Make preprocessor keywords like 'define' first class citizens in the

IdentifierTable, instead of making them resort to strcmp'ing.

llvm-svn: 39002
This commit is contained in:
Chris Lattner 2006-10-17 03:44:32 +00:00
parent b95cca0d33
commit 87d3bec423
6 changed files with 171 additions and 84 deletions

View File

@ -207,6 +207,7 @@ IdentifierInfo &IdentifierTable::get(const char *NameStart,
Identifier->TokInfo.NameLen = Length;
Identifier->TokInfo.Macro = 0;
Identifier->TokInfo.TokenID = tok::identifier;
Identifier->TokInfo.PPID = tok::pp_not_keyword;
Identifier->TokInfo.IsExtension = false;
Identifier->TokInfo.IsPoisoned = false;
Identifier->TokInfo.IsOtherTargetMacro = false;

View File

@ -102,6 +102,12 @@ Preprocessor::PerFileInfo &Preprocessor::getFileInfo(const FileEntry *FE) {
return FileInfo[FE->getUID()];
}
/// AddPPKeyword - Register a preprocessor keyword like "define" "undef" or
/// "elif".
static void AddPPKeyword(tok::PPKeywordKind PPID,
const char *Name, unsigned NameLen, Preprocessor &PP) {
PP.getIdentifierInfo(Name, Name+NameLen)->setPPKeywordID(PPID);
}
/// AddKeywords - Add all keywords to the symbol table.
///
@ -127,9 +133,34 @@ void Preprocessor::AddKeywords() {
((FLAGS) >> CPPShift) & Mask);
#define ALIAS(NAME, TOK) \
AddKeyword(NAME, tok::kw_ ## TOK, 0, 0, 0);
#define PPKEYWORD(NAME) \
AddPPKeyword(tok::pp_##NAME, #NAME, strlen(#NAME), *this);
#include "clang/Basic/TokenKinds.def"
}
/// AddKeyword - This method is used to associate a token ID with specific
/// identifiers because they are language keywords. This causes the lexer to
/// automatically map matching identifiers to specialized token codes.
///
/// The C90/C99/CPP flags are set to 0 if the token should be enabled in the
/// specified langauge, set to 1 if it is an extension in the specified
/// language, and set to 2 if disabled in the specified language.
void Preprocessor::AddKeyword(const std::string &Keyword,
tok::TokenKind TokenCode,
int C90, int C99, int CPP) {
int Flags = Features.CPlusPlus ? CPP : (Features.C99 ? C99 : C90);
// Don't add this keyword if disabled in this language or if an extension
// and extensions are disabled.
if (Flags+Features.NoExtensions >= 2) return;
const char *Str = &Keyword[0];
IdentifierInfo &Info = *getIdentifierInfo(Str, Str+Keyword.size());
Info.setTokenID(TokenCode);
Info.setIsExtensionToken(Flags == 1);
}
/// Diag - Forwarding function for diagnostics. This emits a diagnostic at
/// the specified LexerToken's location, translating the token's start
/// position in the current buffer into a SourcePosition object for rendering.
@ -1405,7 +1436,6 @@ void Preprocessor::HandleDirective(LexerToken &Result) {
Diag(Result, diag::ext_embedded_directive);
switch (Result.getKind()) {
default: break;
case tok::eom:
return; // null directive.
@ -1413,75 +1443,76 @@ void Preprocessor::HandleDirective(LexerToken &Result) {
// FIXME: implement # 7 line numbers!
DiscardUntilEndOfDirective();
return;
case tok::kw_else:
return HandleElseDirective(Result);
case tok::kw_if:
default:
IdentifierInfo *II = Result.getIdentifierInfo();
if (II == 0) break; // Not an identifier.
// Ask what the preprocessor keyword ID is.
switch (II->getPPKeywordID()) {
default: break;
// C99 6.10.1 - Conditional Inclusion.
case tok::pp_if:
return HandleIfDirective(Result, ReadAnyTokensBeforeDirective);
case tok::identifier:
// Get the identifier name without trigraphs or embedded newlines.
const char *Directive = Result.getIdentifierInfo()->getName();
bool isExtension = false;
switch (Result.getIdentifierInfo()->getNameLength()) {
case 4:
if (Directive[0] == 'l' && !strcmp(Directive, "line")) {
case tok::pp_ifdef:
return HandleIfdefDirective(Result, false, true/*not valid for miopt*/);
case tok::pp_ifndef:
return HandleIfdefDirective(Result, true, ReadAnyTokensBeforeDirective);
case tok::pp_elif:
return HandleElifDirective(Result);
case tok::pp_else:
return HandleElseDirective(Result);
case tok::pp_endif:
return HandleEndifDirective(Result);
// C99 6.10.2 - Source File Inclusion.
case tok::pp_include:
return HandleIncludeDirective(Result); // Handle #include.
// C99 6.10.3 - Macro Replacement.
case tok::pp_define:
return HandleDefineDirective(Result, false);
case tok::pp_undef:
return HandleUndefDirective(Result);
// C99 6.10.4 - Line Control.
case tok::pp_line:
// FIXME: implement #line
DiscardUntilEndOfDirective();
return;
}
if (Directive[0] == 'e' && !strcmp(Directive, "elif"))
return HandleElifDirective(Result);
if (Directive[0] == 's' && !strcmp(Directive, "sccs"))
return HandleIdentSCCSDirective(Result);
break;
case 5:
if (Directive[0] == 'e' && !strcmp(Directive, "endif"))
return HandleEndifDirective(Result);
if (Directive[0] == 'i' && !strcmp(Directive, "ifdef"))
return HandleIfdefDirective(Result, false, true/*not valid for miopt*/);
if (Directive[0] == 'u' && !strcmp(Directive, "undef"))
return HandleUndefDirective(Result);
if (Directive[0] == 'e' && !strcmp(Directive, "error"))
// C99 6.10.5 - Error Directive.
case tok::pp_error:
return HandleUserDiagnosticDirective(Result, false);
if (Directive[0] == 'i' && !strcmp(Directive, "ident"))
return HandleIdentSCCSDirective(Result);
break;
case 6:
if (Directive[0] == 'd' && !strcmp(Directive, "define"))
return HandleDefineDirective(Result, false);
if (Directive[0] == 'i' && !strcmp(Directive, "ifndef"))
return HandleIfdefDirective(Result, true, ReadAnyTokensBeforeDirective);
if (Directive[0] == 'i' && !strcmp(Directive, "import"))
return HandleImportDirective(Result);
if (Directive[0] == 'p' && !strcmp(Directive, "pragma"))
// C99 6.10.6 - Pragma Directive.
case tok::pp_pragma:
return HandlePragmaDirective();
if (Directive[0] == 'a' && !strcmp(Directive, "assert"))
isExtension = true; // FIXME: implement #assert
break;
case 7:
if (Directive[0] == 'i' && !strcmp(Directive, "include"))
return HandleIncludeDirective(Result); // Handle #include.
if (Directive[0] == 'w' && !strcmp(Directive, "warning")) {
// GNU Extensions.
case tok::pp_import:
return HandleImportDirective(Result);
case tok::pp_include_next:
return HandleIncludeNextDirective(Result);
case tok::pp_warning:
Diag(Result, diag::ext_pp_warning_directive);
return HandleUserDiagnosticDirective(Result, true);
}
case tok::pp_ident:
return HandleIdentSCCSDirective(Result);
case tok::pp_sccs:
return HandleIdentSCCSDirective(Result);
case tok::pp_assert:
//isExtension = true; // FIXME: implement #assert
break;
case 8:
if (Directive[0] == 'u' && !strcmp(Directive, "unassert")) {
isExtension = true; // FIXME: implement #unassert
}
case tok::pp_unassert:
//isExtension = true; // FIXME: implement #unassert
break;
case 12:
if (Directive[0] == 'i' && !strcmp(Directive, "include_next"))
return HandleIncludeNextDirective(Result); // Handle #include_next.
break;
case 13:
if (Directive[0] == 'd' && !strcmp(Directive, "define_target"))
// clang extensions.
case tok::pp_define_target:
return HandleDefineDirective(Result, true);
break;
case 19:
if (Directive[0] == 'd' && !strcmp(Directive, "define_other_target"))
case tok::pp_define_other_target:
return HandleDefineOtherTargetDirective(Result);
break;
}
break;
}

View File

@ -9,8 +9,8 @@
//
// This file defines the TokenKind database. This includes normal tokens like
// tok::ampamp (corresponding to the && token) as well as keywords for various
// languages. Users of this file must optionally #define the TOK/KEYWORD/ALIAS
// macros to make use of this file.
// languages. Users of this file must optionally #define the TOK, KEYWORD,
// ALIAS, or PPKEYWORD macros to make use of this file.
//
//===----------------------------------------------------------------------===//
@ -23,6 +23,61 @@
#ifndef ALIAS
#define ALIAS(X,Y)
#endif
#ifndef PPKEYWORD
#define PPKEYWORD(X)
#endif
//===----------------------------------------------------------------------===//
// Preprocessor keywords.
//===----------------------------------------------------------------------===//
// These have meaning after a '#' at the start of a line. These define enums in
// the tok::pp_* namespace.
PPKEYWORD(not_keyword)
// C99 6.10.1 - Conditional Inclusion.
PPKEYWORD(if)
PPKEYWORD(ifdef)
PPKEYWORD(ifndef)
PPKEYWORD(elif)
PPKEYWORD(else)
PPKEYWORD(endif)
// C99 6.10.2 - Source File Inclusion.
PPKEYWORD(include)
// C99 6.10.3 - Macro Replacement.
PPKEYWORD(define)
PPKEYWORD(undef)
// C99 6.10.4 - Line Control.
PPKEYWORD(line)
// C99 6.10.5 - Error Directive.
PPKEYWORD(error)
// C99 6.10.6 - Pragma Directive.
PPKEYWORD(pragma)
// GNU Extensions.
PPKEYWORD(import)
PPKEYWORD(include_next)
PPKEYWORD(warning)
PPKEYWORD(ident)
PPKEYWORD(sccs)
PPKEYWORD(assert)
PPKEYWORD(unassert)
// clang extensions.
PPKEYWORD(define_target)
PPKEYWORD(define_other_target)
//===----------------------------------------------------------------------===//
// Language keywords.
//===----------------------------------------------------------------------===//
// These define members of the tok::kw_* namespace.
TOK(unknown) // Not a token.
TOK(eof) // End of file.
@ -239,6 +294,7 @@ ALIAS("__typeof__" , typeof )
ALIAS("__volatile" , volatile )
ALIAS("__volatile__" , volatile )
#undef PPKEYWORD
#undef ALIAS
#undef KEYWORD
#undef TOK

View File

@ -27,6 +27,12 @@ enum TokenKind {
NUM_TOKENS
};
enum PPKeywordKind {
#define PPKEYWORD(X) pp_##X,
#include "clang/Basic/TokenKinds.def"
NUM_PP_KEYWORDS
};
const char *getTokenName(enum TokenKind Kind);
} // end namespace tok

View File

@ -32,6 +32,7 @@ class IdentifierInfo {
unsigned NameLen; // String that is the identifier.
MacroInfo *Macro; // Set if this identifier is #define'd.
tok::TokenKind TokenID : 8; // Front-end token ID or tok::identifier.
tok::PPKeywordKind PPID : 5; // ID for preprocessor command like 'ifdef'.
bool IsExtension : 1; // True if identifier is a lang extension.
bool IsPoisoned : 1; // True if identifier is poisoned.
bool IsOtherTargetMacro : 1; // True if ident is a macro on another target.
@ -64,6 +65,9 @@ public:
tok::TokenKind getTokenID() const { return TokenID; }
void setTokenID(tok::TokenKind ID) { TokenID = ID; }
tok::PPKeywordKind getPPKeywordID() const { return PPID; }
void setPPKeywordID(tok::PPKeywordKind ID) { PPID = ID; }
/// get/setExtension - Initialize information about whether or not this
/// language token is an extension. This controls extension warnings, and is
/// only valid if a custom token ID is set.

View File

@ -276,18 +276,7 @@ public:
/// specified langauge, set to 1 if it is an extension in the specified
/// language, and set to 2 if disabled in the specified language.
void AddKeyword(const std::string &Keyword, tok::TokenKind TokenCode,
int C90, int C99, int CPP) {
int Flags = Features.CPlusPlus ? CPP : (Features.C99 ? C99 : C90);
// Don't add this keyword if disabled in this language or if an extension
// and extensions are disabled.
if (Flags+Features.NoExtensions >= 2) return;
const char *Str = &Keyword[0];
IdentifierInfo &Info = *getIdentifierInfo(Str, Str+Keyword.size());
Info.setTokenID(TokenCode);
Info.setIsExtensionToken(Flags == 1);
}
int C90, int C99, int CPP);
/// AddKeywords - Add all keywords to the symbol table.
///