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_LEX_TOKEN_H
15 #define LLVM_CLANG_LEX_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"
21 #include "llvm/ADT/StringRef.h"
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.
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.
38 /// The location of the token.
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
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
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
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.
62 // This is a pointer to a Decl.
67 /// Kind - The actual flavor of token this is.
70 /// Flags - Bits we track about this token, members of the TokenFlags enum.
74 // Various flags set per token:
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.
90 tok::TokenKind getKind() const { return Kind; }
91 void setKind(tok::TokenKind K) { Kind = K; }
93 /// is/isNot - Predicates to check if this token is a specific kind, as in
94 /// "if (Tok.is(tok::l_brace)) {...}".
95 bool is(tok::TokenKind K) const { return Kind == K; }
96 bool isNot(tok::TokenKind K) const { return Kind != K; }
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).
100 bool isAnyIdentifier() const {
101 return tok::isAnyIdentifier(getKind());
104 /// \brief Return true if this is a "literal", like a numeric
105 /// constant, string, etc.
106 bool isLiteral() const {
107 return tok::isLiteral(getKind());
110 /// \brief Return true if this is any of tok::annot_* kind tokens.
111 bool isAnnotation() const {
112 return tok::isAnnotation(getKind());
115 /// \brief 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");
123 void setLocation(SourceLocation L) { Loc = L; }
124 void setLength(unsigned Len) {
125 assert(!isAnnotation() && "Annotation tokens have no length field");
129 SourceLocation getAnnotationEndLoc() const {
130 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
131 return SourceLocation::getFromRawEncoding(UintData);
133 void setAnnotationEndLoc(SourceLocation L) {
134 assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
135 UintData = L.getRawEncoding();
138 SourceLocation getLastLoc() const {
139 return isAnnotation() ? getAnnotationEndLoc() : getLocation();
142 /// \brief SourceRange of the group of tokens that this annotation token
144 SourceRange getAnnotationRange() const {
145 return SourceRange(getLocation(), getAnnotationEndLoc());
147 void setAnnotationRange(SourceRange R) {
148 setLocation(R.getBegin());
149 setAnnotationEndLoc(R.getEnd());
152 const char *getName() const { return tok::getTokenName(Kind); }
154 /// \brief Reset all flags to cleared.
160 Loc = SourceLocation();
163 IdentifierInfo *getIdentifierInfo() const {
164 assert(isNot(tok::raw_identifier) &&
165 "getIdentifierInfo() on a tok::raw_identifier token!");
166 assert(!isAnnotation() &&
167 "getIdentifierInfo() on an annotation token!");
168 if (isLiteral()) return nullptr;
169 if (is(tok::eof)) return nullptr;
170 return (IdentifierInfo*) PtrData;
172 void setIdentifierInfo(IdentifierInfo *II) {
173 PtrData = (void*) II;
176 const void *getEofData() const {
177 assert(is(tok::eof));
178 return reinterpret_cast<const void *>(PtrData);
180 void setEofData(const void *D) {
181 assert(is(tok::eof));
183 PtrData = const_cast<void *>(D);
186 /// getRawIdentifier - For a raw identifier token (i.e., an identifier
187 /// lexed in raw mode), returns a reference to the text substring in the
189 StringRef getRawIdentifier() const {
190 assert(is(tok::raw_identifier));
191 return StringRef(reinterpret_cast<const char *>(PtrData), getLength());
193 void setRawIdentifierData(const char *Ptr) {
194 assert(is(tok::raw_identifier));
195 PtrData = const_cast<char*>(Ptr);
198 /// getLiteralData - For a literal token (numeric constant, string, etc), this
199 /// returns a pointer to the start of it in the text buffer if known, null
201 const char *getLiteralData() const {
202 assert(isLiteral() && "Cannot get literal data of non-literal");
203 return reinterpret_cast<const char*>(PtrData);
205 void setLiteralData(const char *Ptr) {
206 assert(isLiteral() && "Cannot set literal data of non-literal");
207 PtrData = const_cast<char*>(Ptr);
210 void *getAnnotationValue() const {
211 assert(isAnnotation() && "Used AnnotVal on non-annotation token");
214 void setAnnotationValue(void *val) {
215 assert(isAnnotation() && "Used AnnotVal on non-annotation token");
219 /// \brief Set the specified flag.
220 void setFlag(TokenFlags Flag) {
224 /// \brief Unset the specified flag.
225 void clearFlag(TokenFlags Flag) {
229 /// \brief Return the internal represtation of the flags.
231 /// This is only intended for low-level operations such as writing tokens to
233 unsigned getFlags() const {
237 /// \brief Set a flag to either true or false.
238 void setFlagValue(TokenFlags Flag, bool Val) {
245 /// isAtStartOfLine - Return true if this token is at the start of a line.
247 bool isAtStartOfLine() const { return (Flags & StartOfLine) ? true : false; }
249 /// \brief Return true if this token has whitespace before it.
251 bool hasLeadingSpace() const { return (Flags & LeadingSpace) ? true : false; }
253 /// \brief Return true if this identifier token should never
254 /// be expanded in the future, due to C99 6.10.3.4p2.
255 bool isExpandDisabled() const {
256 return (Flags & DisableExpand) ? true : false;
259 /// \brief Return true if we have an ObjC keyword identifier.
260 bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const;
262 /// \brief Return the ObjC keyword kind.
263 tok::ObjCKeywordKind getObjCKeywordID() const;
265 /// \brief Return true if this token has trigraphs or escaped newlines in it.
266 bool needsCleaning() const { return (Flags & NeedsCleaning) ? true : false; }
268 /// \brief Return true if this token has an empty macro before it.
270 bool hasLeadingEmptyMacro() const {
271 return (Flags & LeadingEmptyMacro) ? true : false;
274 /// \brief Return true if this token is a string or character literal which
276 bool hasUDSuffix() const { return (Flags & HasUDSuffix) ? true : false; }
278 /// Returns true if this token contains a universal character name.
279 bool hasUCN() const { return (Flags & HasUCN) ? true : false; }
281 /// Returns true if this token is formed by macro by stringizing or charizing
283 bool stringifiedInMacro() const {
284 return (Flags & StringifiedInMacro) ? true : false;
288 /// \brief Information about the conditional stack (\#if directives)
289 /// currently active.
290 struct PPConditionalInfo {
291 /// \brief Location where the conditional started.
292 SourceLocation IfLoc;
294 /// \brief True if this was contained in a skipping directive, e.g.,
295 /// in a "\#if 0" block.
298 /// \brief True if we have emitted tokens already, and now we're in
299 /// an \#else block or something. Only useful in Skipping blocks.
302 /// \brief True if we've seen a \#else in this block. If so,
303 /// \#elif/\#else directives are not allowed.
307 } // end namespace clang
311 struct isPodLike<clang::Token> { static const bool value = true; };
312 } // end namespace llvm