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