;
54 bool foldExplicitAnywhere;
55 bool foldCompact;
56 int foldAtElseInt;
57 bool foldAtElse;
OptionsRustOptionsRust58 OptionsRust() {
59 fold = false;
60 foldSyntaxBased = true;
61 foldComment = false;
62 foldCommentMultiline = true;
63 foldCommentExplicit = true;
64 foldExplicitStart = "";
65 foldExplicitEnd = "";
66 foldExplicitAnywhere = false;
67 foldCompact = true;
68 foldAtElseInt = -1;
69 foldAtElse = false;
70 }
71 };
72
73 static const char * const rustWordLists[NUM_RUST_KEYWORD_LISTS + 1] = {
74 "Primary keywords and identifiers",
75 "Built in types",
76 "Other keywords",
77 "Keywords 4",
78 "Keywords 5",
79 "Keywords 6",
80 "Keywords 7",
81 0,
82 };
83
84 struct OptionSetRust : public OptionSet<OptionsRust> {
OptionSetRustOptionSetRust85 OptionSetRust() {
86 DefineProperty("fold", &OptionsRust::fold);
87
88 DefineProperty("fold.comment", &OptionsRust::foldComment);
89
90 DefineProperty("fold.compact", &OptionsRust::foldCompact);
91
92 DefineProperty("fold.at.else", &OptionsRust::foldAtElse);
93
94 DefineProperty("fold.rust.syntax.based", &OptionsRust::foldSyntaxBased,
95 "Set this property to 0 to disable syntax based folding.");
96
97 DefineProperty("fold.rust.comment.multiline", &OptionsRust::foldCommentMultiline,
98 "Set this property to 0 to disable folding multi-line comments when fold.comment=1.");
99
100 DefineProperty("fold.rust.comment.explicit", &OptionsRust::foldCommentExplicit,
101 "Set this property to 0 to disable folding explicit fold points when fold.comment=1.");
102
103 DefineProperty("fold.rust.explicit.start", &OptionsRust::foldExplicitStart,
104 "The string to use for explicit fold start points, replacing the standard //{.");
105
106 DefineProperty("fold.rust.explicit.end", &OptionsRust::foldExplicitEnd,
107 "The string to use for explicit fold end points, replacing the standard //}.");
108
109 DefineProperty("fold.rust.explicit.anywhere", &OptionsRust::foldExplicitAnywhere,
110 "Set this property to 1 to enable explicit fold points anywhere, not just in line comments.");
111
112 DefineProperty("lexer.rust.fold.at.else", &OptionsRust::foldAtElseInt,
113 "This option enables Rust folding on a \"} else {\" line of an if statement.");
114
115 DefineWordListSets(rustWordLists);
116 }
117 };
118
119 class LexerRust : public ILexer {
120 WordList keywords[NUM_RUST_KEYWORD_LISTS];
121 OptionsRust options;
122 OptionSetRust osRust;
123 public:
~LexerRust()124 virtual ~LexerRust() {
125 }
Release()126 void SCI_METHOD Release() {
127 delete this;
128 }
Version() const129 int SCI_METHOD Version() const {
130 return lvOriginal;
131 }
PropertyNames()132 const char * SCI_METHOD PropertyNames() {
133 return osRust.PropertyNames();
134 }
PropertyType(const char * name)135 int SCI_METHOD PropertyType(const char *name) {
136 return osRust.PropertyType(name);
137 }
DescribeProperty(const char * name)138 const char * SCI_METHOD DescribeProperty(const char *name) {
139 return osRust.DescribeProperty(name);
140 }
141 int SCI_METHOD PropertySet(const char *key, const char *val);
DescribeWordListSets()142 const char * SCI_METHOD DescribeWordListSets() {
143 return osRust.DescribeWordListSets();
144 }
145 int SCI_METHOD WordListSet(int n, const char *wl);
146 void SCI_METHOD Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
147 void SCI_METHOD Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess);
PrivateCall(int,void *)148 void * SCI_METHOD PrivateCall(int, void *) {
149 return 0;
150 }
LexerFactoryRust()151 static ILexer *LexerFactoryRust() {
152 return new LexerRust();
153 }
154 };
155
PropertySet(const char * key,const char * val)156 int SCI_METHOD LexerRust::PropertySet(const char *key, const char *val) {
157 if (osRust.PropertySet(&options, key, val)) {
158 return 0;
159 }
160 return -1;
161 }
162
WordListSet(int n,const char * wl)163 int SCI_METHOD LexerRust::WordListSet(int n, const char *wl) {
164 int firstModification = -1;
165 if (n < NUM_RUST_KEYWORD_LISTS) {
166 WordList *wordListN = &keywords[n];
167 WordList wlNew;
168 wlNew.Set(wl);
169 if (*wordListN != wlNew) {
170 wordListN->Set(wl);
171 firstModification = 0;
172 }
173 }
174 return firstModification;
175 }
176
IsWhitespace(int c)177 static bool IsWhitespace(int c) {
178 return c == ' ' || c == '\t' || c == '\r' || c == '\n';
179 }
180
181 /* This isn't quite right for Unicode identifiers */
IsIdentifierStart(int ch)182 static bool IsIdentifierStart(int ch) {
183 return (IsASCII(ch) && (isalpha(ch) || ch == '_')) || !IsASCII(ch);
184 }
185
186 /* This isn't quite right for Unicode identifiers */
IsIdentifierContinue(int ch)187 static bool IsIdentifierContinue(int ch) {
188 return (IsASCII(ch) && (isalnum(ch) || ch == '_')) || !IsASCII(ch);
189 }
190
ScanWhitespace(Accessor & styler,int & pos,int max)191 static void ScanWhitespace(Accessor& styler, int& pos, int max) {
192 while (IsWhitespace(styler.SafeGetCharAt(pos, '\0')) && pos < max) {
193 if (pos == styler.LineEnd(styler.GetLine(pos)))
194 styler.SetLineState(styler.GetLine(pos), 0);
195 pos++;
196 }
197 styler.ColourTo(pos-1, SCE_RUST_DEFAULT);
198 }
199
GrabString(char * s,Accessor & styler,int start,int len)200 static void GrabString(char* s, Accessor& styler, int start, int len) {
201 for (int ii = 0; ii < len; ii++)
202 s[ii] = styler[ii + start];
203 s[len] = '\0';
204 }
205
ScanIdentifier(Accessor & styler,int & pos,WordList * keywords)206 static void ScanIdentifier(Accessor& styler, int& pos, WordList *keywords) {
207 int start = pos;
208 while (IsIdentifierContinue(styler.SafeGetCharAt(pos, '\0')))
209 pos++;
210
211 if (styler.SafeGetCharAt(pos, '\0') == '!') {
212 pos++;
213 styler.ColourTo(pos - 1, SCE_RUST_MACRO);
214 } else {
215 char s[MAX_RUST_IDENT_CHARS + 1];
216 int len = pos - start;
217 len = len > MAX_RUST_IDENT_CHARS ? MAX_RUST_IDENT_CHARS : len;
218 GrabString(s, styler, start, len);
219 bool keyword = false;
220 for (int ii = 0; ii < NUM_RUST_KEYWORD_LISTS; ii++) {
221 if (keywords[ii].InList(s)) {
222 styler.ColourTo(pos - 1, SCE_RUST_WORD + ii);
223 keyword = true;
224 break;
225 }
226 }
227 if (!keyword) {
228 styler.ColourTo(pos - 1, SCE_RUST_IDENTIFIER);
229 }
230 }
231 }
232
233 /* Scans a sequence of digits, returning true if it found any. */
ScanDigits(Accessor & styler,int & pos,int base)234 static bool ScanDigits(Accessor& styler, int& pos, int base) {
235 int old_pos = pos;
236 for (;;) {
237 int c = styler.SafeGetCharAt(pos, '\0');
238 if (IsADigit(c, base) || c == '_')
239 pos++;
240 else
241 break;
242 }
243 return old_pos != pos;
244 }
245
246 /* Scans an integer and floating point literals. */
ScanNumber(Accessor & styler,int & pos)247 static void ScanNumber(Accessor& styler, int& pos) {
248 int base = 10;
249 int c = styler.SafeGetCharAt(pos, '\0');
250 int n = styler.SafeGetCharAt(pos + 1, '\0');
251 bool error = false;
252 /* Scan the prefix, thus determining the base.
253 * 10 is default if there's no prefix. */
254 if (c == '0' && n == 'x') {
255 pos += 2;
256 base = 16;
257 } else if (c == '0' && n == 'b') {
258 pos += 2;
259 base = 2;
260 } else if (c == '0' && n == 'o') {
261 pos += 2;
262 base = 8;
263 }
264
265 /* Scan initial digits. The literal is malformed if there are none. */
266 error |= !ScanDigits(styler, pos, base);
267 /* See if there's an integer suffix. We mimic the Rust's lexer
268 * and munch it even if there was an error above. */
269 c = styler.SafeGetCharAt(pos, '\0');
270 if (c == 'u' || c == 'i') {
271 pos++;
272 c = styler.SafeGetCharAt(pos, '\0');
273 n = styler.SafeGetCharAt(pos + 1, '\0');
274 if (c == '8') {
275 pos++;
276 } else if (c == '1' && n == '6') {
277 pos += 2;
278 } else if (c == '3' && n == '2') {
279 pos += 2;
280 } else if (c == '6' && n == '4') {
281 pos += 2;
282 }
283 /* See if it's a floating point literal. These literals have to be base 10.
284 */
285 } else if (!error) {
286 /* If there's a period, it's a floating point literal unless it's
287 * followed by an identifier (meaning this is a method call, e.g.
288 * `1.foo()`) or another period, in which case it's a range (e.g. 1..2)
289 */
290 n = styler.SafeGetCharAt(pos + 1, '\0');
291 if (c == '.' && !(IsIdentifierStart(n) || n == '.')) {
292 error |= base != 10;
293 pos++;
294 /* It's ok to have no digits after the period. */
295 ScanDigits(styler, pos, 10);
296 }
297
298 /* Look for the exponentiation. */
299 c = styler.SafeGetCharAt(pos, '\0');
300 if (c == 'e' || c == 'E') {
301 error |= base != 10;
302 pos++;
303 c = styler.SafeGetCharAt(pos, '\0');
304 if (c == '-' || c == '+')
305 pos++;
306 /* It is invalid to have no digits in the exponent. */
307 error |= !ScanDigits(styler, pos, 10);
308 }
309
310 /* Scan the floating point suffix. */
311 c = styler.SafeGetCharAt(pos, '\0');
312 if (c == 'f') {
313 error |= base != 10;
314 pos++;
315 c = styler.SafeGetCharAt(pos, '\0');
316 n = styler.SafeGetCharAt(pos + 1, '\0');
317 if (c == '3' && n == '2') {
318 pos += 2;
319 } else if (c == '6' && n == '4') {
320 pos += 2;
321 } else {
322 error = true;
323 }
324 }
325 }
326
327 if (error)
328 styler.ColourTo(pos - 1, SCE_RUST_LEXERROR);
329 else
330 styler.ColourTo(pos - 1, SCE_RUST_NUMBER);
331 }
332
IsOneCharOperator(int c)333 static bool IsOneCharOperator(int c) {
334 return c == ';' || c == ',' || c == '(' || c == ')'
335 || c == '{' || c == '}' || c == '[' || c == ']'
336 || c == '@' || c == '#' || c == '~' || c == '+'
337 || c == '*' || c == '/' || c == '^' || c == '%'
338 || c == '.' || c == ':' || c == '!' || c == '<'
339 || c == '>' || c == '=' || c == '-' || c == '&'
340 || c == '|' || c == '$';
341 }
342
IsTwoCharOperator(int c,int n)343 static bool IsTwoCharOperator(int c, int n) {
344 return (c == '.' && n == '.') || (c == ':' && n == ':')
345 || (c == '!' && n == '=') || (c == '<' && n == '<')
346 || (c == '<' && n == '=') || (c == '>' && n == '>')
347 || (c == '>' && n == '=') || (c == '=' && n == '=')
348 || (c == '=' && n == '>') || (c == '-' && n == '>')
349 || (c == '&' && n == '&') || (c == '|' && n == '|')
350 || (c == '-' && n == '=') || (c == '&' && n == '=')
351 || (c == '|' && n == '=') || (c == '+' && n == '=')
352 || (c == '*' && n == '=') || (c == '/' && n == '=')
353 || (c == '^' && n == '=') || (c == '%' && n == '=');
354 }
355
IsThreeCharOperator(int c,int n,int n2)356 static bool IsThreeCharOperator(int c, int n, int n2) {
357 return (c == '<' && n == '<' && n2 == '=')
358 || (c == '>' && n == '>' && n2 == '=');
359 }
360
IsValidCharacterEscape(int c)361 static bool IsValidCharacterEscape(int c) {
362 return c == 'n' || c == 'r' || c == 't' || c == '\\'
363 || c == '\'' || c == '"' || c == '0';
364 }
365
IsValidStringEscape(int c)366 static bool IsValidStringEscape(int c) {
367 return IsValidCharacterEscape(c) || c == '\n' || c == '\r';
368 }
369
ScanNumericEscape(Accessor & styler,int & pos,int num_digits,bool stop_asap)370 static bool ScanNumericEscape(Accessor &styler, int& pos, int num_digits, bool stop_asap) {
371 for (;;) {
372 int c = styler.SafeGetCharAt(pos, '\0');
373 if (!IsADigit(c, 16))
374 break;
375 num_digits--;
376 pos++;
377 if (num_digits == 0 && stop_asap)
378 return true;
379 }
380 if (num_digits == 0) {
381 return true;
382 } else {
383 return false;
384 }
385 }
386
387 /* This is overly permissive for character literals in order to accept UTF-8 encoded
388 * character literals. */
ScanCharacterLiteralOrLifetime(Accessor & styler,int & pos,bool ascii_only)389 static void ScanCharacterLiteralOrLifetime(Accessor &styler, int& pos, bool ascii_only) {
390 pos++;
391 int c = styler.SafeGetCharAt(pos, '\0');
392 int n = styler.SafeGetCharAt(pos + 1, '\0');
393 bool done = false;
394 bool valid_lifetime = !ascii_only && IsIdentifierStart(c);
395 bool valid_char = true;
396 bool first = true;
397 while (!done) {
398 switch (c) {
399 case '\\':
400 done = true;
401 if (IsValidCharacterEscape(n)) {
402 pos += 2;
403 } else if (n == 'x') {
404 pos += 2;
405 valid_char = ScanNumericEscape(styler, pos, 2, false);
406 } else if (n == 'u' && !ascii_only) {
407 pos += 2;
408 valid_char = ScanNumericEscape(styler, pos, 4, false);
409 } else if (n == 'U' && !ascii_only) {
410 pos += 2;
411 valid_char = ScanNumericEscape(styler, pos, 8, false);
412 } else {
413 valid_char = false;
414 }
415 break;
416 case '\'':
417 valid_char = !first;
418 done = true;
419 break;
420 case '\t':
421 case '\n':
422 case '\r':
423 case '\0':
424 valid_char = false;
425 done = true;
426 break;
427 default:
428 if (ascii_only && !IsASCII((char)c)) {
429 done = true;
430 valid_char = false;
431 } else if (!IsIdentifierContinue(c) && !first) {
432 done = true;
433 } else {
434 pos++;
435 }
436 break;
437 }
438 c = styler.SafeGetCharAt(pos, '\0');
439 n = styler.SafeGetCharAt(pos + 1, '\0');
440
441 first = false;
442 }
443 if (styler.SafeGetCharAt(pos, '\0') == '\'') {
444 valid_lifetime = false;
445 } else {
446 valid_char = false;
447 }
448 if (valid_lifetime) {
449 styler.ColourTo(pos - 1, SCE_RUST_LIFETIME);
450 } else if (valid_char) {
451 pos++;
452 styler.ColourTo(pos - 1, ascii_only ? SCE_RUST_BYTECHARACTER : SCE_RUST_CHARACTER);
453 } else {
454 styler.ColourTo(pos - 1, SCE_RUST_LEXERROR);
455 }
456 }
457
458 enum CommentState {
459 UnknownComment,
460 DocComment,
461 NotDocComment
462 };
463
464 /*
465 * The rule for block-doc comments is as follows: /xxN and /x! (where x is an asterisk, N is a non-asterisk) start doc comments.
466 * Otherwise it's a regular comment.
467 */
ResumeBlockComment(Accessor & styler,int & pos,int max,CommentState state,int level)468 static void ResumeBlockComment(Accessor &styler, int& pos, int max, CommentState state, int level) {
469 int c = styler.SafeGetCharAt(pos, '\0');
470 bool maybe_doc_comment = false;
471 if (c == '*') {
472 int n = styler.SafeGetCharAt(pos + 1, '\0');
473 if (n != '*' && n != '/') {
474 maybe_doc_comment = true;
475 }
476 } else if (c == '!') {
477 maybe_doc_comment = true;
478 }
479
480 for (;;) {
481 int n = styler.SafeGetCharAt(pos + 1, '\0');
482 if (pos == styler.LineEnd(styler.GetLine(pos)))
483 styler.SetLineState(styler.GetLine(pos), level);
484 if (c == '*') {
485 pos++;
486 if (n == '/') {
487 pos++;
488 level--;
489 if (level == 0) {
490 styler.SetLineState(styler.GetLine(pos), 0);
491 if (state == DocComment || (state == UnknownComment && maybe_doc_comment))
492 styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCKDOC);
493 else
494 styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCK);
495 break;
496 }
497 }
498 } else if (c == '/') {
499 pos++;
500 if (n == '*') {
501 pos++;
502 level++;
503 }
504 }
505 else {
506 pos++;
507 }
508 if (pos >= max) {
509 if (state == DocComment || (state == UnknownComment && maybe_doc_comment))
510 styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCKDOC);
511 else
512 styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCK);
513 break;
514 }
515 c = styler.SafeGetCharAt(pos, '\0');
516 }
517 }
518
519 /*
520 * The rule for line-doc comments is as follows... ///N and //! (where N is a non slash) start doc comments.
521 * Otherwise it's a normal line comment.
522 */
ResumeLineComment(Accessor & styler,int & pos,int max,CommentState state)523 static void ResumeLineComment(Accessor &styler, int& pos, int max, CommentState state) {
524 bool maybe_doc_comment = false;
525 int c = styler.SafeGetCharAt(pos, '\0');
526 if (c == '/') {
527 if (pos < max) {
528 pos++;
529 c = styler.SafeGetCharAt(pos, '\0');
530 if (c != '/') {
531 maybe_doc_comment = true;
532 }
533 }
534 } else if (c == '!') {
535 maybe_doc_comment = true;
536 }
537
538 while (pos < max && c != '\n') {
539 if (pos == styler.LineEnd(styler.GetLine(pos)))
540 styler.SetLineState(styler.GetLine(pos), 0);
541 pos++;
542 c = styler.SafeGetCharAt(pos, '\0');
543 }
544
545 if (state == DocComment || (state == UnknownComment && maybe_doc_comment))
546 styler.ColourTo(pos - 1, SCE_RUST_COMMENTLINEDOC);
547 else
548 styler.ColourTo(pos - 1, SCE_RUST_COMMENTLINE);
549 }
550
ScanComments(Accessor & styler,int & pos,int max)551 static void ScanComments(Accessor &styler, int& pos, int max) {
552 pos++;
553 int c = styler.SafeGetCharAt(pos, '\0');
554 pos++;
555 if (c == '/')
556 ResumeLineComment(styler, pos, max, UnknownComment);
557 else if (c == '*')
558 ResumeBlockComment(styler, pos, max, UnknownComment, 1);
559 }
560
ResumeString(Accessor & styler,int & pos,int max,bool ascii_only)561 static void ResumeString(Accessor &styler, int& pos, int max, bool ascii_only) {
562 int c = styler.SafeGetCharAt(pos, '\0');
563 bool error = false;
564 while (c != '"' && !error) {
565 if (pos >= max) {
566 error = true;
567 break;
568 }
569 if (pos == styler.LineEnd(styler.GetLine(pos)))
570 styler.SetLineState(styler.GetLine(pos), 0);
571 if (c == '\\') {
572 int n = styler.SafeGetCharAt(pos + 1, '\0');
573 if (IsValidStringEscape(n)) {
574 pos += 2;
575 } else if (n == 'x') {
576 pos += 2;
577 error = !ScanNumericEscape(styler, pos, 2, true);
578 } else if (n == 'u' && !ascii_only) {
579 pos += 2;
580 error = !ScanNumericEscape(styler, pos, 4, true);
581 } else if (n == 'U' && !ascii_only) {
582 pos += 2;
583 error = !ScanNumericEscape(styler, pos, 8, true);
584 } else {
585 pos += 1;
586 error = true;
587 }
588 } else {
589 if (ascii_only && !IsASCII((char)c))
590 error = true;
591 else
592 pos++;
593 }
594 c = styler.SafeGetCharAt(pos, '\0');
595 }
596 if (!error)
597 pos++;
598 styler.ColourTo(pos - 1, ascii_only ? SCE_RUST_BYTESTRING : SCE_RUST_STRING);
599 }
600
ResumeRawString(Accessor & styler,int & pos,int max,int num_hashes,bool ascii_only)601 static void ResumeRawString(Accessor &styler, int& pos, int max, int num_hashes, bool ascii_only) {
602 for (;;) {
603 if (pos == styler.LineEnd(styler.GetLine(pos)))
604 styler.SetLineState(styler.GetLine(pos), num_hashes);
605
606 int c = styler.SafeGetCharAt(pos, '\0');
607 if (c == '"') {
608 pos++;
609 int trailing_num_hashes = 0;
610 while (styler.SafeGetCharAt(pos, '\0') == '#' && trailing_num_hashes < num_hashes) {
611 trailing_num_hashes++;
612 pos++;
613 }
614 if (trailing_num_hashes == num_hashes) {
615 styler.SetLineState(styler.GetLine(pos), 0);
616 break;
617 }
618 } else if (pos >= max) {
619 break;
620 } else {
621 if (ascii_only && !IsASCII((char)c))
622 break;
623 pos++;
624 }
625 }
626 styler.ColourTo(pos - 1, ascii_only ? SCE_RUST_BYTESTRINGR : SCE_RUST_STRINGR);
627 }
628
ScanRawString(Accessor & styler,int & pos,int max,bool ascii_only)629 static void ScanRawString(Accessor &styler, int& pos, int max, bool ascii_only) {
630 pos++;
631 int num_hashes = 0;
632 while (styler.SafeGetCharAt(pos, '\0') == '#') {
633 num_hashes++;
634 pos++;
635 }
636 if (styler.SafeGetCharAt(pos, '\0') != '"') {
637 styler.ColourTo(pos - 1, SCE_RUST_LEXERROR);
638 } else {
639 pos++;
640 ResumeRawString(styler, pos, max, num_hashes, ascii_only);
641 }
642 }
643
Lex(unsigned int startPos,int length,int initStyle,IDocument * pAccess)644 void SCI_METHOD LexerRust::Lex(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
645 PropSetSimple props;
646 Accessor styler(pAccess, &props);
647 int pos = startPos;
648 int max = pos + length;
649
650 styler.StartAt(pos);
651 styler.StartSegment(pos);
652
653 if (initStyle == SCE_RUST_COMMENTBLOCK || initStyle == SCE_RUST_COMMENTBLOCKDOC) {
654 ResumeBlockComment(styler, pos, max, initStyle == SCE_RUST_COMMENTBLOCKDOC ? DocComment : NotDocComment, styler.GetLineState(styler.GetLine(pos) - 1));
655 } else if (initStyle == SCE_RUST_COMMENTLINE || initStyle == SCE_RUST_COMMENTLINEDOC) {
656 ResumeLineComment(styler, pos, max, initStyle == SCE_RUST_COMMENTLINEDOC ? DocComment : NotDocComment);
657 } else if (initStyle == SCE_RUST_STRING) {
658 ResumeString(styler, pos, max, false);
659 } else if (initStyle == SCE_RUST_BYTESTRING) {
660 ResumeString(styler, pos, max, true);
661 } else if (initStyle == SCE_RUST_STRINGR) {
662 ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), false);
663 } else if (initStyle == SCE_RUST_BYTESTRINGR) {
664 ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), true);
665 }
666
667 while (pos < max) {
668 int c = styler.SafeGetCharAt(pos, '\0');
669 int n = styler.SafeGetCharAt(pos + 1, '\0');
670 int n2 = styler.SafeGetCharAt(pos + 2, '\0');
671
672 if (pos == 0 && c == '#' && n == '!' && n2 != '[') {
673 pos += 2;
674 ResumeLineComment(styler, pos, max, NotDocComment);
675 } else if (IsWhitespace(c)) {
676 ScanWhitespace(styler, pos, max);
677 } else if (c == '/' && (n == '/' || n == '*')) {
678 ScanComments(styler, pos, max);
679 } else if (c == 'r' && (n == '#' || n == '"')) {
680 ScanRawString(styler, pos, max, false);
681 } else if (c == 'b' && n == 'r' && (n2 == '#' || n2 == '"')) {
682 pos++;
683 ScanRawString(styler, pos, max, true);
684 } else if (c == 'b' && n == '"') {
685 pos += 2;
686 ResumeString(styler, pos, max, true);
687 } else if (c == 'b' && n == '\'') {
688 pos++;
689 ScanCharacterLiteralOrLifetime(styler, pos, true);
690 } else if (IsIdentifierStart(c)) {
691 ScanIdentifier(styler, pos, keywords);
692 } else if (IsADigit(c)) {
693 ScanNumber(styler, pos);
694 } else if (IsThreeCharOperator(c, n, n2)) {
695 pos += 3;
696 styler.ColourTo(pos - 1, SCE_RUST_OPERATOR);
697 } else if (IsTwoCharOperator(c, n)) {
698 pos += 2;
699 styler.ColourTo(pos - 1, SCE_RUST_OPERATOR);
700 } else if (IsOneCharOperator(c)) {
701 pos++;
702 styler.ColourTo(pos - 1, SCE_RUST_OPERATOR);
703 } else if (c == '\'') {
704 ScanCharacterLiteralOrLifetime(styler, pos, false);
705 } else if (c == '"') {
706 pos++;
707 ResumeString(styler, pos, max, false);
708 } else {
709 pos++;
710 styler.ColourTo(pos - 1, SCE_RUST_LEXERROR);
711 }
712 }
713 styler.ColourTo(pos - 1, SCE_RUST_DEFAULT);
714 styler.Flush();
715 }
716
Fold(unsigned int startPos,int length,int initStyle,IDocument * pAccess)717 void SCI_METHOD LexerRust::Fold(unsigned int startPos, int length, int initStyle, IDocument *pAccess) {
718
719 if (!options.fold)
720 return;
721
722 LexAccessor styler(pAccess);
723
724 unsigned int endPos = startPos + length;
725 int visibleChars = 0;
726 bool inLineComment = false;
727 int lineCurrent = styler.GetLine(startPos);
728 int levelCurrent = SC_FOLDLEVELBASE;
729 if (lineCurrent > 0)
730 levelCurrent = styler.LevelAt(lineCurrent-1) >> 16;
731 unsigned int lineStartNext = styler.LineStart(lineCurrent+1);
732 int levelMinCurrent = levelCurrent;
733 int levelNext = levelCurrent;
734 char chNext = styler[startPos];
735 int styleNext = styler.StyleAt(startPos);
736 int style = initStyle;
737 const bool userDefinedFoldMarkers = !options.foldExplicitStart.empty() && !options.foldExplicitEnd.empty();
738 for (unsigned int i = startPos; i < endPos; i++) {
739 char ch = chNext;
740 chNext = styler.SafeGetCharAt(i + 1);
741 int stylePrev = style;
742 style = styleNext;
743 styleNext = styler.StyleAt(i + 1);
744 bool atEOL = i == (lineStartNext-1);
745 if ((style == SCE_RUST_COMMENTLINE) || (style == SCE_RUST_COMMENTLINEDOC))
746 inLineComment = true;
747 if (options.foldComment && options.foldCommentMultiline && IsStreamCommentStyle(style) && !inLineComment) {
748 if (!IsStreamCommentStyle(stylePrev)) {
749 levelNext++;
750 } else if (!IsStreamCommentStyle(styleNext) && !atEOL) {
751 // Comments don't end at end of line and the next character may be unstyled.
752 levelNext--;
753 }
754 }
755 if (options.foldComment && options.foldCommentExplicit && ((style == SCE_RUST_COMMENTLINE) || options.foldExplicitAnywhere)) {
756 if (userDefinedFoldMarkers) {
757 if (styler.Match(i, options.foldExplicitStart.c_str())) {
758 levelNext++;
759 } else if (styler.Match(i, options.foldExplicitEnd.c_str())) {
760 levelNext--;
761 }
762 } else {
763 if ((ch == '/') && (chNext == '/')) {
764 char chNext2 = styler.SafeGetCharAt(i + 2);
765 if (chNext2 == '{') {
766 levelNext++;
767 } else if (chNext2 == '}') {
768 levelNext--;
769 }
770 }
771 }
772 }
773 if (options.foldSyntaxBased && (style == SCE_RUST_OPERATOR)) {
774 if (ch == '{') {
775 // Measure the minimum before a '{' to allow
776 // folding on "} else {"
777 if (levelMinCurrent > levelNext) {
778 levelMinCurrent = levelNext;
779 }
780 levelNext++;
781 } else if (ch == '}') {
782 levelNext--;
783 }
784 }
785 if (!IsASpace(ch))
786 visibleChars++;
787 if (atEOL || (i == endPos-1)) {
788 int levelUse = levelCurrent;
789 if (options.foldSyntaxBased && options.foldAtElse) {
790 levelUse = levelMinCurrent;
791 }
792 int lev = levelUse | levelNext << 16;
793 if (visibleChars == 0 && options.foldCompact)
794 lev |= SC_FOLDLEVELWHITEFLAG;
795 if (levelUse < levelNext)
796 lev |= SC_FOLDLEVELHEADERFLAG;
797 if (lev != styler.LevelAt(lineCurrent)) {
798 styler.SetLevel(lineCurrent, lev);
799 }
800 lineCurrent++;
801 lineStartNext = styler.LineStart(lineCurrent+1);
802 levelCurrent = levelNext;
803 levelMinCurrent = levelCurrent;
804 if (atEOL && (i == static_cast<unsigned int>(styler.Length()-1))) {
805 // There is an empty line at end of file so give it same level and empty
806 styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG);
807 }
808 visibleChars = 0;
809 inLineComment = false;
810 }
811 }
812 }
813
814 LexerModule lmRust(SCLEX_RUST, LexerRust::LexerFactoryRust, "rust", rustWordLists);
815