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