| 1 |
/** @file LexRust.cxx |
| 2 |
** Lexer for Rust. |
| 3 |
** |
| 4 |
** Copyright (c) 2013 by SiegeLord <[email protected]> |
| 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 <[email protected]> |
| 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 <string_view> |
| 19 |
#include <map> |
| 20 |
#include <functional> |
| 21 |
|
| 22 |
#include "ILexer.h" |
| 23 |
#include "Scintilla.h" |
| 24 |
#include "SciLexer.h" |
| 25 |
|
| 26 |
#include "PropSetSimple.h" |
| 27 |
#include "WordList.h" |
| 28 |
#include "LexAccessor.h" |
| 29 |
#include "Accessor.h" |
| 30 |
#include "StyleContext.h" |
| 31 |
#include "CharacterSet.h" |
| 32 |
#include "LexerModule.h" |
| 33 |
#include "OptionSet.h" |
| 34 |
#include "DefaultLexer.h" |
| 35 |
|
| 36 |
using namespace Scintilla; |
| 37 |
using namespace Lexilla; |
| 38 |
|
| 39 |
namespace { |
| 40 |
|
| 41 |
constexpr int NUM_RUST_KEYWORD_LISTS = 7; |
| 42 |
constexpr int MAX_RUST_IDENT_CHARS = 1023; |
| 43 |
|
| 44 |
|
| 45 |
enum class StringType : int { |
| 46 |
STRING = SCE_RUST_STRING, |
| 47 |
BYTESTRING = SCE_RUST_BYTESTRING, |
| 48 |
CSTRING = SCE_RUST_CSTRING, |
| 49 |
|
| 50 |
RAW_STRING = SCE_RUST_STRINGR, |
| 51 |
RAW_BYTESTRING = SCE_RUST_BYTESTRINGR, |
| 52 |
RAW_CSTRING = SCE_RUST_CSTRINGR |
| 53 |
}; |
| 54 |
|
| 55 |
static bool IsStreamCommentStyle(int style) noexcept { |
| 56 |
return style == SCE_RUST_COMMENTBLOCK || |
| 57 |
style == SCE_RUST_COMMENTBLOCKDOC; |
| 58 |
} |
| 59 |
|
| 60 |
// Options used for LexerRust |
| 61 |
struct OptionsRust { |
| 62 |
bool fold; |
| 63 |
bool foldSyntaxBased; |
| 64 |
bool foldComment; |
| 65 |
bool foldCommentMultiline; |
| 66 |
bool foldCommentExplicit; |
| 67 |
std::string foldExplicitStart; |
| 68 |
std::string foldExplicitEnd; |
| 69 |
bool foldExplicitAnywhere; |
| 70 |
bool foldCompact; |
| 71 |
int foldAtElseInt; // This variable is not used |
| 72 |
bool foldAtElse; |
| 73 |
OptionsRust() { |
| 74 |
fold = false; |
| 75 |
foldSyntaxBased = true; |
| 76 |
foldComment = false; |
| 77 |
foldCommentMultiline = true; |
| 78 |
foldCommentExplicit = true; |
| 79 |
foldExplicitStart = ""; |
| 80 |
foldExplicitEnd = ""; |
| 81 |
foldExplicitAnywhere = false; |
| 82 |
foldCompact = true; |
| 83 |
foldAtElseInt = -1; |
| 84 |
foldAtElse = false; |
| 85 |
} |
| 86 |
}; |
| 87 |
|
| 88 |
static const char * const rustWordLists[NUM_RUST_KEYWORD_LISTS + 1] = { |
| 89 |
"Primary keywords and identifiers", |
| 90 |
"Built in types", |
| 91 |
"Other keywords", |
| 92 |
"Keywords 4", |
| 93 |
"Keywords 5", |
| 94 |
"Keywords 6", |
| 95 |
"Keywords 7", |
| 96 |
0, |
| 97 |
}; |
| 98 |
|
| 99 |
struct OptionSetRust : public OptionSet<OptionsRust> { |
| 100 |
OptionSetRust() { |
| 101 |
DefineProperty("fold", &OptionsRust::fold); |
| 102 |
|
| 103 |
DefineProperty("fold.comment", &OptionsRust::foldComment); |
| 104 |
|
| 105 |
DefineProperty("fold.compact", &OptionsRust::foldCompact); |
| 106 |
|
| 107 |
DefineProperty("fold.at.else", &OptionsRust::foldAtElse); |
| 108 |
|
| 109 |
DefineProperty("fold.rust.syntax.based", &OptionsRust::foldSyntaxBased, |
| 110 |
"Set this property to 0 to disable syntax based folding."); |
| 111 |
|
| 112 |
DefineProperty("fold.rust.comment.multiline", &OptionsRust::foldCommentMultiline, |
| 113 |
"Set this property to 0 to disable folding multi-line comments when fold.comment=1."); |
| 114 |
|
| 115 |
DefineProperty("fold.rust.comment.explicit", &OptionsRust::foldCommentExplicit, |
| 116 |
"Set this property to 0 to disable folding explicit fold points when fold.comment=1."); |
| 117 |
|
| 118 |
DefineProperty("fold.rust.explicit.start", &OptionsRust::foldExplicitStart, |
| 119 |
"The string to use for explicit fold start points, replacing the standard //{."); |
| 120 |
|
| 121 |
DefineProperty("fold.rust.explicit.end", &OptionsRust::foldExplicitEnd, |
| 122 |
"The string to use for explicit fold end points, replacing the standard //}."); |
| 123 |
|
| 124 |
DefineProperty("fold.rust.explicit.anywhere", &OptionsRust::foldExplicitAnywhere, |
| 125 |
"Set this property to 1 to enable explicit fold points anywhere, not just in line comments."); |
| 126 |
|
| 127 |
DefineProperty("lexer.rust.fold.at.else", &OptionsRust::foldAtElseInt, |
| 128 |
"This option enables Rust folding on a \"} else {\" line of an if statement."); |
| 129 |
|
| 130 |
DefineWordListSets(rustWordLists); |
| 131 |
} |
| 132 |
}; |
| 133 |
|
| 134 |
class LexerRust : public DefaultLexer { |
| 135 |
WordList keywords[NUM_RUST_KEYWORD_LISTS]; |
| 136 |
OptionsRust options; |
| 137 |
OptionSetRust osRust; |
| 138 |
public: |
| 139 |
LexerRust() : DefaultLexer("rust", SCLEX_RUST) { |
| 140 |
} |
| 141 |
virtual ~LexerRust() { |
| 142 |
} |
| 143 |
void SCI_METHOD Release() override { |
| 144 |
delete this; |
| 145 |
} |
| 146 |
int SCI_METHOD Version() const override { |
| 147 |
return lvRelease5; |
| 148 |
} |
| 149 |
const char * SCI_METHOD PropertyNames() override { |
| 150 |
return osRust.PropertyNames(); |
| 151 |
} |
| 152 |
int SCI_METHOD PropertyType(const char *name) override { |
| 153 |
return osRust.PropertyType(name); |
| 154 |
} |
| 155 |
const char * SCI_METHOD DescribeProperty(const char *name) override { |
| 156 |
return osRust.DescribeProperty(name); |
| 157 |
} |
| 158 |
Sci_Position SCI_METHOD PropertySet(const char *key, const char *val) override; |
| 159 |
const char * SCI_METHOD PropertyGet(const char *key) override { |
| 160 |
return osRust.PropertyGet(key); |
| 161 |
} |
| 162 |
const char * SCI_METHOD DescribeWordListSets() override { |
| 163 |
return osRust.DescribeWordListSets(); |
| 164 |
} |
| 165 |
Sci_Position SCI_METHOD WordListSet(int n, const char *wl) override; |
| 166 |
void SCI_METHOD Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override; |
| 167 |
void SCI_METHOD Fold(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override; |
| 168 |
void * SCI_METHOD PrivateCall(int, void *) override { |
| 169 |
return 0; |
| 170 |
} |
| 171 |
static ILexer5 *LexerFactoryRust() { |
| 172 |
return new LexerRust(); |
| 173 |
} |
| 174 |
}; |
| 175 |
|
| 176 |
Sci_Position SCI_METHOD LexerRust::PropertySet(const char *key, const char *val) { |
| 177 |
if (osRust.PropertySet(&options, key, val)) { |
| 178 |
return 0; |
| 179 |
} |
| 180 |
return -1; |
| 181 |
} |
| 182 |
|
| 183 |
Sci_Position SCI_METHOD LexerRust::WordListSet(int n, const char *wl) { |
| 184 |
Sci_Position firstModification = -1; |
| 185 |
if (n < NUM_RUST_KEYWORD_LISTS) { |
| 186 |
WordList *wordListN = &keywords[n]; |
| 187 |
if (wordListN->Set(wl)) { |
| 188 |
firstModification = 0; |
| 189 |
} |
| 190 |
} |
| 191 |
return firstModification; |
| 192 |
} |
| 193 |
|
| 194 |
static bool IsWhitespace(int c) { |
| 195 |
return c == ' ' || c == '\t' || c == '\r' || c == '\n'; |
| 196 |
} |
| 197 |
|
| 198 |
/* This isn't quite right for Unicode identifiers */ |
| 199 |
static bool IsIdentifierStart(int ch) { |
| 200 |
return (IsASCII(ch) && (isalpha(ch) || ch == '_')) || !IsASCII(ch); |
| 201 |
} |
| 202 |
|
| 203 |
/* This isn't quite right for Unicode identifiers */ |
| 204 |
static bool IsIdentifierContinue(int ch) { |
| 205 |
return (IsASCII(ch) && (isalnum(ch) || ch == '_')) || !IsASCII(ch); |
| 206 |
} |
| 207 |
|
| 208 |
static void ScanWhitespace(Accessor& styler, Sci_Position& pos, Sci_Position max) { |
| 209 |
while (IsWhitespace(styler.SafeGetCharAt(pos, '\0')) && pos < max) { |
| 210 |
if (pos == styler.LineEnd(styler.GetLine(pos))) |
| 211 |
styler.SetLineState(styler.GetLine(pos), 0); |
| 212 |
pos++; |
| 213 |
} |
| 214 |
styler.ColourTo(pos-1, SCE_RUST_DEFAULT); |
| 215 |
} |
| 216 |
|
| 217 |
static void GrabString(char* s, Accessor& styler, Sci_Position start, Sci_Position len) { |
| 218 |
for (Sci_Position ii = 0; ii < len; ii++) |
| 219 |
s[ii] = styler[ii + start]; |
| 220 |
s[len] = '\0'; |
| 221 |
} |
| 222 |
|
| 223 |
static void ScanRawIdentifier(Accessor& styler, Sci_Position& pos) { |
| 224 |
const Sci_Position start = pos; |
| 225 |
while (IsIdentifierContinue(styler.SafeGetCharAt(pos, '\0'))) |
| 226 |
pos++; |
| 227 |
|
| 228 |
char s[MAX_RUST_IDENT_CHARS + 1]; |
| 229 |
Sci_Position len = pos - start; |
| 230 |
len = len > MAX_RUST_IDENT_CHARS ? MAX_RUST_IDENT_CHARS : len; |
| 231 |
GrabString(s, styler, start, len); |
| 232 |
// restricted values https://doc.rust-lang.org/reference/identifiers.html#raw-identifiers |
| 233 |
if (strcmp(s, "crate") != 0 && strcmp(s, "self") != 0 && |
| 234 |
strcmp(s, "super") != 0 && strcmp(s, "Self") != 0) { |
| 235 |
styler.ColourTo(pos - 1, SCE_RUST_IDENTIFIER); |
| 236 |
} else { |
| 237 |
styler.ColourTo(pos - 1, SCE_RUST_LEXERROR); |
| 238 |
} |
| 239 |
} |
| 240 |
|
| 241 |
static void ScanIdentifier(Accessor& styler, Sci_Position& pos, WordList *keywords) { |
| 242 |
const Sci_Position start = pos; |
| 243 |
while (IsIdentifierContinue(styler.SafeGetCharAt(pos, '\0'))) |
| 244 |
pos++; |
| 245 |
|
| 246 |
if (styler.SafeGetCharAt(pos, '\0') == '!') { |
| 247 |
pos++; |
| 248 |
styler.ColourTo(pos - 1, SCE_RUST_MACRO); |
| 249 |
} else { |
| 250 |
char s[MAX_RUST_IDENT_CHARS + 1]; |
| 251 |
Sci_Position len = pos - start; |
| 252 |
len = len > MAX_RUST_IDENT_CHARS ? MAX_RUST_IDENT_CHARS : len; |
| 253 |
GrabString(s, styler, start, len); |
| 254 |
bool keyword = false; |
| 255 |
for (int ii = 0; ii < NUM_RUST_KEYWORD_LISTS; ii++) { |
| 256 |
if (keywords[ii].InList(s)) { |
| 257 |
styler.ColourTo(pos - 1, SCE_RUST_WORD + ii); |
| 258 |
keyword = true; |
| 259 |
break; |
| 260 |
} |
| 261 |
} |
| 262 |
if (!keyword) { |
| 263 |
styler.ColourTo(pos - 1, SCE_RUST_IDENTIFIER); |
| 264 |
} |
| 265 |
} |
| 266 |
} |
| 267 |
|
| 268 |
/* Scans a sequence of digits, returning true if it found any. */ |
| 269 |
static bool ScanDigits(Accessor& styler, Sci_Position& pos, int base) { |
| 270 |
const Sci_Position old_pos = pos; |
| 271 |
for (;;) { |
| 272 |
const int c = styler.SafeGetCharAt(pos, '\0'); |
| 273 |
if (IsADigit(c, base) || c == '_') |
| 274 |
pos++; |
| 275 |
else |
| 276 |
break; |
| 277 |
} |
| 278 |
return old_pos != pos; |
| 279 |
} |
| 280 |
|
| 281 |
/* Scans an integer and floating point literals. */ |
| 282 |
static void ScanNumber(Accessor& styler, Sci_Position& pos) { |
| 283 |
int base = 10; |
| 284 |
int c = styler.SafeGetCharAt(pos, '\0'); |
| 285 |
int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 286 |
bool error = false; |
| 287 |
/* Scan the prefix, thus determining the base. |
| 288 |
* 10 is default if there's no prefix. */ |
| 289 |
if (c == '0' && n == 'x') { |
| 290 |
pos += 2; |
| 291 |
base = 16; |
| 292 |
} else if (c == '0' && n == 'b') { |
| 293 |
pos += 2; |
| 294 |
base = 2; |
| 295 |
} else if (c == '0' && n == 'o') { |
| 296 |
pos += 2; |
| 297 |
base = 8; |
| 298 |
} |
| 299 |
|
| 300 |
/* Scan initial digits. The literal is malformed if there are none. */ |
| 301 |
error |= !ScanDigits(styler, pos, base); |
| 302 |
/* See if there's an integer suffix. We mimic the Rust's lexer |
| 303 |
* and munch it even if there was an error above. */ |
| 304 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 305 |
if (c == 'u' || c == 'i') { |
| 306 |
pos++; |
| 307 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 308 |
n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 309 |
if (c == '8') { |
| 310 |
pos++; |
| 311 |
} else if (c == '1' && n == '6') { |
| 312 |
pos += 2; |
| 313 |
} else if (c == '3' && n == '2') { |
| 314 |
pos += 2; |
| 315 |
} else if (c == '6' && n == '4') { |
| 316 |
pos += 2; |
| 317 |
} else if (styler.Match(pos, "128")) { |
| 318 |
pos += 3; |
| 319 |
} else if (styler.Match(pos, "size")) { |
| 320 |
pos += 4; |
| 321 |
} else { |
| 322 |
error = true; |
| 323 |
} |
| 324 |
/* See if it's a floating point literal. These literals have to be base 10. |
| 325 |
*/ |
| 326 |
} else if (!error) { |
| 327 |
/* If there's a period, it's a floating point literal unless it's |
| 328 |
* followed by an identifier (meaning this is a method call, e.g. |
| 329 |
* `1.foo()`) or another period, in which case it's a range (e.g. 1..2) |
| 330 |
*/ |
| 331 |
n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 332 |
if (c == '.' && !(IsIdentifierStart(n) || n == '.')) { |
| 333 |
error |= base != 10; |
| 334 |
pos++; |
| 335 |
/* It's ok to have no digits after the period. */ |
| 336 |
ScanDigits(styler, pos, 10); |
| 337 |
} |
| 338 |
|
| 339 |
/* Look for the exponentiation. */ |
| 340 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 341 |
if (c == 'e' || c == 'E') { |
| 342 |
error |= base != 10; |
| 343 |
pos++; |
| 344 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 345 |
if (c == '-' || c == '+') |
| 346 |
pos++; |
| 347 |
/* It is invalid to have no digits in the exponent. */ |
| 348 |
error |= !ScanDigits(styler, pos, 10); |
| 349 |
} |
| 350 |
|
| 351 |
/* Scan the floating point suffix. */ |
| 352 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 353 |
if (c == 'f') { |
| 354 |
error |= base != 10; |
| 355 |
pos++; |
| 356 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 357 |
n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 358 |
if (c == '3' && n == '2') { |
| 359 |
pos += 2; |
| 360 |
} else if (c == '6' && n == '4') { |
| 361 |
pos += 2; |
| 362 |
} else { |
| 363 |
error = true; |
| 364 |
} |
| 365 |
} |
| 366 |
} |
| 367 |
|
| 368 |
if (error) |
| 369 |
styler.ColourTo(pos - 1, SCE_RUST_LEXERROR); |
| 370 |
else |
| 371 |
styler.ColourTo(pos - 1, SCE_RUST_NUMBER); |
| 372 |
} |
| 373 |
|
| 374 |
static bool IsOneCharOperator(int c) { |
| 375 |
return c == ';' || c == ',' || c == '(' || c == ')' |
| 376 |
|| c == '{' || c == '}' || c == '[' || c == ']' |
| 377 |
|| c == '@' || c == '#' || c == '~' || c == '+' |
| 378 |
|| c == '*' || c == '/' || c == '^' || c == '%' |
| 379 |
|| c == '.' || c == ':' || c == '!' || c == '<' |
| 380 |
|| c == '>' || c == '=' || c == '-' || c == '&' |
| 381 |
|| c == '|' || c == '$' || c == '?'; |
| 382 |
} |
| 383 |
|
| 384 |
static bool IsTwoCharOperator(int c, int n) { |
| 385 |
return (c == '.' && n == '.') || (c == ':' && n == ':') |
| 386 |
|| (c == '!' && n == '=') || (c == '<' && n == '<') |
| 387 |
|| (c == '<' && n == '=') || (c == '>' && n == '>') |
| 388 |
|| (c == '>' && n == '=') || (c == '=' && n == '=') |
| 389 |
|| (c == '=' && n == '>') || (c == '-' && n == '>') |
| 390 |
|| (c == '&' && n == '&') || (c == '|' && n == '|') |
| 391 |
|| (c == '-' && n == '=') || (c == '&' && n == '=') |
| 392 |
|| (c == '|' && n == '=') || (c == '+' && n == '=') |
| 393 |
|| (c == '*' && n == '=') || (c == '/' && n == '=') |
| 394 |
|| (c == '^' && n == '=') || (c == '%' && n == '='); |
| 395 |
} |
| 396 |
|
| 397 |
static bool IsThreeCharOperator(int c, int n, int n2) { |
| 398 |
return (c == '<' && n == '<' && n2 == '=') |
| 399 |
|| (c == '>' && n == '>' && n2 == '='); |
| 400 |
} |
| 401 |
|
| 402 |
static bool IsValidCharacterEscape(int c) { |
| 403 |
return c == 'n' || c == 'r' || c == 't' || c == '\\' |
| 404 |
|| c == '\'' || c == '"' || c == '0'; |
| 405 |
} |
| 406 |
|
| 407 |
static bool IsValidStringEscape(int c) { |
| 408 |
return IsValidCharacterEscape(c) || c == '\n' || c == '\r'; |
| 409 |
} |
| 410 |
|
| 411 |
static bool ScanNumericEscape(Accessor &styler, Sci_Position& pos, Sci_Position num_digits, bool stop_asap) { |
| 412 |
for (;;) { |
| 413 |
const int c = styler.SafeGetCharAt(pos, '\0'); |
| 414 |
if (!IsADigit(c, 16)) |
| 415 |
break; |
| 416 |
num_digits--; |
| 417 |
pos++; |
| 418 |
if (num_digits == 0 && stop_asap) |
| 419 |
return true; |
| 420 |
} |
| 421 |
if (num_digits == 0) { |
| 422 |
return true; |
| 423 |
} else { |
| 424 |
return false; |
| 425 |
} |
| 426 |
} |
| 427 |
|
| 428 |
/* This is overly permissive for character literals in order to accept UTF-8 encoded |
| 429 |
* character literals. */ |
| 430 |
static void ScanCharacterLiteralOrLifetime(Accessor &styler, Sci_Position& pos, bool ascii_only) { |
| 431 |
pos++; |
| 432 |
int c = styler.SafeGetCharAt(pos, '\0'); |
| 433 |
int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 434 |
bool done = false; |
| 435 |
bool valid_lifetime = !ascii_only && IsIdentifierStart(c); |
| 436 |
bool valid_char = true; |
| 437 |
bool first = true; |
| 438 |
while (!done) { |
| 439 |
switch (c) { |
| 440 |
case '\\': |
| 441 |
done = true; |
| 442 |
if (IsValidCharacterEscape(n)) { |
| 443 |
pos += 2; |
| 444 |
} else if (n == 'x') { |
| 445 |
pos += 2; |
| 446 |
valid_char = ScanNumericEscape(styler, pos, 2, false); |
| 447 |
} else if (n == 'u' && !ascii_only) { |
| 448 |
pos += 2; |
| 449 |
if (styler.SafeGetCharAt(pos, '\0') != '{') { |
| 450 |
// old-style |
| 451 |
valid_char = ScanNumericEscape(styler, pos, 4, false); |
| 452 |
} else { |
| 453 |
int n_digits = 0; |
| 454 |
while (IsADigit(styler.SafeGetCharAt(++pos, '\0'), 16) && n_digits++ < 6) { |
| 455 |
} |
| 456 |
if (n_digits > 0 && styler.SafeGetCharAt(pos, '\0') == '}') |
| 457 |
pos++; |
| 458 |
else |
| 459 |
valid_char = false; |
| 460 |
} |
| 461 |
} else if (n == 'U' && !ascii_only) { |
| 462 |
pos += 2; |
| 463 |
valid_char = ScanNumericEscape(styler, pos, 8, false); |
| 464 |
} else { |
| 465 |
valid_char = false; |
| 466 |
} |
| 467 |
break; |
| 468 |
case '\'': |
| 469 |
valid_char = !first; |
| 470 |
done = true; |
| 471 |
break; |
| 472 |
case '\t': |
| 473 |
case '\n': |
| 474 |
case '\r': |
| 475 |
case '\0': |
| 476 |
valid_char = false; |
| 477 |
done = true; |
| 478 |
break; |
| 479 |
default: |
| 480 |
if (ascii_only && !IsASCII((char)c)) { |
| 481 |
done = true; |
| 482 |
valid_char = false; |
| 483 |
} else if (!IsIdentifierContinue(c) && !first) { |
| 484 |
done = true; |
| 485 |
} else { |
| 486 |
pos++; |
| 487 |
} |
| 488 |
break; |
| 489 |
} |
| 490 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 491 |
n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 492 |
|
| 493 |
first = false; |
| 494 |
} |
| 495 |
if (styler.SafeGetCharAt(pos, '\0') == '\'') { |
| 496 |
valid_lifetime = false; |
| 497 |
} else { |
| 498 |
valid_char = false; |
| 499 |
} |
| 500 |
if (valid_lifetime) { |
| 501 |
styler.ColourTo(pos - 1, SCE_RUST_LIFETIME); |
| 502 |
} else if (valid_char) { |
| 503 |
pos++; |
| 504 |
styler.ColourTo(pos - 1, ascii_only ? SCE_RUST_BYTECHARACTER : SCE_RUST_CHARACTER); |
| 505 |
} else { |
| 506 |
styler.ColourTo(pos - 1, SCE_RUST_LEXERROR); |
| 507 |
} |
| 508 |
} |
| 509 |
|
| 510 |
enum CommentState { |
| 511 |
UnknownComment, |
| 512 |
DocComment, |
| 513 |
NotDocComment |
| 514 |
}; |
| 515 |
|
| 516 |
/* |
| 517 |
* 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. |
| 518 |
* Otherwise it's a regular comment. |
| 519 |
*/ |
| 520 |
static void ResumeBlockComment(Accessor &styler, Sci_Position& pos, Sci_Position max, CommentState state, int level) { |
| 521 |
int c = styler.SafeGetCharAt(pos, '\0'); |
| 522 |
bool maybe_doc_comment = false; |
| 523 |
if (c == '*') { |
| 524 |
const int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 525 |
if (n != '*' && n != '/') { |
| 526 |
maybe_doc_comment = true; |
| 527 |
} |
| 528 |
} else if (c == '!') { |
| 529 |
maybe_doc_comment = true; |
| 530 |
} |
| 531 |
|
| 532 |
for (;;) { |
| 533 |
const int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 534 |
if (pos == styler.LineEnd(styler.GetLine(pos))) |
| 535 |
styler.SetLineState(styler.GetLine(pos), level); |
| 536 |
if (c == '*') { |
| 537 |
pos++; |
| 538 |
if (n == '/') { |
| 539 |
pos++; |
| 540 |
level--; |
| 541 |
if (level == 0) { |
| 542 |
styler.SetLineState(styler.GetLine(pos), 0); |
| 543 |
if (state == DocComment || (state == UnknownComment && maybe_doc_comment)) |
| 544 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCKDOC); |
| 545 |
else |
| 546 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCK); |
| 547 |
break; |
| 548 |
} |
| 549 |
} |
| 550 |
} else if (c == '/') { |
| 551 |
pos++; |
| 552 |
if (n == '*') { |
| 553 |
pos++; |
| 554 |
level++; |
| 555 |
} |
| 556 |
} |
| 557 |
else if (pos < max) { |
| 558 |
pos++; |
| 559 |
} |
| 560 |
if (pos >= max) { |
| 561 |
if (state == DocComment || (state == UnknownComment && maybe_doc_comment)) |
| 562 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCKDOC); |
| 563 |
else |
| 564 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTBLOCK); |
| 565 |
break; |
| 566 |
} |
| 567 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 568 |
} |
| 569 |
} |
| 570 |
|
| 571 |
/* |
| 572 |
* The rule for line-doc comments is as follows... ///N and //! (where N is a non slash) start doc comments. |
| 573 |
* Otherwise it's a normal line comment. |
| 574 |
*/ |
| 575 |
static void ResumeLineComment(Accessor &styler, Sci_Position& pos, Sci_Position max, CommentState state) { |
| 576 |
bool maybe_doc_comment = false; |
| 577 |
int c = styler.SafeGetCharAt(pos, '\0'); |
| 578 |
if (c == '/') { |
| 579 |
if (pos < max) { |
| 580 |
pos++; |
| 581 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 582 |
if (c != '/') { |
| 583 |
maybe_doc_comment = true; |
| 584 |
} |
| 585 |
} |
| 586 |
} else if (c == '!') { |
| 587 |
maybe_doc_comment = true; |
| 588 |
} |
| 589 |
|
| 590 |
pos = styler.LineEnd(styler.GetLine(pos)); |
| 591 |
styler.SetLineState(styler.GetLine(pos), SCE_RUST_DEFAULT); |
| 592 |
|
| 593 |
if (state == DocComment || (state == UnknownComment && maybe_doc_comment)) |
| 594 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTLINEDOC); |
| 595 |
else |
| 596 |
styler.ColourTo(pos - 1, SCE_RUST_COMMENTLINE); |
| 597 |
} |
| 598 |
|
| 599 |
static void ScanComments(Accessor &styler, Sci_Position& pos, Sci_Position max) { |
| 600 |
pos++; |
| 601 |
const int c = styler.SafeGetCharAt(pos, '\0'); |
| 602 |
pos++; |
| 603 |
if (c == '/') |
| 604 |
ResumeLineComment(styler, pos, max, UnknownComment); |
| 605 |
else if (c == '*') |
| 606 |
ResumeBlockComment(styler, pos, max, UnknownComment, 1); |
| 607 |
} |
| 608 |
|
| 609 |
static void ResumeString(Accessor &styler, Sci_Position& pos, Sci_Position max, StringType string_type) { |
| 610 |
int c = styler.SafeGetCharAt(pos, '\0'); |
| 611 |
bool error = false; |
| 612 |
while (c != '"' && !error) { |
| 613 |
if (pos >= max) { |
| 614 |
error = true; |
| 615 |
break; |
| 616 |
} |
| 617 |
if (pos == styler.LineEnd(styler.GetLine(pos))) |
| 618 |
styler.SetLineState(styler.GetLine(pos), 0); |
| 619 |
if (c == '\\') { |
| 620 |
const int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 621 |
if (IsValidStringEscape(n)) { |
| 622 |
pos += 2; |
| 623 |
} else if (n == 'x') { |
| 624 |
pos += 2; |
| 625 |
error = !ScanNumericEscape(styler, pos, 2, true); |
| 626 |
} else if (n == 'u' && (string_type != StringType::BYTESTRING)) { |
| 627 |
pos += 2; |
| 628 |
if (styler.SafeGetCharAt(pos, '\0') != '{') { |
| 629 |
// old-style |
| 630 |
error = !ScanNumericEscape(styler, pos, 4, true); |
| 631 |
} else { |
| 632 |
int n_digits = 0; |
| 633 |
while (IsADigit(styler.SafeGetCharAt(++pos, '\0'), 16) && n_digits++ < 6) { |
| 634 |
} |
| 635 |
if (n_digits > 0 && styler.SafeGetCharAt(pos, '\0') == '}') |
| 636 |
pos++; |
| 637 |
else |
| 638 |
error = true; |
| 639 |
} |
| 640 |
} else if (n == 'U' && (string_type != StringType::BYTESTRING)) { |
| 641 |
pos += 2; |
| 642 |
error = !ScanNumericEscape(styler, pos, 8, true); |
| 643 |
} else { |
| 644 |
pos += 1; |
| 645 |
error = true; |
| 646 |
} |
| 647 |
} else { |
| 648 |
if (string_type == StringType::BYTESTRING && !IsASCII((char)c)) |
| 649 |
error = true; |
| 650 |
else |
| 651 |
pos++; |
| 652 |
} |
| 653 |
c = styler.SafeGetCharAt(pos, '\0'); |
| 654 |
} |
| 655 |
if (!error) |
| 656 |
pos++; |
| 657 |
|
| 658 |
styler.ColourTo(pos - 1, static_cast<int>(string_type)); |
| 659 |
} |
| 660 |
|
| 661 |
static void ResumeRawString(Accessor &styler, Sci_Position& pos, Sci_Position max, int num_hashes, StringType string_type) { |
| 662 |
for (;;) { |
| 663 |
if (pos == styler.LineEnd(styler.GetLine(pos))) |
| 664 |
styler.SetLineState(styler.GetLine(pos), num_hashes); |
| 665 |
|
| 666 |
const int c = styler.SafeGetCharAt(pos, '\0'); |
| 667 |
if (c == '"') { |
| 668 |
pos++; |
| 669 |
int trailing_num_hashes = 0; |
| 670 |
while (styler.SafeGetCharAt(pos, '\0') == '#' && trailing_num_hashes < num_hashes) { |
| 671 |
trailing_num_hashes++; |
| 672 |
pos++; |
| 673 |
} |
| 674 |
if (trailing_num_hashes == num_hashes) { |
| 675 |
styler.SetLineState(styler.GetLine(pos), 0); |
| 676 |
break; |
| 677 |
} |
| 678 |
} else if (pos >= max) { |
| 679 |
break; |
| 680 |
} else { |
| 681 |
if ((string_type == StringType::RAW_BYTESTRING) && !IsASCII((char)c)) |
| 682 |
break; |
| 683 |
pos++; |
| 684 |
} |
| 685 |
} |
| 686 |
|
| 687 |
styler.ColourTo(pos - 1, static_cast<int>(string_type)); |
| 688 |
} |
| 689 |
|
| 690 |
static void ScanRawString(Accessor &styler, Sci_Position& pos, Sci_Position max, StringType string_type) { |
| 691 |
pos++; |
| 692 |
int num_hashes = 0; |
| 693 |
while (styler.SafeGetCharAt(pos, '\0') == '#') { |
| 694 |
num_hashes++; |
| 695 |
pos++; |
| 696 |
} |
| 697 |
if (styler.SafeGetCharAt(pos, '\0') != '"') { |
| 698 |
styler.ColourTo(pos - 1, SCE_RUST_LEXERROR); |
| 699 |
} else { |
| 700 |
pos++; |
| 701 |
ResumeRawString(styler, pos, max, num_hashes, string_type); |
| 702 |
} |
| 703 |
} |
| 704 |
|
| 705 |
void SCI_METHOD LexerRust::Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) { |
| 706 |
PropSetSimple props; |
| 707 |
Accessor styler(pAccess, &props); |
| 708 |
Sci_Position pos = startPos; |
| 709 |
const Sci_Position max = pos + length; |
| 710 |
|
| 711 |
styler.StartAt(pos); |
| 712 |
styler.StartSegment(pos); |
| 713 |
|
| 714 |
if (initStyle == SCE_RUST_COMMENTBLOCK || initStyle == SCE_RUST_COMMENTBLOCKDOC) { |
| 715 |
ResumeBlockComment(styler, pos, max, initStyle == SCE_RUST_COMMENTBLOCKDOC ? DocComment : NotDocComment, styler.GetLineState(styler.GetLine(pos) - 1)); |
| 716 |
} else if (initStyle == SCE_RUST_COMMENTLINE || initStyle == SCE_RUST_COMMENTLINEDOC) { |
| 717 |
ResumeLineComment(styler, pos, max, initStyle == SCE_RUST_COMMENTLINEDOC ? DocComment : NotDocComment); |
| 718 |
} else if (initStyle == SCE_RUST_STRING) { |
| 719 |
ResumeString(styler, pos, max, StringType::STRING); |
| 720 |
} else if (initStyle == SCE_RUST_BYTESTRING) { |
| 721 |
ResumeString(styler, pos, max, StringType::BYTESTRING); |
| 722 |
} else if (initStyle == SCE_RUST_STRINGR) { |
| 723 |
ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), StringType::RAW_STRING); |
| 724 |
} else if (initStyle == SCE_RUST_BYTESTRINGR) { |
| 725 |
ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), StringType::RAW_BYTESTRING); |
| 726 |
} else if (initStyle == SCE_RUST_CSTRING) { |
| 727 |
ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), StringType::CSTRING); |
| 728 |
} else if (initStyle == SCE_RUST_CSTRINGR) { |
| 729 |
ResumeRawString(styler, pos, max, styler.GetLineState(styler.GetLine(pos) - 1), StringType::RAW_CSTRING); |
| 730 |
} |
| 731 |
while (pos < max) { |
| 732 |
const int c = styler.SafeGetCharAt(pos, '\0'); |
| 733 |
const int n = styler.SafeGetCharAt(pos + 1, '\0'); |
| 734 |
const int n2 = styler.SafeGetCharAt(pos + 2, '\0'); |
| 735 |
|
| 736 |
if (pos == 0 && c == '#' && n == '!' && n2 != '[') { |
| 737 |
pos += 2; |
| 738 |
ResumeLineComment(styler, pos, max, NotDocComment); |
| 739 |
} else if (IsWhitespace(c)) { |
| 740 |
ScanWhitespace(styler, pos, max); |
| 741 |
} else if (c == '/' && (n == '/' || n == '*')) { |
| 742 |
ScanComments(styler, pos, max); |
| 743 |
} else if (c == 'r' && (n == '#' && IsIdentifierStart(n2))) { |
| 744 |
pos += 2; |
| 745 |
ScanRawIdentifier(styler, pos); |
| 746 |
} else if (c == 'r' && (n == '#' || n == '"')) { |
| 747 |
ScanRawString(styler, pos, max, StringType::RAW_STRING); |
| 748 |
} else if (c == 'b' && n == 'r' && (n2 == '#' || n2 == '"')) { |
| 749 |
pos++; |
| 750 |
ScanRawString(styler, pos, max, StringType::RAW_BYTESTRING); |
| 751 |
} else if (c == 'b' && n == '"') { |
| 752 |
pos += 2; |
| 753 |
ResumeString(styler, pos, max, StringType::BYTESTRING); |
| 754 |
} else if (c == 'c' && n == 'r' && (n2 == '#' || n2 == '"')) { |
| 755 |
pos++; |
| 756 |
ScanRawString(styler, pos, max, StringType::RAW_CSTRING); |
| 757 |
} else if (c == 'c' && n == '"') { |
| 758 |
pos += 2; |
| 759 |
ResumeString(styler, pos, max, StringType::CSTRING); |
| 760 |
} else if (c == 'b' && n == '\'') { |
| 761 |
pos++; |
| 762 |
ScanCharacterLiteralOrLifetime(styler, pos, true); |
| 763 |
} else if (IsIdentifierStart(c)) { |
| 764 |
ScanIdentifier(styler, pos, keywords); |
| 765 |
} else if (IsADigit(c)) { |
| 766 |
ScanNumber(styler, pos); |
| 767 |
} else if (IsThreeCharOperator(c, n, n2)) { |
| 768 |
pos += 3; |
| 769 |
styler.ColourTo(pos - 1, SCE_RUST_OPERATOR); |
| 770 |
} else if (IsTwoCharOperator(c, n)) { |
| 771 |
pos += 2; |
| 772 |
styler.ColourTo(pos - 1, SCE_RUST_OPERATOR); |
| 773 |
} else if (IsOneCharOperator(c)) { |
| 774 |
pos++; |
| 775 |
styler.ColourTo(pos - 1, SCE_RUST_OPERATOR); |
| 776 |
} else if (c == '\'') { |
| 777 |
ScanCharacterLiteralOrLifetime(styler, pos, false); |
| 778 |
} else if (c == '"') { |
| 779 |
pos++; |
| 780 |
ResumeString(styler, pos, max, StringType::STRING); |
| 781 |
} else { |
| 782 |
pos++; |
| 783 |
styler.ColourTo(pos - 1, SCE_RUST_LEXERROR); |
| 784 |
} |
| 785 |
} |
| 786 |
styler.ColourTo(pos - 1, SCE_RUST_DEFAULT); |
| 787 |
styler.Flush(); |
| 788 |
} |
| 789 |
|
| 790 |
void SCI_METHOD LexerRust::Fold(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) { |
| 791 |
|
| 792 |
if (!options.fold) |
| 793 |
return; |
| 794 |
|
| 795 |
LexAccessor styler(pAccess); |
| 796 |
|
| 797 |
const Sci_PositionU endPos = startPos + length; |
| 798 |
int visibleChars = 0; |
| 799 |
bool inLineComment = false; |
| 800 |
Sci_Position lineCurrent = styler.GetLine(startPos); |
| 801 |
int levelCurrent = SC_FOLDLEVELBASE; |
| 802 |
if (lineCurrent > 0) |
| 803 |
levelCurrent = styler.LevelAt(lineCurrent-1) >> 16; |
| 804 |
Sci_PositionU lineStartNext = styler.LineStart(lineCurrent+1); |
| 805 |
int levelMinCurrent = levelCurrent; |
| 806 |
int levelNext = levelCurrent; |
| 807 |
char chNext = styler[startPos]; |
| 808 |
int styleNext = styler.StyleAt(startPos); |
| 809 |
int style = initStyle; |
| 810 |
const bool userDefinedFoldMarkers = !options.foldExplicitStart.empty() && !options.foldExplicitEnd.empty(); |
| 811 |
for (Sci_PositionU i = startPos; i < endPos; i++) { |
| 812 |
const char ch = chNext; |
| 813 |
chNext = styler.SafeGetCharAt(i + 1); |
| 814 |
const int stylePrev = style; |
| 815 |
style = styleNext; |
| 816 |
styleNext = styler.StyleAt(i + 1); |
| 817 |
const bool atEOL = i == (lineStartNext-1); |
| 818 |
if ((style == SCE_RUST_COMMENTLINE) || (style == SCE_RUST_COMMENTLINEDOC)) |
| 819 |
inLineComment = true; |
| 820 |
if (options.foldComment && options.foldCommentMultiline && IsStreamCommentStyle(style) && !inLineComment) { |
| 821 |
if (!IsStreamCommentStyle(stylePrev)) { |
| 822 |
levelNext++; |
| 823 |
} else if (!IsStreamCommentStyle(styleNext) && !atEOL) { |
| 824 |
// Comments don't end at end of line and the next character may be unstyled. |
| 825 |
levelNext--; |
| 826 |
} |
| 827 |
} |
| 828 |
if (options.foldComment && options.foldCommentExplicit && ((style == SCE_RUST_COMMENTLINE) || options.foldExplicitAnywhere)) { |
| 829 |
if (userDefinedFoldMarkers) { |
| 830 |
if (styler.Match(i, options.foldExplicitStart.c_str())) { |
| 831 |
levelNext++; |
| 832 |
} else if (styler.Match(i, options.foldExplicitEnd.c_str())) { |
| 833 |
levelNext--; |
| 834 |
} |
| 835 |
} else { |
| 836 |
if ((ch == '/') && (chNext == '/')) { |
| 837 |
const char chNext2 = styler.SafeGetCharAt(i + 2); |
| 838 |
if (chNext2 == '{') { |
| 839 |
levelNext++; |
| 840 |
} else if (chNext2 == '}') { |
| 841 |
levelNext--; |
| 842 |
} |
| 843 |
} |
| 844 |
} |
| 845 |
} |
| 846 |
if (options.foldSyntaxBased && (style == SCE_RUST_OPERATOR)) { |
| 847 |
if (ch == '{') { |
| 848 |
// Measure the minimum before a '{' to allow |
| 849 |
// folding on "} else {" |
| 850 |
if (levelMinCurrent > levelNext) { |
| 851 |
levelMinCurrent = levelNext; |
| 852 |
} |
| 853 |
levelNext++; |
| 854 |
} else if (ch == '}') { |
| 855 |
levelNext--; |
| 856 |
} |
| 857 |
} |
| 858 |
if (!IsASpace(ch)) |
| 859 |
visibleChars++; |
| 860 |
if (atEOL || (i == endPos-1)) { |
| 861 |
int levelUse = levelCurrent; |
| 862 |
if (options.foldSyntaxBased && options.foldAtElse) { |
| 863 |
levelUse = levelMinCurrent; |
| 864 |
} |
| 865 |
int lev = levelUse | levelNext << 16; |
| 866 |
if (visibleChars == 0 && options.foldCompact) |
| 867 |
lev |= SC_FOLDLEVELWHITEFLAG; |
| 868 |
if (levelUse < levelNext) |
| 869 |
lev |= SC_FOLDLEVELHEADERFLAG; |
| 870 |
if (lev != styler.LevelAt(lineCurrent)) { |
| 871 |
styler.SetLevel(lineCurrent, lev); |
| 872 |
} |
| 873 |
lineCurrent++; |
| 874 |
lineStartNext = styler.LineStart(lineCurrent+1); |
| 875 |
levelCurrent = levelNext; |
| 876 |
levelMinCurrent = levelCurrent; |
| 877 |
if (atEOL && (i == static_cast<Sci_PositionU>(styler.Length()-1))) { |
| 878 |
// There is an empty line at end of file so give it same level and empty |
| 879 |
styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG); |
| 880 |
} |
| 881 |
visibleChars = 0; |
| 882 |
inLineComment = false; |
| 883 |
} |
| 884 |
} |
| 885 |
} |
| 886 |
|
| 887 |
} |
| 888 |
|
| 889 |
extern const LexerModule lmRust(SCLEX_RUST, LexerRust::LexerFactoryRust, "rust", rustWordLists); |