1 //===--- Token.h - Token interface ------------------------------*- C++ -*-===// 2 // 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 4 // See https://llvm.org/LICENSE.txt for license information. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 6 // 7 //===----------------------------------------------------------------------===// 8 // 9 // This file defines the Token interface. 10 // 11 //===----------------------------------------------------------------------===// 12 13 #ifndef LLVM_CLANG_LEX_TOKEN_H 14 #define LLVM_CLANG_LEX_TOKEN_H 15 16 #include "clang/Basic/SourceLocation.h" 17 #include "clang/Basic/TokenKinds.h" 18 #include "llvm/ADT/ArrayRef.h" 19 #include "llvm/ADT/StringRef.h" 20 #include <cassert> 21 22 namespace clang { 23 24 class IdentifierInfo; 25 26 /// Token - This structure provides full information about a lexed token. 27 /// It is not intended to be space efficient, it is intended to return as much 28 /// information as possible about each returned token. This is expected to be 29 /// compressed into a smaller form if memory footprint is important. 30 /// 31 /// The parser can create a special "annotation token" representing a stream of 32 /// tokens that were parsed and semantically resolved, e.g.: "foo::MyClass<int>" 33 /// can be represented by a single typename annotation token that carries 34 /// information about the SourceRange of the tokens and the type object. 35 class Token { 36 /// The location of the token. This is actually a SourceLocation. 37 SourceLocation::UIntTy Loc; 38 39 // Conceptually these next two fields could be in a union. However, this 40 // causes gcc 4.2 to pessimize LexTokenInternal, a very performance critical 41 // routine. Keeping as separate members with casts until a more beautiful fix 42 // presents itself. 43 44 /// UintData - This holds either the length of the token text, when 45 /// a normal token, or the end of the SourceRange when an annotation 46 /// token. 47 SourceLocation::UIntTy UintData; 48 49 /// PtrData - This is a union of four different pointer types, which depends 50 /// on what type of token this is: 51 /// Identifiers, keywords, etc: 52 /// This is an IdentifierInfo*, which contains the uniqued identifier 53 /// spelling. 54 /// Literals: isLiteral() returns true. 55 /// This is a pointer to the start of the token in a text buffer, which 56 /// may be dirty (have trigraphs / escaped newlines). 57 /// Annotations (resolved type names, C++ scopes, etc): isAnnotation(). 58 /// This is a pointer to sema-specific data for the annotation token. 59 /// Eof: 60 // This is a pointer to a Decl. 61 /// Other: 62 /// This is null. 63 void *PtrData; 64 65 /// Kind - The actual flavor of token this is. 66 tok::TokenKind Kind; 67 68 /// Flags - Bits we track about this token, members of the TokenFlags enum. 69 unsigned short Flags; 70 71 public: 72 // Various flags set per token: 73 enum TokenFlags { 74 StartOfLine = 0x01, // At start of line or only after whitespace 75 // (considering the line after macro expansion). 76 LeadingSpace = 0x02, // Whitespace exists before this token (considering 77 // whitespace after macro expansion). 78 DisableExpand = 0x04, // This identifier may never be macro expanded. 79 NeedsCleaning = 0x08, // Contained an escaped newline or trigraph. 80 LeadingEmptyMacro = 0x10, // Empty macro exists before this token. 81 HasUDSuffix = 0x20, // This string or character literal has a ud-suffix. 82 HasUCN = 0x40, // This identifier contains a UCN. 83 IgnoredComma = 0x80, // This comma is not a macro argument separator (MS). 84 StringifiedInMacro = 0x100, // This string or character literal is formed by 85 // macro stringizing or charizing operator. 86 CommaAfterElided = 0x200, // The comma following this token was elided (MS). 87 IsEditorPlaceholder = 0x400, // This identifier is a placeholder. 88 IsReinjected = 0x800, // A phase 4 token that was produced before and 89 // re-added, e.g. via EnterTokenStream. Annotation 90 // tokens are *not* reinjected. 91 }; 92 getKind()93 tok::TokenKind getKind() const { return Kind; } setKind(tok::TokenKind K)94 void setKind(tok::TokenKind K) { Kind = K; } 95 96 /// is/isNot - Predicates to check if this token is a specific kind, as in 97 /// "if (Tok.is(tok::l_brace)) {...}". is(tok::TokenKind K)98 bool is(tok::TokenKind K) const { return Kind == K; } isNot(tok::TokenKind K)99 bool isNot(tok::TokenKind K) const { return Kind != K; } isOneOf(tok::TokenKind K1,tok::TokenKind K2)100 bool isOneOf(tok::TokenKind K1, tok::TokenKind K2) const { 101 return is(K1) || is(K2); 102 } isOneOf(tok::TokenKind K1,Ts...Ks)103 template <typename... Ts> bool isOneOf(tok::TokenKind K1, Ts... Ks) const { 104 return is(K1) || isOneOf(Ks...); 105 } 106 107 /// Return true if this is a raw identifier (when lexing 108 /// in raw mode) or a non-keyword identifier (when lexing in non-raw mode). isAnyIdentifier()109 bool isAnyIdentifier() const { 110 return tok::isAnyIdentifier(getKind()); 111 } 112 113 /// Return true if this is a "literal", like a numeric 114 /// constant, string, etc. isLiteral()115 bool isLiteral() const { 116 return tok::isLiteral(getKind()); 117 } 118 119 /// Return true if this is any of tok::annot_* kind tokens. isAnnotation()120 bool isAnnotation() const { return tok::isAnnotation(getKind()); } 121 122 /// Return true if the token is a keyword that is parsed in the same 123 /// position as a standard attribute, but that has semantic meaning 124 /// and so cannot be a true attribute. isRegularKeywordAttribute()125 bool isRegularKeywordAttribute() const { 126 return tok::isRegularKeywordAttribute(getKind()); 127 } 128 129 /// Return a source location identifier for the specified 130 /// offset in the current file. getLocation()131 SourceLocation getLocation() const { 132 return SourceLocation::getFromRawEncoding(Loc); 133 } getLength()134 unsigned getLength() const { 135 assert(!isAnnotation() && "Annotation tokens have no length field"); 136 return UintData; 137 } 138 setLocation(SourceLocation L)139 void setLocation(SourceLocation L) { Loc = L.getRawEncoding(); } setLength(unsigned Len)140 void setLength(unsigned Len) { 141 assert(!isAnnotation() && "Annotation tokens have no length field"); 142 UintData = Len; 143 } 144 getAnnotationEndLoc()145 SourceLocation getAnnotationEndLoc() const { 146 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token"); 147 return SourceLocation::getFromRawEncoding(UintData ? UintData : Loc); 148 } setAnnotationEndLoc(SourceLocation L)149 void setAnnotationEndLoc(SourceLocation L) { 150 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token"); 151 UintData = L.getRawEncoding(); 152 } 153 getLastLoc()154 SourceLocation getLastLoc() const { 155 return isAnnotation() ? getAnnotationEndLoc() : getLocation(); 156 } 157 getEndLoc()158 SourceLocation getEndLoc() const { 159 return isAnnotation() ? getAnnotationEndLoc() 160 : getLocation().getLocWithOffset(getLength()); 161 } 162 163 /// SourceRange of the group of tokens that this annotation token 164 /// represents. getAnnotationRange()165 SourceRange getAnnotationRange() const { 166 return SourceRange(getLocation(), getAnnotationEndLoc()); 167 } setAnnotationRange(SourceRange R)168 void setAnnotationRange(SourceRange R) { 169 setLocation(R.getBegin()); 170 setAnnotationEndLoc(R.getEnd()); 171 } 172 getName()173 const char *getName() const { return tok::getTokenName(Kind); } 174 175 /// Reset all flags to cleared. startToken()176 void startToken() { 177 Kind = tok::unknown; 178 Flags = 0; 179 PtrData = nullptr; 180 UintData = 0; 181 Loc = SourceLocation().getRawEncoding(); 182 } 183 hasPtrData()184 bool hasPtrData() const { return PtrData != nullptr; } 185 getIdentifierInfo()186 IdentifierInfo *getIdentifierInfo() const { 187 assert(isNot(tok::raw_identifier) && 188 "getIdentifierInfo() on a tok::raw_identifier token!"); 189 assert(!isAnnotation() && 190 "getIdentifierInfo() on an annotation token!"); 191 if (isLiteral()) return nullptr; 192 if (is(tok::eof)) return nullptr; 193 return (IdentifierInfo*) PtrData; 194 } setIdentifierInfo(IdentifierInfo * II)195 void setIdentifierInfo(IdentifierInfo *II) { 196 PtrData = (void*) II; 197 } 198 getEofData()199 const void *getEofData() const { 200 assert(is(tok::eof)); 201 return reinterpret_cast<const void *>(PtrData); 202 } setEofData(const void * D)203 void setEofData(const void *D) { 204 assert(is(tok::eof)); 205 assert(!PtrData); 206 PtrData = const_cast<void *>(D); 207 } 208 209 /// getRawIdentifier - For a raw identifier token (i.e., an identifier 210 /// lexed in raw mode), returns a reference to the text substring in the 211 /// buffer if known. getRawIdentifier()212 StringRef getRawIdentifier() const { 213 assert(is(tok::raw_identifier)); 214 return StringRef(reinterpret_cast<const char *>(PtrData), getLength()); 215 } setRawIdentifierData(const char * Ptr)216 void setRawIdentifierData(const char *Ptr) { 217 assert(is(tok::raw_identifier)); 218 PtrData = const_cast<char*>(Ptr); 219 } 220 221 /// getLiteralData - For a literal token (numeric constant, string, etc), this 222 /// returns a pointer to the start of it in the text buffer if known, null 223 /// otherwise. getLiteralData()224 const char *getLiteralData() const { 225 assert(isLiteral() && "Cannot get literal data of non-literal"); 226 return reinterpret_cast<const char*>(PtrData); 227 } setLiteralData(const char * Ptr)228 void setLiteralData(const char *Ptr) { 229 assert(isLiteral() && "Cannot set literal data of non-literal"); 230 PtrData = const_cast<char*>(Ptr); 231 } 232 getAnnotationValue()233 void *getAnnotationValue() const { 234 assert(isAnnotation() && "Used AnnotVal on non-annotation token"); 235 return PtrData; 236 } setAnnotationValue(void * val)237 void setAnnotationValue(void *val) { 238 assert(isAnnotation() && "Used AnnotVal on non-annotation token"); 239 PtrData = val; 240 } 241 242 /// Set the specified flag. setFlag(TokenFlags Flag)243 void setFlag(TokenFlags Flag) { 244 Flags |= Flag; 245 } 246 247 /// Get the specified flag. getFlag(TokenFlags Flag)248 bool getFlag(TokenFlags Flag) const { 249 return (Flags & Flag) != 0; 250 } 251 252 /// Unset the specified flag. clearFlag(TokenFlags Flag)253 void clearFlag(TokenFlags Flag) { 254 Flags &= ~Flag; 255 } 256 257 /// Return the internal represtation of the flags. 258 /// 259 /// This is only intended for low-level operations such as writing tokens to 260 /// disk. getFlags()261 unsigned getFlags() const { 262 return Flags; 263 } 264 265 /// Set a flag to either true or false. setFlagValue(TokenFlags Flag,bool Val)266 void setFlagValue(TokenFlags Flag, bool Val) { 267 if (Val) 268 setFlag(Flag); 269 else 270 clearFlag(Flag); 271 } 272 273 /// isAtStartOfLine - Return true if this token is at the start of a line. 274 /// isAtStartOfLine()275 bool isAtStartOfLine() const { return getFlag(StartOfLine); } 276 277 /// Return true if this token has whitespace before it. 278 /// hasLeadingSpace()279 bool hasLeadingSpace() const { return getFlag(LeadingSpace); } 280 281 /// Return true if this identifier token should never 282 /// be expanded in the future, due to C99 6.10.3.4p2. isExpandDisabled()283 bool isExpandDisabled() const { return getFlag(DisableExpand); } 284 285 /// Return true if we have an ObjC keyword identifier. 286 bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const; 287 288 /// Return the ObjC keyword kind. 289 tok::ObjCKeywordKind getObjCKeywordID() const; 290 291 /// Return true if this token has trigraphs or escaped newlines in it. needsCleaning()292 bool needsCleaning() const { return getFlag(NeedsCleaning); } 293 294 /// Return true if this token has an empty macro before it. 295 /// hasLeadingEmptyMacro()296 bool hasLeadingEmptyMacro() const { return getFlag(LeadingEmptyMacro); } 297 298 /// Return true if this token is a string or character literal which 299 /// has a ud-suffix. hasUDSuffix()300 bool hasUDSuffix() const { return getFlag(HasUDSuffix); } 301 302 /// Returns true if this token contains a universal character name. hasUCN()303 bool hasUCN() const { return getFlag(HasUCN); } 304 305 /// Returns true if this token is formed by macro by stringizing or charizing 306 /// operator. stringifiedInMacro()307 bool stringifiedInMacro() const { return getFlag(StringifiedInMacro); } 308 309 /// Returns true if the comma after this token was elided. commaAfterElided()310 bool commaAfterElided() const { return getFlag(CommaAfterElided); } 311 312 /// Returns true if this token is an editor placeholder. 313 /// 314 /// Editor placeholders are produced by the code-completion engine and are 315 /// represented as characters between '<#' and '#>' in the source code. The 316 /// lexer uses identifier tokens to represent placeholders. isEditorPlaceholder()317 bool isEditorPlaceholder() const { return getFlag(IsEditorPlaceholder); } 318 }; 319 320 /// Information about the conditional stack (\#if directives) 321 /// currently active. 322 struct PPConditionalInfo { 323 /// Location where the conditional started. 324 SourceLocation IfLoc; 325 326 /// True if this was contained in a skipping directive, e.g., 327 /// in a "\#if 0" block. 328 bool WasSkipping; 329 330 /// True if we have emitted tokens already, and now we're in 331 /// an \#else block or something. Only useful in Skipping blocks. 332 bool FoundNonSkip; 333 334 /// True if we've seen a \#else in this block. If so, 335 /// \#elif/\#else directives are not allowed. 336 bool FoundElse; 337 }; 338 339 // Extra information needed for annonation tokens. 340 struct PragmaLoopHintInfo { 341 Token PragmaName; 342 Token Option; 343 ArrayRef<Token> Toks; 344 }; 345 } // end namespace clang 346 347 #endif // LLVM_CLANG_LEX_TOKEN_H 348