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_TOKEN_H 15 #define LLVM_CLANG_TOKEN_H 16 17 #include "clang/Basic/TemplateKinds.h" 18 #include "clang/Basic/TokenKinds.h" 19 #include "clang/Basic/SourceLocation.h" 20 #include "clang/Basic/OperatorKinds.h" 21 #include <cstdlib> 22 23 namespace clang { 24 25 class IdentifierInfo; 26 27 /// Token - This structure provides full information about a lexed token. 28 /// It is not intended to be space efficient, it is intended to return as much 29 /// information as possible about each returned token. This is expected to be 30 /// compressed into a smaller form if memory footprint is important. 31 /// 32 /// The parser can create a special "annotation token" representing a stream of 33 /// tokens that were parsed and semantically resolved, e.g.: "foo::MyClass<int>" 34 /// can be represented by a single typename annotation token that carries 35 /// information about the SourceRange of the tokens and the type object. 36 class Token { 37 /// The location of the token. 38 SourceLocation Loc; 39 40 // Conceptually these next two fields could be in a union. However, this 41 // causes gcc 4.2 to pessimize LexTokenInternal, a very performance critical 42 // routine. Keeping as separate members with casts until a more beautiful fix 43 // presents itself. 44 45 /// UintData - This holds either the length of the token text, when 46 /// a normal token, or the end of the SourceRange when an annotation 47 /// token. 48 unsigned UintData; 49 50 /// PtrData - This is a union of four different pointer types, which depends 51 /// on what type of token this is: 52 /// Identifiers, keywords, etc: 53 /// This is an IdentifierInfo*, which contains the uniqued identifier 54 /// spelling. 55 /// Literals: isLiteral() returns true. 56 /// This is a pointer to the start of the token in a text buffer, which 57 /// may be dirty (have trigraphs / escaped newlines). 58 /// Annotations (resolved type names, C++ scopes, etc): isAnnotation(). 59 /// This is a pointer to sema-specific data for the annotation token. 60 /// Other: 61 /// This is null. 62 void *PtrData; 63 64 /// Kind - The actual flavor of token this is. 65 /// 66 unsigned short Kind; 67 68 /// Flags - Bits we track about this token, members of the TokenFlags enum. 69 unsigned char Flags; 70 public: 71 72 // Various flags set per token: 73 enum TokenFlags { 74 StartOfLine = 0x01, // At start of line or only after whitespace. 75 LeadingSpace = 0x02, // Whitespace exists before this token. 76 DisableExpand = 0x04, // This identifier may never be macro expanded. 77 NeedsCleaning = 0x08, // Contained an escaped newline or trigraph. 78 LeadingEmptyMacro = 0x10, // Empty macro exists before this token. 79 HasUDSuffix = 0x20 // This string or character literal has a ud-suffix. 80 }; 81 82 tok::TokenKind getKind() const { return (tok::TokenKind)Kind; } 83 void setKind(tok::TokenKind K) { Kind = K; } 84 85 /// is/isNot - Predicates to check if this token is a specific kind, as in 86 /// "if (Tok.is(tok::l_brace)) {...}". 87 bool is(tok::TokenKind K) const { return Kind == (unsigned) K; } 88 bool isNot(tok::TokenKind K) const { return Kind != (unsigned) K; } 89 90 /// isAnyIdentifier - Return true if this is a raw identifier (when lexing 91 /// in raw mode) or a non-keyword identifier (when lexing in non-raw mode). 92 bool isAnyIdentifier() const { 93 return is(tok::identifier) || is(tok::raw_identifier); 94 } 95 96 /// isLiteral - Return true if this is a "literal", like a numeric 97 /// constant, string, etc. 98 bool isLiteral() const { 99 return is(tok::numeric_constant) || is(tok::char_constant) || 100 is(tok::wide_char_constant) || is(tok::utf16_char_constant) || 101 is(tok::utf32_char_constant) || is(tok::string_literal) || 102 is(tok::wide_string_literal) || is(tok::utf8_string_literal) || 103 is(tok::utf16_string_literal) || is(tok::utf32_string_literal) || 104 is(tok::angle_string_literal); 105 } 106 107 bool isAnnotation() const { 108 #define ANNOTATION(NAME) \ 109 if (is(tok::annot_##NAME)) \ 110 return true; 111 #include "clang/Basic/TokenKinds.def" 112 return false; 113 } 114 115 /// getLocation - Return a source location identifier for the specified 116 /// offset in the current file. 117 SourceLocation getLocation() const { return Loc; } 118 unsigned getLength() const { 119 assert(!isAnnotation() && "Annotation tokens have no length field"); 120 return UintData; 121 } 122 123 void setLocation(SourceLocation L) { Loc = L; } 124 void setLength(unsigned Len) { 125 assert(!isAnnotation() && "Annotation tokens have no length field"); 126 UintData = Len; 127 } 128 129 SourceLocation getAnnotationEndLoc() const { 130 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token"); 131 return SourceLocation::getFromRawEncoding(UintData); 132 } 133 void setAnnotationEndLoc(SourceLocation L) { 134 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token"); 135 UintData = L.getRawEncoding(); 136 } 137 138 SourceLocation getLastLoc() const { 139 return isAnnotation() ? getAnnotationEndLoc() : getLocation(); 140 } 141 142 /// getAnnotationRange - SourceRange of the group of tokens that this 143 /// annotation token represents. 144 SourceRange getAnnotationRange() const { 145 return SourceRange(getLocation(), getAnnotationEndLoc()); 146 } 147 void setAnnotationRange(SourceRange R) { 148 setLocation(R.getBegin()); 149 setAnnotationEndLoc(R.getEnd()); 150 } 151 152 const char *getName() const { 153 return tok::getTokenName( (tok::TokenKind) Kind); 154 } 155 156 /// startToken - Reset all flags to cleared. 157 /// 158 void startToken() { 159 Kind = tok::unknown; 160 Flags = 0; 161 PtrData = 0; 162 UintData = 0; 163 Loc = SourceLocation(); 164 } 165 166 IdentifierInfo *getIdentifierInfo() const { 167 assert(isNot(tok::raw_identifier) && 168 "getIdentifierInfo() on a tok::raw_identifier token!"); 169 assert(!isAnnotation() && 170 "getIdentifierInfo() on an annotation token!"); 171 if (isLiteral()) return 0; 172 return (IdentifierInfo*) PtrData; 173 } 174 void setIdentifierInfo(IdentifierInfo *II) { 175 PtrData = (void*) II; 176 } 177 178 /// getRawIdentifierData - For a raw identifier token (i.e., an identifier 179 /// lexed in raw mode), returns a pointer to the start of it in the text 180 /// buffer if known, null otherwise. 181 const char *getRawIdentifierData() const { 182 assert(is(tok::raw_identifier)); 183 return reinterpret_cast<const char*>(PtrData); 184 } 185 void setRawIdentifierData(const char *Ptr) { 186 assert(is(tok::raw_identifier)); 187 PtrData = const_cast<char*>(Ptr); 188 } 189 190 /// getLiteralData - For a literal token (numeric constant, string, etc), this 191 /// returns a pointer to the start of it in the text buffer if known, null 192 /// otherwise. 193 const char *getLiteralData() const { 194 assert(isLiteral() && "Cannot get literal data of non-literal"); 195 return reinterpret_cast<const char*>(PtrData); 196 } 197 void setLiteralData(const char *Ptr) { 198 assert(isLiteral() && "Cannot set literal data of non-literal"); 199 PtrData = const_cast<char*>(Ptr); 200 } 201 202 void *getAnnotationValue() const { 203 assert(isAnnotation() && "Used AnnotVal on non-annotation token"); 204 return PtrData; 205 } 206 void setAnnotationValue(void *val) { 207 assert(isAnnotation() && "Used AnnotVal on non-annotation token"); 208 PtrData = val; 209 } 210 211 /// setFlag - Set the specified flag. 212 void setFlag(TokenFlags Flag) { 213 Flags |= Flag; 214 } 215 216 /// clearFlag - Unset the specified flag. 217 void clearFlag(TokenFlags Flag) { 218 Flags &= ~Flag; 219 } 220 221 /// getFlags - Return the internal represtation of the flags. 222 /// Only intended for low-level operations such as writing tokens to 223 // disk. 224 unsigned getFlags() const { 225 return Flags; 226 } 227 228 /// setFlagValue - Set a flag to either true or false. 229 void setFlagValue(TokenFlags Flag, bool Val) { 230 if (Val) 231 setFlag(Flag); 232 else 233 clearFlag(Flag); 234 } 235 236 /// isAtStartOfLine - Return true if this token is at the start of a line. 237 /// 238 bool isAtStartOfLine() const { return (Flags & StartOfLine) ? true : false; } 239 240 /// hasLeadingSpace - Return true if this token has whitespace before it. 241 /// 242 bool hasLeadingSpace() const { return (Flags & LeadingSpace) ? true : false; } 243 244 /// isExpandDisabled - Return true if this identifier token should never 245 /// be expanded in the future, due to C99 6.10.3.4p2. 246 bool isExpandDisabled() const { 247 return (Flags & DisableExpand) ? true : false; 248 } 249 250 /// isObjCAtKeyword - Return true if we have an ObjC keyword identifier. 251 bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const; 252 253 /// getObjCKeywordID - Return the ObjC keyword kind. 254 tok::ObjCKeywordKind getObjCKeywordID() const; 255 256 /// needsCleaning - Return true if this token has trigraphs or escaped 257 /// newlines in it. 258 /// 259 bool needsCleaning() const { return (Flags & NeedsCleaning) ? true : false; } 260 261 /// \brief Return true if this token has an empty macro before it. 262 /// 263 bool hasLeadingEmptyMacro() const { 264 return (Flags & LeadingEmptyMacro) ? true : false; 265 } 266 267 /// \brief Return true if this token is a string or character literal which 268 /// has a ud-suffix. 269 bool hasUDSuffix() const { return (Flags & HasUDSuffix) ? true : false; } 270 }; 271 272 /// PPConditionalInfo - Information about the conditional stack (#if directives) 273 /// currently active. 274 struct PPConditionalInfo { 275 /// IfLoc - Location where the conditional started. 276 /// 277 SourceLocation IfLoc; 278 279 /// WasSkipping - True if this was contained in a skipping directive, e.g. 280 /// in a "#if 0" block. 281 bool WasSkipping; 282 283 /// FoundNonSkip - True if we have emitted tokens already, and now we're in 284 /// an #else block or something. Only useful in Skipping blocks. 285 bool FoundNonSkip; 286 287 /// FoundElse - True if we've seen a #else in this block. If so, 288 /// #elif/#else directives are not allowed. 289 bool FoundElse; 290 }; 291 292 } // end namespace clang 293 294 namespace llvm { 295 template <> 296 struct isPodLike<clang::Token> { static const bool value = true; }; 297 } // end namespace llvm 298 299 #endif 300
