1 //===--- Token.h - Token interface ------------------------------*- C++ -*-===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 // This file defines the Token interface.
12 //===----------------------------------------------------------------------===//
14 #ifndef LLVM_CLANG_TOKEN_H
15 #define LLVM_CLANG_TOKEN_H
17 #include "clang/Basic/OperatorKinds.h"
18 #include "clang/Basic/SourceLocation.h"
19 #include "clang/Basic/TemplateKinds.h"
20 #include "clang/Basic/TokenKinds.h"
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.
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.
37 /// The location of the token.
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
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
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
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.
64 /// Kind - The actual flavor of token this is.
68 /// Flags - Bits we track about this token, members of the TokenFlags enum.
72 // Various flags set per token:
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).
86 tok::TokenKind getKind() const { return (tok::TokenKind)Kind; }
87 void setKind(tok::TokenKind K) { Kind = K; }
89 /// is/isNot - Predicates to check if this token is a specific kind, as in
90 /// "if (Tok.is(tok::l_brace)) {...}".
91 bool is(tok::TokenKind K) const { return Kind == (unsigned) K; }
92 bool isNot(tok::TokenKind K) const { return Kind != (unsigned) K; }
94 /// \brief Return true if this is a raw identifier (when lexing
95 /// in raw mode) or a non-keyword identifier (when lexing in non-raw mode).
96 bool isAnyIdentifier() const {
97 return tok::isAnyIdentifier(getKind());
100 /// \brief Return true if this is a "literal", like a numeric
101 /// constant, string, etc.
102 bool isLiteral() const {
103 return tok::isLiteral(getKind());
106 /// \brief Return true if this is any of tok::annot_* kind tokens.
107 bool isAnnotation() const {
108 return tok::isAnnotation(getKind());
111 /// \brief Return a source location identifier for the specified
112 /// offset in the current file.
113 SourceLocation getLocation() const { return Loc; }
114 unsigned getLength() const {
115 assert(!isAnnotation() && "Annotation tokens have no length field");
119 void setLocation(SourceLocation L) { Loc = L; }
120 void setLength(unsigned Len) {
121 assert(!isAnnotation() && "Annotation tokens have no length field");
125 SourceLocation getAnnotationEndLoc() const {
126 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
127 return SourceLocation::getFromRawEncoding(UintData);
129 void setAnnotationEndLoc(SourceLocation L) {
130 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
131 UintData = L.getRawEncoding();
134 SourceLocation getLastLoc() const {
135 return isAnnotation() ? getAnnotationEndLoc() : getLocation();
138 /// \brief SourceRange of the group of tokens that this annotation token
140 SourceRange getAnnotationRange() const {
141 return SourceRange(getLocation(), getAnnotationEndLoc());
143 void setAnnotationRange(SourceRange R) {
144 setLocation(R.getBegin());
145 setAnnotationEndLoc(R.getEnd());
148 const char *getName() const {
149 return tok::getTokenName( (tok::TokenKind) Kind);
152 /// \brief Reset all flags to cleared.
158 Loc = SourceLocation();
161 IdentifierInfo *getIdentifierInfo() const {
162 assert(isNot(tok::raw_identifier) &&
163 "getIdentifierInfo() on a tok::raw_identifier token!");
164 assert(!isAnnotation() &&
165 "getIdentifierInfo() on an annotation token!");
166 if (isLiteral()) return 0;
167 return (IdentifierInfo*) PtrData;
169 void setIdentifierInfo(IdentifierInfo *II) {
170 PtrData = (void*) II;
173 /// getRawIdentifierData - For a raw identifier token (i.e., an identifier
174 /// lexed in raw mode), returns a pointer to the start of it in the text
175 /// buffer if known, null otherwise.
176 const char *getRawIdentifierData() const {
177 assert(is(tok::raw_identifier));
178 return reinterpret_cast<const char*>(PtrData);
180 void setRawIdentifierData(const char *Ptr) {
181 assert(is(tok::raw_identifier));
182 PtrData = const_cast<char*>(Ptr);
185 /// getLiteralData - For a literal token (numeric constant, string, etc), this
186 /// returns a pointer to the start of it in the text buffer if known, null
188 const char *getLiteralData() const {
189 assert(isLiteral() && "Cannot get literal data of non-literal");
190 return reinterpret_cast<const char*>(PtrData);
192 void setLiteralData(const char *Ptr) {
193 assert(isLiteral() && "Cannot set literal data of non-literal");
194 PtrData = const_cast<char*>(Ptr);
197 void *getAnnotationValue() const {
198 assert(isAnnotation() && "Used AnnotVal on non-annotation token");
201 void setAnnotationValue(void *val) {
202 assert(isAnnotation() && "Used AnnotVal on non-annotation token");
206 /// \brief Set the specified flag.
207 void setFlag(TokenFlags Flag) {
211 /// \brief Unset the specified flag.
212 void clearFlag(TokenFlags Flag) {
216 /// \brief Return the internal represtation of the flags.
218 /// This is only intended for low-level operations such as writing tokens to
220 unsigned getFlags() const {
224 /// \brief Set a flag to either true or false.
225 void setFlagValue(TokenFlags Flag, bool Val) {
232 /// isAtStartOfLine - Return true if this token is at the start of a line.
234 bool isAtStartOfLine() const { return (Flags & StartOfLine) ? true : false; }
236 /// \brief Return true if this token has whitespace before it.
238 bool hasLeadingSpace() const { return (Flags & LeadingSpace) ? true : false; }
240 /// \brief Return true if this identifier token should never
241 /// be expanded in the future, due to C99 6.10.3.4p2.
242 bool isExpandDisabled() const {
243 return (Flags & DisableExpand) ? true : false;
246 /// \brief Return true if we have an ObjC keyword identifier.
247 bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const;
249 /// \brief Return the ObjC keyword kind.
250 tok::ObjCKeywordKind getObjCKeywordID() const;
252 /// \brief Return true if this token has trigraphs or escaped newlines in it.
253 bool needsCleaning() const { return (Flags & NeedsCleaning) ? true : false; }
255 /// \brief Return true if this token has an empty macro before it.
257 bool hasLeadingEmptyMacro() const {
258 return (Flags & LeadingEmptyMacro) ? true : false;
261 /// \brief Return true if this token is a string or character literal which
263 bool hasUDSuffix() const { return (Flags & HasUDSuffix) ? true : false; }
265 /// Returns true if this token contains a universal character name.
266 bool hasUCN() const { return (Flags & HasUCN) ? true : false; }
269 /// \brief Information about the conditional stack (\#if directives)
270 /// currently active.
271 struct PPConditionalInfo {
272 /// \brief Location where the conditional started.
273 SourceLocation IfLoc;
275 /// \brief True if this was contained in a skipping directive, e.g.,
276 /// in a "\#if 0" block.
279 /// \brief True if we have emitted tokens already, and now we're in
280 /// an \#else block or something. Only useful in Skipping blocks.
283 /// \brief True if we've seen a \#else in this block. If so,
284 /// \#elif/\#else directives are not allowed.
288 } // end namespace clang
292 struct isPodLike<clang::Token> { static const bool value = true; };
293 } // end namespace llvm