| 1 | /** |
| 2 | * @file LexFSharp.cxx |
| 3 | * Lexer for F# 5.0 |
| 4 | * Copyright (c) 2021 Robert Di Pardo <dipardo.r@gmail.com> |
| 5 | * Parts of LexerFSharp::Lex were adapted from LexCaml.cxx by Robert Roessler ("RR"). |
| 6 | * Parts of LexerFSharp::Fold were adapted from LexCPP.cxx by Neil Hodgson and Udo Lechner. |
| 7 | * The License.txt file describes the conditions under which this software may be distributed. |
| 8 | */ |
| 9 | // clang-format off |
| 10 | #include <cstdlib> |
| 11 | #include <cassert> |
| 12 | |
| 13 | #include <string> |
| 14 | #include <string_view> |
| 15 | #include <map> |
| 16 | #include <functional> |
| 17 | |
| 18 | #include "ILexer.h" |
| 19 | #include "Scintilla.h" |
| 20 | #include "SciLexer.h" |
| 21 | |
| 22 | #include "WordList.h" |
| 23 | #include "LexAccessor.h" |
| 24 | #include "StyleContext.h" |
| 25 | #include "CharacterSet.h" |
| 26 | #include "LexerModule.h" |
| 27 | #include "OptionSet.h" |
| 28 | #include "DefaultLexer.h" |
| 29 | // clang-format on |
| 30 | |
| 31 | using namespace Scintilla; |
| 32 | using namespace Lexilla; |
| 33 | |
| 34 | static const char *const lexerName = "fsharp" ; |
| 35 | static constexpr int WORDLIST_SIZE = 5; |
| 36 | static const char *const fsharpWordLists[] = { |
| 37 | "standard language keywords" , |
| 38 | "core functions, including those in the FSharp.Collections namespace" , |
| 39 | "built-in types, core namespaces, modules" , |
| 40 | "optional" , |
| 41 | "optional" , |
| 42 | nullptr, |
| 43 | }; |
| 44 | static constexpr int keywordClasses[] = { |
| 45 | SCE_FSHARP_KEYWORD, SCE_FSHARP_KEYWORD2, SCE_FSHARP_KEYWORD3, SCE_FSHARP_KEYWORD4, SCE_FSHARP_KEYWORD5, |
| 46 | }; |
| 47 | |
| 48 | namespace { |
| 49 | |
| 50 | struct OptionsFSharp { |
| 51 | bool fold; |
| 52 | bool foldCompact; |
| 53 | bool ; |
| 54 | bool ; |
| 55 | bool ; |
| 56 | bool foldPreprocessor; |
| 57 | bool foldImports; |
| 58 | OptionsFSharp() { |
| 59 | fold = true; |
| 60 | foldCompact = true; |
| 61 | foldComment = true; |
| 62 | foldCommentStream = true; |
| 63 | foldCommentMultiLine = true; |
| 64 | foldPreprocessor = false; |
| 65 | foldImports = true; |
| 66 | } |
| 67 | }; |
| 68 | |
| 69 | struct OptionSetFSharp : public OptionSet<OptionsFSharp> { |
| 70 | OptionSetFSharp() { |
| 71 | DefineProperty("fold" , &OptionsFSharp::fold); |
| 72 | DefineProperty("fold.compact" , &OptionsFSharp::foldCompact); |
| 73 | DefineProperty("fold.comment" , &OptionsFSharp::foldComment, |
| 74 | "Setting this option to 0 disables comment folding in F# files." ); |
| 75 | |
| 76 | DefineProperty("fold.fsharp.comment.stream" , &OptionsFSharp::foldCommentStream, |
| 77 | "Setting this option to 0 disables folding of ML-style comments in F# files when " |
| 78 | "fold.comment=1." ); |
| 79 | |
| 80 | DefineProperty("fold.fsharp.comment.multiline" , &OptionsFSharp::foldCommentMultiLine, |
| 81 | "Setting this option to 0 disables folding of grouped line comments in F# files when " |
| 82 | "fold.comment=1." ); |
| 83 | |
| 84 | DefineProperty("fold.fsharp.preprocessor" , &OptionsFSharp::foldPreprocessor, |
| 85 | "Setting this option to 1 enables folding of F# compiler directives." ); |
| 86 | |
| 87 | DefineProperty("fold.fsharp.imports" , &OptionsFSharp::foldImports, |
| 88 | "Setting this option to 0 disables folding of F# import declarations." ); |
| 89 | |
| 90 | DefineWordListSets(fsharpWordLists); |
| 91 | } |
| 92 | }; |
| 93 | |
| 94 | const CharacterSet setOperators = CharacterSet(CharacterSet::setNone, "~^'-+*/%=@|&<>()[]{};,:!?" ); |
| 95 | const CharacterSet setClosingTokens = CharacterSet(CharacterSet::setNone, ")}]" ); |
| 96 | const CharacterSet setFormatSpecs = CharacterSet(CharacterSet::setNone, ".%aAbcdeEfFgGiMoOstuxX0123456789" ); |
| 97 | const CharacterSet setFormatFlags = CharacterSet(CharacterSet::setNone, ".-+0 " ); |
| 98 | const CharacterSet numericMetaChars1 = CharacterSet(CharacterSet::setNone, "_IbeEflmnosuxy" ); |
| 99 | const CharacterSet numericMetaChars2 = CharacterSet(CharacterSet::setNone, "lnsy" ); |
| 100 | std::map<int, int> numericPrefixes = { { 'b', 2 }, { 'o', 8 }, { 'x', 16 } }; |
| 101 | constexpr Sci_Position ZERO_LENGTH = -1; |
| 102 | |
| 103 | struct FSharpString { |
| 104 | Sci_Position startPos; |
| 105 | int startChar; |
| 106 | FSharpString() { |
| 107 | startPos = ZERO_LENGTH; |
| 108 | startChar = '"'; |
| 109 | } |
| 110 | constexpr bool HasLength() const { |
| 111 | return startPos > ZERO_LENGTH; |
| 112 | } |
| 113 | constexpr bool CanInterpolate() const { |
| 114 | return startChar == '$'; |
| 115 | } |
| 116 | }; |
| 117 | |
| 118 | class UnicodeChar { |
| 119 | enum class Notation { none, asciiDec, asciiHex, utf16, utf32 }; |
| 120 | Notation type = Notation::none; |
| 121 | // single-byte Unicode char (000 - 255) |
| 122 | int asciiDigits[3] = { 0 }; |
| 123 | int maxDigit = '9'; |
| 124 | int toEnd = 0; |
| 125 | bool invalid = false; |
| 126 | |
| 127 | public: |
| 128 | UnicodeChar() noexcept = default; |
| 129 | explicit UnicodeChar(const int prefix) { |
| 130 | if (IsADigit(prefix)) { |
| 131 | *asciiDigits = prefix; |
| 132 | if (*asciiDigits >= '0' && *asciiDigits <= '2') { |
| 133 | type = Notation::asciiDec; |
| 134 | // count first digit as "prefix" |
| 135 | toEnd = 2; |
| 136 | } |
| 137 | } else if (prefix == 'x' || prefix == 'u' || prefix == 'U') { |
| 138 | switch (prefix) { |
| 139 | case 'x': |
| 140 | type = Notation::asciiHex; |
| 141 | toEnd = 2; |
| 142 | break; |
| 143 | case 'u': |
| 144 | type = Notation::utf16; |
| 145 | toEnd = 4; |
| 146 | break; |
| 147 | case 'U': |
| 148 | type = Notation::utf32; |
| 149 | toEnd = 8; |
| 150 | break; |
| 151 | } |
| 152 | } |
| 153 | } |
| 154 | void Parse(const int ch) { |
| 155 | invalid = false; |
| 156 | switch (type) { |
| 157 | case Notation::asciiDec: { |
| 158 | maxDigit = (*asciiDigits < '2') ? '9' : (asciiDigits[1] <= '4') ? '9' : '5'; |
| 159 | if (IsADigit(ch) && asciiDigits[1] <= maxDigit && ch <= maxDigit) { |
| 160 | asciiDigits[1] = ch; |
| 161 | toEnd--; |
| 162 | } else { |
| 163 | invalid = true; |
| 164 | } |
| 165 | break; |
| 166 | } |
| 167 | case Notation::asciiHex: |
| 168 | case Notation::utf16: |
| 169 | if (IsADigit(ch, 16)) { |
| 170 | toEnd--; |
| 171 | } else { |
| 172 | invalid = true; |
| 173 | } |
| 174 | break; |
| 175 | case Notation::utf32: |
| 176 | if ((toEnd > 6 && ch == '0') || (toEnd <= 6 && IsADigit(ch, 16))) { |
| 177 | toEnd--; |
| 178 | } else { |
| 179 | invalid = true; |
| 180 | } |
| 181 | break; |
| 182 | case Notation::none: |
| 183 | break; |
| 184 | } |
| 185 | } |
| 186 | constexpr bool AtEnd() noexcept { |
| 187 | return invalid || type == Notation::none || (type != Notation::none && toEnd < 0); |
| 188 | } |
| 189 | }; |
| 190 | |
| 191 | inline bool (StyleContext &cxt) { |
| 192 | // match (* ... *), but allow point-free usage of the `*` operator, |
| 193 | // e.g. List.fold (*) 1 [ 1; 2; 3 ] |
| 194 | return (cxt.Match('(', '*') && cxt.GetRelative(2) != ')'); |
| 195 | } |
| 196 | |
| 197 | inline bool (const StyleContext &cxt) { |
| 198 | return (cxt.ch == ')' && cxt.chPrev == '*'); |
| 199 | } |
| 200 | |
| 201 | inline bool (const StyleContext &cxt) { |
| 202 | // style shebang lines as comments in F# scripts: |
| 203 | // https://fsharp.org/specs/language-spec/4.1/FSharpSpec-4.1-latest.pdf#page=30&zoom=auto,-98,537 |
| 204 | return cxt.Match('/', '/') || cxt.Match('#', '!'); |
| 205 | } |
| 206 | |
| 207 | inline bool MatchLineNumberStart(StyleContext &cxt) { |
| 208 | return cxt.atLineStart && (cxt.MatchIgnoreCase("#line" ) || |
| 209 | (cxt.ch == '#' && (IsADigit(cxt.chNext) || IsADigit(cxt.GetRelative(2))))); |
| 210 | } |
| 211 | |
| 212 | inline bool MatchPPDirectiveStart(const StyleContext &cxt) { |
| 213 | return (cxt.atLineStart && cxt.ch == '#' && iswordstart(cxt.chNext)); |
| 214 | } |
| 215 | |
| 216 | inline bool MatchTypeAttributeStart(const StyleContext &cxt) { |
| 217 | return cxt.Match('[', '<'); |
| 218 | } |
| 219 | |
| 220 | inline bool MatchTypeAttributeEnd(const StyleContext &cxt) { |
| 221 | return (cxt.ch == ']' && cxt.chPrev == '>'); |
| 222 | } |
| 223 | |
| 224 | inline bool MatchQuotedExpressionStart(const StyleContext &cxt) { |
| 225 | return cxt.Match('<', '@'); |
| 226 | } |
| 227 | |
| 228 | inline bool MatchQuotedExpressionEnd(const StyleContext &cxt) { |
| 229 | return (cxt.ch == '>' && cxt.chPrev == '@'); |
| 230 | } |
| 231 | |
| 232 | inline bool MatchStringStart(const StyleContext &cxt) { |
| 233 | return (cxt.ch == '"' || cxt.Match('@', '"') || cxt.Match('$', '"') || cxt.Match('`', '`')); |
| 234 | } |
| 235 | |
| 236 | inline bool FollowsEscapedBackslash(StyleContext &cxt) { |
| 237 | int count = 0; |
| 238 | for (Sci_Position offset = 1; cxt.GetRelative(-offset) == '\\'; offset++) |
| 239 | count++; |
| 240 | return count % 2 != 0; |
| 241 | } |
| 242 | |
| 243 | inline bool MatchStringEnd(StyleContext &cxt, const FSharpString &fsStr) { |
| 244 | return (fsStr.HasLength() && |
| 245 | // end of quoted identifier? |
| 246 | ((cxt.ch == '`' && cxt.chPrev == '`') || |
| 247 | // end of literal or interpolated triple-quoted string? |
| 248 | ((fsStr.startChar == '"' || (fsStr.CanInterpolate() && cxt.chPrev != '$')) && |
| 249 | cxt.MatchIgnoreCase("\"\"\"" )) || |
| 250 | // end of verbatim string? |
| 251 | (fsStr.startChar == '@' && |
| 252 | // embedded quotes must be in pairs |
| 253 | cxt.ch == '"' && cxt.chNext != '"' && |
| 254 | (cxt.chPrev != '"' || (cxt.chPrev == '"' && |
| 255 | // empty verbatim string? |
| 256 | (cxt.GetRelative(-2) == '@' || |
| 257 | // pair of quotes at end of string? |
| 258 | (cxt.GetRelative(-2) == '"' && cxt.GetRelative(-3) != '@'))))))) || |
| 259 | (!fsStr.HasLength() && cxt.ch == '"' && |
| 260 | ((cxt.chPrev != '\\' || (cxt.GetRelative(-2) == '\\' && !FollowsEscapedBackslash(cxt))) || |
| 261 | // treat backslashes as char literals in verbatim strings |
| 262 | (fsStr.startChar == '@' && cxt.chPrev == '\\'))); |
| 263 | } |
| 264 | |
| 265 | inline bool MatchCharacterStart(StyleContext &cxt) { |
| 266 | // don't style generic type parameters: 'a, 'b, 'T, etc. |
| 267 | return (cxt.ch == '\'' && !(cxt.chPrev == ':' || cxt.GetRelative(-2) == ':')); |
| 268 | } |
| 269 | |
| 270 | inline bool CanEmbedQuotes(StyleContext &cxt) { |
| 271 | // allow unescaped double quotes inside literal or interpolated triple-quoted strings, verbatim strings, |
| 272 | // and quoted identifiers: |
| 273 | // - https://docs.microsoft.com/en-us/dotnet/fsharp/language-reference/strings |
| 274 | // - https://docs.microsoft.com/en-us/dotnet/fsharp/language-reference/interpolated-strings#syntax |
| 275 | // - https://fsharp.org/specs/language-spec/4.1/FSharpSpec-4.1-latest.pdf#page=25&zoom=auto,-98,600 |
| 276 | return cxt.MatchIgnoreCase("$\"\"\"" ) || cxt.MatchIgnoreCase("\"\"\"" ) || cxt.Match('@', '"') || |
| 277 | cxt.Match('`', '`'); |
| 278 | } |
| 279 | |
| 280 | inline bool IsNumber(StyleContext &cxt, const int base = 10) { |
| 281 | return IsADigit(cxt.ch, base) || (IsADigit(cxt.chPrev, base) && numericMetaChars1.Contains(cxt.ch)) || |
| 282 | (IsADigit(cxt.GetRelative(-2), base) && numericMetaChars2.Contains(cxt.ch)); |
| 283 | } |
| 284 | |
| 285 | inline bool IsFloat(const StyleContext &cxt) { |
| 286 | return (cxt.ch == '.' && IsADigit(cxt.chPrev)) || |
| 287 | ((cxt.ch == '+' || cxt.ch == '-' ) && IsADigit(cxt.chNext)); |
| 288 | } |
| 289 | |
| 290 | inline bool IsLineEnd(StyleContext &cxt, const Sci_Position offset) { |
| 291 | const int ch = cxt.GetRelative(offset, '\n'); |
| 292 | return (ch == '\r' || ch == '\n'); |
| 293 | } |
| 294 | |
| 295 | class LexerFSharp : public DefaultLexer { |
| 296 | WordList keywords[WORDLIST_SIZE]; |
| 297 | OptionsFSharp options; |
| 298 | OptionSetFSharp optionSet; |
| 299 | |
| 300 | public: |
| 301 | explicit LexerFSharp() : DefaultLexer(lexerName, SCLEX_FSHARP) { |
| 302 | } |
| 303 | static ILexer5 *LexerFactoryFSharp() { |
| 304 | return new LexerFSharp(); |
| 305 | } |
| 306 | virtual ~LexerFSharp() { |
| 307 | } |
| 308 | void SCI_METHOD Release() noexcept override { |
| 309 | delete this; |
| 310 | } |
| 311 | int SCI_METHOD Version() const noexcept override { |
| 312 | return lvRelease5; |
| 313 | } |
| 314 | const char *SCI_METHOD GetName() noexcept override { |
| 315 | return lexerName; |
| 316 | } |
| 317 | int SCI_METHOD GetIdentifier() noexcept override { |
| 318 | return SCLEX_FSHARP; |
| 319 | } |
| 320 | int SCI_METHOD LineEndTypesSupported() noexcept override { |
| 321 | return SC_LINE_END_TYPE_DEFAULT; |
| 322 | } |
| 323 | void *SCI_METHOD PrivateCall(int, void *) noexcept override { |
| 324 | return nullptr; |
| 325 | } |
| 326 | const char *SCI_METHOD DescribeWordListSets() override { |
| 327 | return optionSet.DescribeWordListSets(); |
| 328 | } |
| 329 | const char *SCI_METHOD PropertyNames() override { |
| 330 | return optionSet.PropertyNames(); |
| 331 | } |
| 332 | int SCI_METHOD PropertyType(const char *name) override { |
| 333 | return optionSet.PropertyType(name); |
| 334 | } |
| 335 | const char *SCI_METHOD DescribeProperty(const char *name) override { |
| 336 | return optionSet.DescribeProperty(name); |
| 337 | } |
| 338 | const char *SCI_METHOD PropertyGet(const char *key) override { |
| 339 | return optionSet.PropertyGet(key); |
| 340 | } |
| 341 | Sci_Position SCI_METHOD PropertySet(const char *key, const char *val) override { |
| 342 | if (optionSet.PropertySet(&options, key, val)) { |
| 343 | return 0; |
| 344 | } |
| 345 | return ZERO_LENGTH; |
| 346 | } |
| 347 | Sci_Position SCI_METHOD WordListSet(int n, const char *wl) override; |
| 348 | void SCI_METHOD Lex(Sci_PositionU start, Sci_Position length, int initStyle, IDocument *pAccess) override; |
| 349 | void SCI_METHOD Fold(Sci_PositionU start, Sci_Position length, int initStyle,IDocument *pAccess) override; |
| 350 | }; |
| 351 | |
| 352 | Sci_Position SCI_METHOD LexerFSharp::WordListSet(int n, const char *wl) { |
| 353 | WordList *wordListN = nullptr; |
| 354 | Sci_Position firstModification = ZERO_LENGTH; |
| 355 | |
| 356 | if (n < WORDLIST_SIZE) { |
| 357 | wordListN = &keywords[n]; |
| 358 | } |
| 359 | if (wordListN) { |
| 360 | WordList wlNew; |
| 361 | wlNew.Set(wl); |
| 362 | |
| 363 | if (*wordListN != wlNew) { |
| 364 | wordListN->Set(wl); |
| 365 | firstModification = 0; |
| 366 | } |
| 367 | } |
| 368 | return firstModification; |
| 369 | } |
| 370 | |
| 371 | void SCI_METHOD LexerFSharp::Lex(Sci_PositionU start, Sci_Position length, int initStyle, IDocument *pAccess) { |
| 372 | LexAccessor styler(pAccess); |
| 373 | StyleContext sc(start, static_cast<Sci_PositionU>(length), initStyle, styler); |
| 374 | Sci_PositionU cursor = 0; |
| 375 | UnicodeChar uniCh = UnicodeChar(); |
| 376 | FSharpString fsStr = FSharpString(); |
| 377 | constexpr Sci_Position MAX_WORD_LEN = 64; |
| 378 | constexpr int SPACE = ' '; |
| 379 | int currentBase = 10; |
| 380 | |
| 381 | while (sc.More()) { |
| 382 | Sci_PositionU colorSpan = sc.currentPos - 1; |
| 383 | int state = -1; |
| 384 | bool advance = true; |
| 385 | |
| 386 | switch (sc.state & 0xff) { |
| 387 | case SCE_FSHARP_DEFAULT: |
| 388 | cursor = sc.currentPos; |
| 389 | |
| 390 | if (MatchLineNumberStart(sc)) { |
| 391 | state = SCE_FSHARP_LINENUM; |
| 392 | } else if (MatchPPDirectiveStart(sc)) { |
| 393 | state = SCE_FSHARP_PREPROCESSOR; |
| 394 | } else if (MatchLineComment(sc)) { |
| 395 | state = SCE_FSHARP_COMMENTLINE; |
| 396 | sc.Forward(); |
| 397 | sc.ch = SPACE; |
| 398 | } else if (MatchStreamCommentStart(sc)) { |
| 399 | state = SCE_FSHARP_COMMENT; |
| 400 | sc.Forward(); |
| 401 | sc.ch = SPACE; |
| 402 | } else if (MatchTypeAttributeStart(sc)) { |
| 403 | state = SCE_FSHARP_ATTRIBUTE; |
| 404 | sc.Forward(); |
| 405 | } else if (MatchQuotedExpressionStart(sc)) { |
| 406 | state = SCE_FSHARP_QUOTATION; |
| 407 | sc.Forward(); |
| 408 | } else if (MatchCharacterStart(sc)) { |
| 409 | state = SCE_FSHARP_CHARACTER; |
| 410 | } else if (MatchStringStart(sc)) { |
| 411 | fsStr.startChar = sc.ch; |
| 412 | fsStr.startPos = ZERO_LENGTH; |
| 413 | if (CanEmbedQuotes(sc)) { |
| 414 | // double quotes after this position should be non-terminating |
| 415 | fsStr.startPos = static_cast<Sci_Position>(sc.currentPos - cursor); |
| 416 | } |
| 417 | if (sc.ch == '`') { |
| 418 | state = SCE_FSHARP_QUOT_IDENTIFIER; |
| 419 | } else if (sc.ch == '@') { |
| 420 | state = SCE_FSHARP_VERBATIM; |
| 421 | } else { |
| 422 | state = SCE_FSHARP_STRING; |
| 423 | } |
| 424 | } else if (IsADigit(sc.ch, currentBase) || |
| 425 | ((sc.ch == '+' || sc.ch == '-') && IsADigit(sc.chNext))) { |
| 426 | state = SCE_FSHARP_NUMBER; |
| 427 | if (sc.ch == '0') { |
| 428 | const int prefix = sc.chNext; |
| 429 | if (numericPrefixes.find(prefix) != numericPrefixes.end()) { |
| 430 | currentBase = numericPrefixes[prefix]; |
| 431 | } |
| 432 | } |
| 433 | } else if (setOperators.Contains(sc.ch) && |
| 434 | // don't use operator style in async keywords (e.g. `return!`) |
| 435 | !(sc.ch == '!' && iswordstart(sc.chPrev)) && |
| 436 | // don't use operator style in member access, array/string indexing |
| 437 | !(sc.ch == '.' && (sc.chPrev == '\"' || iswordstart(sc.chPrev)) && |
| 438 | (iswordstart(sc.chNext) || sc.chNext == '['))) { |
| 439 | state = SCE_FSHARP_OPERATOR; |
| 440 | } else if (iswordstart(sc.ch)) { |
| 441 | state = SCE_FSHARP_IDENTIFIER; |
| 442 | } else { |
| 443 | state = SCE_FSHARP_DEFAULT; |
| 444 | } |
| 445 | break; |
| 446 | case SCE_FSHARP_LINENUM: |
| 447 | case SCE_FSHARP_PREPROCESSOR: |
| 448 | case SCE_FSHARP_COMMENTLINE: |
| 449 | if (sc.MatchLineEnd()) { |
| 450 | state = SCE_FSHARP_DEFAULT; |
| 451 | advance = false; |
| 452 | } |
| 453 | break; |
| 454 | case SCE_FSHARP_COMMENT: |
| 455 | case SCE_FSHARP_ATTRIBUTE: |
| 456 | case SCE_FSHARP_QUOTATION: |
| 457 | if (MatchStreamCommentEnd(sc) || MatchTypeAttributeEnd(sc) || MatchQuotedExpressionEnd(sc)) { |
| 458 | state = SCE_FSHARP_DEFAULT; |
| 459 | colorSpan++; |
| 460 | } |
| 461 | break; |
| 462 | case SCE_FSHARP_CHARACTER: |
| 463 | if (sc.chPrev == '\\' && sc.GetRelative(-2) != '\\') { |
| 464 | uniCh = UnicodeChar(sc.ch); |
| 465 | } else if (sc.ch == '\'' && |
| 466 | ((sc.chPrev == ' ' && sc.GetRelative(-2) == '\'') || sc.chPrev != '\\' || |
| 467 | (sc.chPrev == '\\' && sc.GetRelative(-2) == '\\'))) { |
| 468 | // byte literal? |
| 469 | if (sc.Match('\'', 'B')) { |
| 470 | sc.Forward(); |
| 471 | colorSpan++; |
| 472 | } |
| 473 | if (!sc.atLineEnd) { |
| 474 | colorSpan++; |
| 475 | } else { |
| 476 | sc.ChangeState(SCE_FSHARP_IDENTIFIER); |
| 477 | } |
| 478 | state = SCE_FSHARP_DEFAULT; |
| 479 | } else { |
| 480 | uniCh.Parse(sc.ch); |
| 481 | if (uniCh.AtEnd() && (sc.currentPos - cursor) >= 2) { |
| 482 | // terminate now, since we left the char behind |
| 483 | sc.ChangeState(SCE_FSHARP_IDENTIFIER); |
| 484 | advance = false; |
| 485 | } |
| 486 | } |
| 487 | break; |
| 488 | case SCE_FSHARP_STRING: |
| 489 | case SCE_FSHARP_VERBATIM: |
| 490 | case SCE_FSHARP_QUOT_IDENTIFIER: |
| 491 | if (MatchStringEnd(sc, fsStr)) { |
| 492 | const Sci_Position strLen = static_cast<Sci_Position>(sc.currentPos - cursor); |
| 493 | // backtrack to start of string |
| 494 | for (Sci_Position i = -strLen; i < 0; i++) { |
| 495 | const int startQuote = sc.GetRelative(i); |
| 496 | if (startQuote == '\"' || (startQuote == '`' && sc.GetRelative(i - 1) == '`')) { |
| 497 | // byte array? |
| 498 | if (sc.Match('\"', 'B')) { |
| 499 | sc.Forward(); |
| 500 | colorSpan++; |
| 501 | } |
| 502 | if (!sc.atLineEnd) { |
| 503 | colorSpan++; |
| 504 | } else { |
| 505 | sc.ChangeState(SCE_FSHARP_IDENTIFIER); |
| 506 | } |
| 507 | state = SCE_FSHARP_DEFAULT; |
| 508 | break; |
| 509 | } |
| 510 | } |
| 511 | } else if (sc.ch == '%' && |
| 512 | !(fsStr.startChar == '`' || sc.MatchIgnoreCase("% " ) || sc.MatchIgnoreCase("% \"" )) && |
| 513 | (setFormatSpecs.Contains(sc.chNext) || setFormatFlags.Contains(sc.chNext))) { |
| 514 | if (fsStr.CanInterpolate() && sc.chNext != '%') { |
| 515 | for (Sci_Position i = 2; i < length && !IsLineEnd(sc, i); i++) { |
| 516 | if (sc.GetRelative(i) == '{') { |
| 517 | state = setFormatSpecs.Contains(sc.GetRelative(i - 1)) |
| 518 | ? SCE_FSHARP_FORMAT_SPEC |
| 519 | : state; |
| 520 | break; |
| 521 | } |
| 522 | } |
| 523 | } else { |
| 524 | state = SCE_FSHARP_FORMAT_SPEC; |
| 525 | } |
| 526 | } |
| 527 | break; |
| 528 | case SCE_FSHARP_IDENTIFIER: |
| 529 | if (!(iswordstart(sc.ch) || sc.ch == '\'')) { |
| 530 | const Sci_Position wordLen = static_cast<Sci_Position>(sc.currentPos - cursor); |
| 531 | if (wordLen < MAX_WORD_LEN) { |
| 532 | // wordLength is believable as keyword, [re-]construct token - RR |
| 533 | char token[MAX_WORD_LEN] = { 0 }; |
| 534 | for (Sci_Position i = -wordLen; i < 0; i++) { |
| 535 | token[wordLen + i] = static_cast<char>(sc.GetRelative(i)); |
| 536 | } |
| 537 | token[wordLen] = '\0'; |
| 538 | // a snake_case_identifier can never be a keyword |
| 539 | if (!(sc.ch == '_' || sc.GetRelative(-wordLen - 1) == '_')) { |
| 540 | for (int i = 0; i < WORDLIST_SIZE; i++) { |
| 541 | if (keywords[i].InList(token)) { |
| 542 | sc.ChangeState(keywordClasses[i]); |
| 543 | break; |
| 544 | } |
| 545 | } |
| 546 | } |
| 547 | } |
| 548 | state = SCE_FSHARP_DEFAULT; |
| 549 | advance = false; |
| 550 | } |
| 551 | break; |
| 552 | case SCE_FSHARP_OPERATOR: |
| 553 | // special-case "()" and "[]" tokens as KEYWORDS - RR |
| 554 | if (setClosingTokens.Contains(sc.ch) && |
| 555 | ((sc.ch == ')' && sc.chPrev == '(') || (sc.ch == ']' && sc.chPrev == '['))) { |
| 556 | sc.ChangeState(SCE_FSHARP_KEYWORD); |
| 557 | colorSpan++; |
| 558 | } else { |
| 559 | advance = false; |
| 560 | } |
| 561 | state = SCE_FSHARP_DEFAULT; |
| 562 | break; |
| 563 | case SCE_FSHARP_NUMBER: |
| 564 | state = (IsNumber(sc, currentBase) || IsFloat(sc)) |
| 565 | ? SCE_FSHARP_NUMBER |
| 566 | // change style even when operators aren't spaced |
| 567 | : setOperators.Contains(sc.ch) ? SCE_FSHARP_OPERATOR : SCE_FSHARP_DEFAULT; |
| 568 | currentBase = (state == SCE_FSHARP_NUMBER) ? currentBase : 10; |
| 569 | break; |
| 570 | case SCE_FSHARP_FORMAT_SPEC: |
| 571 | if (!setFormatSpecs.Contains(sc.chNext) || |
| 572 | !(setFormatFlags.Contains(sc.ch) || IsADigit(sc.ch)) || |
| 573 | (setFormatFlags.Contains(sc.ch) && sc.ch == sc.chNext)) { |
| 574 | colorSpan++; |
| 575 | state = (fsStr.startChar == '@') ? SCE_FSHARP_VERBATIM : SCE_FSHARP_STRING; |
| 576 | } |
| 577 | break; |
| 578 | } |
| 579 | |
| 580 | if (state >= SCE_FSHARP_DEFAULT) { |
| 581 | styler.ColourTo(colorSpan, sc.state); |
| 582 | sc.ChangeState(state); |
| 583 | } |
| 584 | |
| 585 | if (advance) { |
| 586 | sc.Forward(); |
| 587 | } |
| 588 | } |
| 589 | |
| 590 | sc.Complete(); |
| 591 | } |
| 592 | |
| 593 | bool LineContains(LexAccessor &styler, const char *word, const Sci_Position start, const Sci_Position end = 1, |
| 594 | const int chAttr = SCE_FSHARP_DEFAULT); |
| 595 | |
| 596 | void FoldLexicalGroup(LexAccessor &styler, int &levelNext, const Sci_Position lineCurrent, const char *word, |
| 597 | const int chAttr); |
| 598 | |
| 599 | void SCI_METHOD LexerFSharp::Fold(Sci_PositionU start, Sci_Position length, int initStyle, IDocument *pAccess) { |
| 600 | if (!options.fold) { |
| 601 | return; |
| 602 | } |
| 603 | |
| 604 | LexAccessor styler(pAccess); |
| 605 | const Sci_Position startPos = static_cast<Sci_Position>(start); |
| 606 | const Sci_PositionU endPos = start + length; |
| 607 | Sci_Position lineCurrent = styler.GetLine(startPos); |
| 608 | Sci_Position lineNext = lineCurrent + 1; |
| 609 | Sci_Position lineStartNext = styler.LineStart(lineNext); |
| 610 | int style = initStyle; |
| 611 | int styleNext = styler.StyleAt(startPos); |
| 612 | char chNext = styler[startPos]; |
| 613 | int levelNext; |
| 614 | int levelCurrent = SC_FOLDLEVELBASE; |
| 615 | int visibleChars = 0; |
| 616 | |
| 617 | if (lineCurrent > 0) { |
| 618 | levelCurrent = styler.LevelAt(lineCurrent - 1) >> 0x10; |
| 619 | } |
| 620 | |
| 621 | levelNext = levelCurrent; |
| 622 | |
| 623 | for (Sci_PositionU i = start; i < endPos; i++) { |
| 624 | const Sci_Position currentPos = static_cast<Sci_Position>(i); |
| 625 | const bool atEOL = currentPos == (lineStartNext - 1); |
| 626 | const bool atLineOrDocEnd = (atEOL || (i == (endPos - 1))); |
| 627 | const bool atDosOrMacEOL = (atEOL || styler.SafeGetCharAt(currentPos) == '\r'); |
| 628 | const int stylePrev = style; |
| 629 | const char ch = chNext; |
| 630 | const bool = (stylePrev == SCE_FSHARP_COMMENTLINE); |
| 631 | style = styleNext; |
| 632 | styleNext = styler.StyleAt(currentPos + 1); |
| 633 | chNext = styler.SafeGetCharAt(currentPos + 1); |
| 634 | |
| 635 | if (options.foldComment) { |
| 636 | if (options.foldCommentMultiLine && inLineComment && atDosOrMacEOL && |
| 637 | (lineCurrent > 0 || styler.StyleAt(lineStartNext) == SCE_FSHARP_COMMENTLINE)) { |
| 638 | FoldLexicalGroup(styler, levelNext, lineCurrent, "//" , SCE_FSHARP_COMMENTLINE); |
| 639 | } |
| 640 | |
| 641 | if (options.foldCommentStream && style == SCE_FSHARP_COMMENT && !inLineComment) { |
| 642 | if (stylePrev != SCE_FSHARP_COMMENT) { |
| 643 | levelNext++; |
| 644 | } else if (styleNext != SCE_FSHARP_COMMENT && !atEOL) { |
| 645 | levelNext--; |
| 646 | } |
| 647 | } |
| 648 | } |
| 649 | |
| 650 | if (options.foldPreprocessor && style == SCE_FSHARP_PREPROCESSOR) { |
| 651 | if (styler.Match(currentPos, "#if" )) { |
| 652 | levelNext++; |
| 653 | } else if (styler.Match(currentPos, "#endif" )) { |
| 654 | levelNext--; |
| 655 | } |
| 656 | } |
| 657 | |
| 658 | if (options.foldImports && atLineOrDocEnd && |
| 659 | LineContains(styler, "open " , lineCurrent, 1, SCE_FSHARP_KEYWORD)) { |
| 660 | FoldLexicalGroup(styler, levelNext, lineCurrent, "open " , SCE_FSHARP_KEYWORD); |
| 661 | } |
| 662 | |
| 663 | if (!IsASpace(ch)) { |
| 664 | visibleChars++; |
| 665 | } |
| 666 | |
| 667 | if (atLineOrDocEnd) { |
| 668 | int levelUse = levelCurrent; |
| 669 | int lev = levelUse | levelNext << 16; |
| 670 | |
| 671 | if (visibleChars == 0 && options.foldCompact) { |
| 672 | lev |= SC_FOLDLEVELWHITEFLAG; |
| 673 | } |
| 674 | if (levelUse < levelNext) { |
| 675 | lev |= SC_FOLDLEVELHEADERFLAG; |
| 676 | } |
| 677 | if (lev != styler.LevelAt(lineCurrent)) { |
| 678 | styler.SetLevel(lineCurrent, lev); |
| 679 | } |
| 680 | |
| 681 | visibleChars = 0; |
| 682 | lineCurrent++; |
| 683 | lineNext = lineCurrent + 1; |
| 684 | lineStartNext = styler.LineStart(lineNext); |
| 685 | levelCurrent = levelNext; |
| 686 | |
| 687 | if (atEOL && (currentPos == (styler.Length() - 1))) { |
| 688 | styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG); |
| 689 | } |
| 690 | } |
| 691 | } |
| 692 | } |
| 693 | |
| 694 | bool LineContains(LexAccessor &styler, const char *word, const Sci_Position start, const Sci_Position end, |
| 695 | const int chAttr) { |
| 696 | bool found = false; |
| 697 | bool requireStyle = (chAttr > SCE_FSHARP_DEFAULT); |
| 698 | const Sci_Position limit = (end > 1) ? end : start; |
| 699 | for (Sci_Position i = start; i < styler.LineEnd(limit); i++) { |
| 700 | if (styler.Match(i, word)) { |
| 701 | found = requireStyle ? styler.StyleAt(i) == chAttr : true; |
| 702 | break; |
| 703 | } |
| 704 | } |
| 705 | return found; |
| 706 | } |
| 707 | |
| 708 | void FoldLexicalGroup(LexAccessor &styler, int &levelNext, const Sci_Position lineCurrent, const char *word, |
| 709 | const int chAttr) { |
| 710 | const Sci_Position linePrev = lineCurrent - 1; |
| 711 | const Sci_Position lineNext = lineCurrent + 1; |
| 712 | const Sci_Position lineStartPrev = styler.LineStart(linePrev); |
| 713 | const Sci_Position lineStartNext = styler.LineStart(lineNext); |
| 714 | const bool atFoldGroupStart = |
| 715 | (lineCurrent == 0 || !LineContains(styler, word, lineStartPrev, linePrev, chAttr)) && |
| 716 | LineContains(styler, word, lineStartNext, lineNext, chAttr); |
| 717 | const bool atFoldGroupEnd = |
| 718 | LineContains(styler, word, lineStartPrev, linePrev, chAttr) && |
| 719 | !LineContains(styler, word, lineStartNext, lineNext, chAttr); |
| 720 | |
| 721 | if (atFoldGroupStart) { |
| 722 | levelNext++; |
| 723 | } else if (atFoldGroupEnd && levelNext > SC_FOLDLEVELBASE) { |
| 724 | levelNext--; |
| 725 | } |
| 726 | } |
| 727 | } // namespace |
| 728 | |
| 729 | LexerModule lmFSharp(SCLEX_FSHARP, LexerFSharp::LexerFactoryFSharp, "fsharp" , fsharpWordLists); |
| 730 | |