Home | History | Annotate | Download | only in Lex
      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   };
     80 
     81   tok::TokenKind getKind() const { return (tok::TokenKind)Kind; }
     82   void setKind(tok::TokenKind K) { Kind = K; }
     83 
     84   /// is/isNot - Predicates to check if this token is a specific kind, as in
     85   /// "if (Tok.is(tok::l_brace)) {...}".
     86   bool is(tok::TokenKind K) const { return Kind == (unsigned) K; }
     87   bool isNot(tok::TokenKind K) const { return Kind != (unsigned) K; }
     88 
     89   /// isAnyIdentifier - Return true if this is a raw identifier (when lexing
     90   /// in raw mode) or a non-keyword identifier (when lexing in non-raw mode).
     91   bool isAnyIdentifier() const {
     92     return is(tok::identifier) || is(tok::raw_identifier);
     93   }
     94 
     95   /// isLiteral - Return true if this is a "literal", like a numeric
     96   /// constant, string, etc.
     97   bool isLiteral() const {
     98     return is(tok::numeric_constant) || is(tok::char_constant) ||
     99            is(tok::wide_char_constant) || is(tok::utf16_char_constant) ||
    100            is(tok::utf32_char_constant) || is(tok::string_literal) ||
    101            is(tok::wide_string_literal) || is(tok::utf8_string_literal) ||
    102            is(tok::utf16_string_literal) || is(tok::utf32_string_literal) ||
    103            is(tok::angle_string_literal);
    104   }
    105 
    106   bool isAnnotation() const {
    107 #define ANNOTATION(NAME) \
    108     if (is(tok::annot_##NAME)) \
    109       return true;
    110 #include "clang/Basic/TokenKinds.def"
    111     return false;
    112   }
    113 
    114   /// getLocation - Return a source location identifier for the specified
    115   /// offset in the current file.
    116   SourceLocation getLocation() const { return Loc; }
    117   unsigned getLength() const {
    118     assert(!isAnnotation() && "Annotation tokens have no length field");
    119     return UintData;
    120   }
    121 
    122   void setLocation(SourceLocation L) { Loc = L; }
    123   void setLength(unsigned Len) {
    124     assert(!isAnnotation() && "Annotation tokens have no length field");
    125     UintData = Len;
    126   }
    127 
    128   SourceLocation getAnnotationEndLoc() const {
    129     assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
    130     return SourceLocation::getFromRawEncoding(UintData);
    131   }
    132   void setAnnotationEndLoc(SourceLocation L) {
    133     assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
    134     UintData = L.getRawEncoding();
    135   }
    136 
    137   SourceLocation getLastLoc() const {
    138     return isAnnotation() ? getAnnotationEndLoc() : getLocation();
    139   }
    140 
    141   /// getAnnotationRange - SourceRange of the group of tokens that this
    142   /// annotation token represents.
    143   SourceRange getAnnotationRange() const {
    144     return SourceRange(getLocation(), getAnnotationEndLoc());
    145   }
    146   void setAnnotationRange(SourceRange R) {
    147     setLocation(R.getBegin());
    148     setAnnotationEndLoc(R.getEnd());
    149   }
    150 
    151   const char *getName() const {
    152     return tok::getTokenName( (tok::TokenKind) Kind);
    153   }
    154 
    155   /// startToken - Reset all flags to cleared.
    156   ///
    157   void startToken() {
    158     Kind = tok::unknown;
    159     Flags = 0;
    160     PtrData = 0;
    161     UintData = 0;
    162     Loc = SourceLocation();
    163   }
    164 
    165   IdentifierInfo *getIdentifierInfo() const {
    166     assert(isNot(tok::raw_identifier) &&
    167            "getIdentifierInfo() on a tok::raw_identifier token!");
    168     assert(!isAnnotation() &&
    169            "getIdentifierInfo() on an annotation token!");
    170     if (isLiteral()) return 0;
    171     return (IdentifierInfo*) PtrData;
    172   }
    173   void setIdentifierInfo(IdentifierInfo *II) {
    174     PtrData = (void*) II;
    175   }
    176 
    177   /// getRawIdentifierData - For a raw identifier token (i.e., an identifier
    178   /// lexed in raw mode), returns a pointer to the start of it in the text
    179   /// buffer if known, null otherwise.
    180   const char *getRawIdentifierData() const {
    181     assert(is(tok::raw_identifier));
    182     return reinterpret_cast<const char*>(PtrData);
    183   }
    184   void setRawIdentifierData(const char *Ptr) {
    185     assert(is(tok::raw_identifier));
    186     PtrData = const_cast<char*>(Ptr);
    187   }
    188 
    189   /// getLiteralData - For a literal token (numeric constant, string, etc), this
    190   /// returns a pointer to the start of it in the text buffer if known, null
    191   /// otherwise.
    192   const char *getLiteralData() const {
    193     assert(isLiteral() && "Cannot get literal data of non-literal");
    194     return reinterpret_cast<const char*>(PtrData);
    195   }
    196   void setLiteralData(const char *Ptr) {
    197     assert(isLiteral() && "Cannot set literal data of non-literal");
    198     PtrData = const_cast<char*>(Ptr);
    199   }
    200 
    201   void *getAnnotationValue() const {
    202     assert(isAnnotation() && "Used AnnotVal on non-annotation token");
    203     return PtrData;
    204   }
    205   void setAnnotationValue(void *val) {
    206     assert(isAnnotation() && "Used AnnotVal on non-annotation token");
    207     PtrData = val;
    208   }
    209 
    210   /// setFlag - Set the specified flag.
    211   void setFlag(TokenFlags Flag) {
    212     Flags |= Flag;
    213   }
    214 
    215   /// clearFlag - Unset the specified flag.
    216   void clearFlag(TokenFlags Flag) {
    217     Flags &= ~Flag;
    218   }
    219 
    220   /// getFlags - Return the internal represtation of the flags.
    221   ///  Only intended for low-level operations such as writing tokens to
    222   //   disk.
    223   unsigned getFlags() const {
    224     return Flags;
    225   }
    226 
    227   /// setFlagValue - Set a flag to either true or false.
    228   void setFlagValue(TokenFlags Flag, bool Val) {
    229     if (Val)
    230       setFlag(Flag);
    231     else
    232       clearFlag(Flag);
    233   }
    234 
    235   /// isAtStartOfLine - Return true if this token is at the start of a line.
    236   ///
    237   bool isAtStartOfLine() const { return (Flags & StartOfLine) ? true : false; }
    238 
    239   /// hasLeadingSpace - Return true if this token has whitespace before it.
    240   ///
    241   bool hasLeadingSpace() const { return (Flags & LeadingSpace) ? true : false; }
    242 
    243   /// isExpandDisabled - Return true if this identifier token should never
    244   /// be expanded in the future, due to C99 6.10.3.4p2.
    245   bool isExpandDisabled() const {
    246     return (Flags & DisableExpand) ? true : false;
    247   }
    248 
    249   /// isObjCAtKeyword - Return true if we have an ObjC keyword identifier.
    250   bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const;
    251 
    252   /// getObjCKeywordID - Return the ObjC keyword kind.
    253   tok::ObjCKeywordKind getObjCKeywordID() const;
    254 
    255   /// needsCleaning - Return true if this token has trigraphs or escaped
    256   /// newlines in it.
    257   ///
    258   bool needsCleaning() const { return (Flags & NeedsCleaning) ? true : false; }
    259 
    260   /// \brief Return true if this token has an empty macro before it.
    261   ///
    262   bool hasLeadingEmptyMacro() const {
    263     return (Flags & LeadingEmptyMacro) ? true : false;
    264   }
    265 
    266 };
    267 
    268 /// PPConditionalInfo - Information about the conditional stack (#if directives)
    269 /// currently active.
    270 struct PPConditionalInfo {
    271   /// IfLoc - Location where the conditional started.
    272   ///
    273   SourceLocation IfLoc;
    274 
    275   /// WasSkipping - True if this was contained in a skipping directive, e.g.
    276   /// in a "#if 0" block.
    277   bool WasSkipping;
    278 
    279   /// FoundNonSkip - True if we have emitted tokens already, and now we're in
    280   /// an #else block or something.  Only useful in Skipping blocks.
    281   bool FoundNonSkip;
    282 
    283   /// FoundElse - True if we've seen a #else in this block.  If so,
    284   /// #elif/#else directives are not allowed.
    285   bool FoundElse;
    286 };
    287 
    288 }  // end namespace clang
    289 
    290 namespace llvm {
    291   template <>
    292   struct isPodLike<clang::Token> { static const bool value = true; };
    293 }  // end namespace llvm
    294 
    295 #endif
    296