RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-23 03:02:36 +01:00
Code Issues Projects Releases Wiki Activity

reimplement the tblgen lexer with a simple hand-written lexer. This eliminates

Browse Source 
one dependency on flex and gets rid of two ".cvs" files.

llvm-svn: 44210
This commit is contained in:
Chris Lattner 2007-11-18 02:57:27 +00:00
parent 2a8ef3f29a
commit 07a4b4d5fe
6 changed files with 516 additions and 2494 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2011
utils/TableGen/FileLexer.cpp.cvs
View File

File diff suppressed because it is too large Load Diff

240
utils/TableGen/FileLexer.l
View File

@ -1,240 +0,0 @@
/*===-- FileLexer.l - Scanner for TableGen Files ----------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file defines a simple flex scanner for TableGen files. This is pretty
// straight-forward, except for the magic to handle file inclusion.
//
//===----------------------------------------------------------------------===*/
%option prefix="File"
%option yylineno
%option nostdinit
%option never-interactive
%option batch
%option nodefault
%option 8bit
%option outfile="Lexer.cpp"
%option ecs
%option noreject
%option noyymore
%x in_comment
%{
#include "llvm/Config/config.h"
#include "llvm/Support/Streams.h"
#include "Record.h"
typedef std::pair<llvm::Record*, std::vector<llvm::Init*>*> SubClassRefTy;
#include "FileParser.h"
int Fileparse();
namespace llvm {
// Global variable recording the location of the include directory
std::vector<std::string> IncludeDirectories;
/// ParseInt - This has to handle the special case of binary numbers 0b0101
///
static int ParseInt(const char *Str) {
if (Str[0] == '0' && Str[1] == 'b')
return strtoll(Str+2, 0, 2);
return strtoll(Str, 0, 0);
}
static int CommentDepth = 0;
struct IncludeRec {
std::string Filename;
FILE *File;
unsigned LineNo;
YY_BUFFER_STATE Buffer;
IncludeRec(const std::string &FN, FILE *F)
: Filename(FN), File(F), LineNo(0){
}
};
static std::vector<IncludeRec> IncludeStack;
std::ostream &err() {
if (IncludeStack.empty()) {
cerr << "At end of input: ";
return *cerr.stream();
}
for (unsigned i = 0, e = IncludeStack.size()-1; i != e; ++i)
cerr << "Included from " << IncludeStack[i].Filename << ":"
<< IncludeStack[i].LineNo << ":\n";
cerr << "Parsing " << IncludeStack.back().Filename << ":"
<< Filelineno << ": ";
return *cerr.stream();
}
/// ParseFile - this function begins the parsing of the specified tablegen file.
///
void ParseFile(const std::string &Filename,
const std::vector<std::string> &IncludeDirs) {
FILE *F = stdin;
if (Filename != "-") {
F = fopen(Filename.c_str(), "r");
if (F == 0) {
cerr << "Could not open input file '" + Filename + "'!\n";
exit (1);
}
IncludeStack.push_back(IncludeRec(Filename, F));
} else {
IncludeStack.push_back(IncludeRec("<stdin>", stdin));
}
// Record the location of the include directory so that the lexer can find
// it later.
IncludeDirectories = IncludeDirs;
Filein = F;
Filelineno = 1;
Fileparse();
Filein = stdin;
}
/// HandleInclude - This function is called when an include directive is
/// encountered in the input stream...
///
static void HandleInclude(const char *Buffer) {
unsigned Length = yyleng;
assert(Buffer[Length-1] == '"');
Buffer += strlen("include ");
Length -= strlen("include ");
while (*Buffer != '"') {
++Buffer;
--Length;
}
assert(Length >= 2 && "Double quotes not found?");
std::string Filename(Buffer+1, Buffer+Length-1);
//cerr << "Filename = '" << Filename << "'\n";
// Save the line number and lex buffer of the includer...
IncludeStack.back().LineNo = Filelineno;
IncludeStack.back().Buffer = YY_CURRENT_BUFFER;
// Open the new input file...
yyin = fopen(Filename.c_str(), "r");
if (yyin == 0) {
// If we couldn't find the file in the current directory, look for it in
// the include directories.
//
std::string NextFilename;
for (unsigned i = 0, e = IncludeDirectories.size(); i != e; ++i) {
NextFilename = IncludeDirectories[i] + "/" + Filename;
if ((yyin = fopen(NextFilename.c_str(), "r")))
break;
}
if (yyin == 0) {
err() << "Could not find include file '" << Filename << "'!\n";
exit(1);
}
Filename = NextFilename;
}
// Add the file to our include stack...
IncludeStack.push_back(IncludeRec(Filename, yyin));
Filelineno = 1; // Reset line numbering...
//yyrestart(yyin); // Start lexing the new file...
yy_switch_to_buffer(yy_create_buffer(yyin, YY_BUF_SIZE));
}
/// yywrap - This is called when the lexer runs out of input in one of the
/// files. Switch back to an includer if an includee has run out of input.
///
extern "C"
int yywrap(void) {
if (IncludeStack.back().File != stdin)
fclose(IncludeStack.back().File);
IncludeStack.pop_back();
if (IncludeStack.empty()) return 1; // Top-level file is done.
// Otherwise, we need to switch back to a file which included the current one.
Filelineno = IncludeStack.back().LineNo; // Restore current line number
yy_switch_to_buffer(IncludeStack.back().Buffer);
return 0;
}
} // End llvm namespace
using namespace llvm;
%}
Comment \/\/.*
Identifier [a-zA-Z_][0-9a-zA-Z_]*
Integer [-+]?[0-9]+|0x[0-9a-fA-F]+|0b[01]+
CodeFragment \[\{([^}]+|\}[^\]])*\}\]
StringVal \"[^"]*\"
IncludeStr include[ \t\n]+\"[^"]*\"
%%
{Comment} { /* Ignore comments */ }
{IncludeStr} { HandleInclude(yytext); }
{CodeFragment} { Filelval.StrVal = new std::string(yytext+2, yytext+yyleng-2);
return CODEFRAGMENT; }
int { return INT; }
bit { return BIT; }
bits { return BITS; }
string { return STRING; }
list { return LIST; }
code { return CODE; }
dag { return DAG; }
class { return CLASS; }
def { return DEF; }
defm { return DEFM; }
multiclass { return MULTICLASS; }
field { return FIELD; }
let { return LET; }
in { return IN; }
!con { return CONCATTOK; }
!sra { return SRATOK; }
!srl { return SRLTOK; }
!shl { return SHLTOK; }
!strconcat { return STRCONCATTOK; }
{Identifier} { Filelval.StrVal = new std::string(yytext, yytext+yyleng);
return ID; }
${Identifier} { Filelval.StrVal = new std::string(yytext+1, yytext+yyleng);
return VARNAME; }
{StringVal} { Filelval.StrVal = new std::string(yytext+1, yytext+yyleng-1);
return STRVAL; }
{Integer} { Filelval.IntVal = ParseInt(Filetext); return INTVAL; }
[ \t\n\r]+ { /* Ignore whitespace */ }
"/*" { BEGIN(in_comment); CommentDepth++; }
<in_comment>[^*/]* {} /* eat anything that's not a '*' or '/' */
<in_comment>"*"+[^*/]* {} /* eat up '*'s not followed by '/'s */
<in_comment>"/*" { ++CommentDepth; }
<in_comment>"/"+[^*/]* {} /* eat up /'s not followed by *'s */
<in_comment>"*"+"/" { if (!--CommentDepth) { BEGIN(INITIAL); } }
<in_comment><<EOF>> { err() << "Unterminated comment!\n"; exit(1); }
. { return Filetext[0]; }
%%

240
utils/TableGen/FileLexer.l.cvs
View File

@ -1,240 +0,0 @@
/*===-- FileLexer.l - Scanner for TableGen Files ----------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file defines a simple flex scanner for TableGen files. This is pretty
// straight-forward, except for the magic to handle file inclusion.
//
//===----------------------------------------------------------------------===*/
%option prefix="File"
%option yylineno
%option nostdinit
%option never-interactive
%option batch
%option nodefault
%option 8bit
%option outfile="Lexer.cpp"
%option ecs
%option noreject
%option noyymore
%x in_comment
%{
#include "llvm/Config/config.h"
#include "llvm/Support/Streams.h"
#include "Record.h"
typedef std::pair<llvm::Record*, std::vector<llvm::Init*>*> SubClassRefTy;
#include "FileParser.h"
int Fileparse();
namespace llvm {
// Global variable recording the location of the include directory
std::vector<std::string> IncludeDirectories;
/// ParseInt - This has to handle the special case of binary numbers 0b0101
///
static int ParseInt(const char *Str) {
if (Str[0] == '0' && Str[1] == 'b')
return strtoll(Str+2, 0, 2);
return strtoll(Str, 0, 0);
}
static int CommentDepth = 0;
struct IncludeRec {
std::string Filename;
FILE *File;
unsigned LineNo;
YY_BUFFER_STATE Buffer;
IncludeRec(const std::string &FN, FILE *F)
: Filename(FN), File(F), LineNo(0){
}
};
static std::vector<IncludeRec> IncludeStack;
std::ostream &err() {
if (IncludeStack.empty()) {
cerr << "At end of input: ";
return *cerr.stream();
}
for (unsigned i = 0, e = IncludeStack.size()-1; i != e; ++i)
cerr << "Included from " << IncludeStack[i].Filename << ":"
<< IncludeStack[i].LineNo << ":\n";
cerr << "Parsing " << IncludeStack.back().Filename << ":"
<< Filelineno << ": ";
return *cerr.stream();
}
/// ParseFile - this function begins the parsing of the specified tablegen file.
///
void ParseFile(const std::string &Filename,
const std::vector<std::string> &IncludeDirs) {
FILE *F = stdin;
if (Filename != "-") {
F = fopen(Filename.c_str(), "r");
if (F == 0) {
cerr << "Could not open input file '" + Filename + "'!\n";
exit (1);
}
IncludeStack.push_back(IncludeRec(Filename, F));
} else {
IncludeStack.push_back(IncludeRec("<stdin>", stdin));
}
// Record the location of the include directory so that the lexer can find
// it later.
IncludeDirectories = IncludeDirs;
Filein = F;
Filelineno = 1;
Fileparse();
Filein = stdin;
}
/// HandleInclude - This function is called when an include directive is
/// encountered in the input stream...
///
static void HandleInclude(const char *Buffer) {
unsigned Length = yyleng;
assert(Buffer[Length-1] == '"');
Buffer += strlen("include ");
Length -= strlen("include ");
while (*Buffer != '"') {
++Buffer;
--Length;
}
assert(Length >= 2 && "Double quotes not found?");
std::string Filename(Buffer+1, Buffer+Length-1);
//cerr << "Filename = '" << Filename << "'\n";
// Save the line number and lex buffer of the includer...
IncludeStack.back().LineNo = Filelineno;
IncludeStack.back().Buffer = YY_CURRENT_BUFFER;
// Open the new input file...
yyin = fopen(Filename.c_str(), "r");
if (yyin == 0) {
// If we couldn't find the file in the current directory, look for it in
// the include directories.
//
std::string NextFilename;
for (unsigned i = 0, e = IncludeDirectories.size(); i != e; ++i) {
NextFilename = IncludeDirectories[i] + "/" + Filename;
if ((yyin = fopen(NextFilename.c_str(), "r")))
break;
}
if (yyin == 0) {
err() << "Could not find include file '" << Filename << "'!\n";
exit(1);
}
Filename = NextFilename;
}
// Add the file to our include stack...
IncludeStack.push_back(IncludeRec(Filename, yyin));
Filelineno = 1; // Reset line numbering...
//yyrestart(yyin); // Start lexing the new file...
yy_switch_to_buffer(yy_create_buffer(yyin, YY_BUF_SIZE));
}
/// yywrap - This is called when the lexer runs out of input in one of the
/// files. Switch back to an includer if an includee has run out of input.
///
extern "C"
int yywrap(void) {
if (IncludeStack.back().File != stdin)
fclose(IncludeStack.back().File);
IncludeStack.pop_back();
if (IncludeStack.empty()) return 1; // Top-level file is done.
// Otherwise, we need to switch back to a file which included the current one.
Filelineno = IncludeStack.back().LineNo; // Restore current line number
yy_switch_to_buffer(IncludeStack.back().Buffer);
return 0;
}
} // End llvm namespace
using namespace llvm;
%}
Comment \/\/.*
Identifier [a-zA-Z_][0-9a-zA-Z_]*
Integer [-+]?[0-9]+|0x[0-9a-fA-F]+|0b[01]+
CodeFragment \[\{([^}]+|\}[^\]])*\}\]
StringVal \"[^"]*\"
IncludeStr include[ \t\n]+\"[^"]*\"
%%
{Comment} { /* Ignore comments */ }
{IncludeStr} { HandleInclude(yytext); }
{CodeFragment} { Filelval.StrVal = new std::string(yytext+2, yytext+yyleng-2);
return CODEFRAGMENT; }
int { return INT; }
bit { return BIT; }
bits { return BITS; }
string { return STRING; }
list { return LIST; }
code { return CODE; }
dag { return DAG; }
class { return CLASS; }
def { return DEF; }
defm { return DEFM; }
multiclass { return MULTICLASS; }
field { return FIELD; }
let { return LET; }
in { return IN; }
!con { return CONCATTOK; }
!sra { return SRATOK; }
!srl { return SRLTOK; }
!shl { return SHLTOK; }
!strconcat { return STRCONCATTOK; }
{Identifier} { Filelval.StrVal = new std::string(yytext, yytext+yyleng);
return ID; }
${Identifier} { Filelval.StrVal = new std::string(yytext+1, yytext+yyleng);
return VARNAME; }
{StringVal} { Filelval.StrVal = new std::string(yytext+1, yytext+yyleng-1);
return STRVAL; }
{Integer} { Filelval.IntVal = ParseInt(Filetext); return INTVAL; }
[ \t\n\r]+ { /* Ignore whitespace */ }
"/*" { BEGIN(in_comment); CommentDepth++; }
<in_comment>[^*/]* {} /* eat anything that's not a '*' or '/' */
<in_comment>"*"+[^*/]* {} /* eat up '*'s not followed by '/'s */
<in_comment>"/*" { ++CommentDepth; }
<in_comment>"/"+[^*/]* {} /* eat up /'s not followed by *'s */
<in_comment>"*"+"/" { if (!--CommentDepth) { BEGIN(INITIAL); } }
<in_comment><<EOF>> { err() << "Unterminated comment!\n"; exit(1); }
. { return Filetext[0]; }
%%

6
utils/TableGen/Makefile
View File

@ -11,8 +11,7 @@ LEVEL = ../..
TOOLNAME = tblgen
NO_INSTALL = 1;
USEDLIBS = LLVMSupport.a LLVMSystem.a
EXTRA_DIST = FileLexer.cpp.cvs FileLexer.l.cvs \
FileParser.cpp.cvs FileParser.h.cvs FileParser.y.cvs
EXTRA_DIST = FileParser.cpp.cvs FileParser.h.cvs FileParser.y.cvs
REQUIRES_EH := 1
REQUIRES_RTTI := 1
@ -27,4 +26,5 @@ CompileCommonOpts := $(filter-out -Wno-long-long,$(CompileCommonOpts))
# (which depend on the source file) won't get generated until bison is done
# generating the C source and header files for the parser.
#
$(ObjDir)/FileLexer.o : $(PROJ_SRC_DIR)/FileParser.h
$(ObjDir)/TGLexer.o : $(PROJ_SRC_DIR)/FileParser.h

439
utils/TableGen/TGLexer.cpp Normal file
View File

@ -0,0 +1,439 @@
//===- TGLexer.cpp - Lexer for TableGen -----------------------------------===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by Chris Lattner and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// Implement the Lexer for TableGen.
//
//===----------------------------------------------------------------------===//
#include "Record.h"
#include "llvm/Support/Streams.h"
#include "Record.h"
#include "TGLexer.h"
#include "llvm/Support/MemoryBuffer.h"
typedef std::pair<llvm::Record*, std::vector<llvm::Init*>*> SubClassRefTy;
#include "FileParser.h"
#include <cctype>
using namespace llvm;
// FIXME: REMOVE THIS.
#define YYEOF 0
#define YYERROR -2
TGLexer::TGLexer(MemoryBuffer *StartBuf) : CurLineNo(1), CurBuf(StartBuf) {
CurPtr = CurBuf->getBufferStart();
}
TGLexer::~TGLexer() {
while (!IncludeStack.empty()) {
delete IncludeStack.back().Buffer;
IncludeStack.pop_back();
}
delete CurBuf;
}
std::ostream &TGLexer::err() {
PrintIncludeStack(*cerr.stream());
return *cerr.stream();
}
void TGLexer::PrintIncludeStack(std::ostream &OS) {
for (unsigned i = 0, e = IncludeStack.size(); i != e; ++i)
OS << "Included from " << IncludeStack[i].Buffer->getBufferIdentifier()
<< ":" << IncludeStack[i].LineNo << ":\n";
OS << "Parsing " << CurBuf->getBufferIdentifier() << ":"
<< CurLineNo << ": ";
}
int TGLexer::getNextChar() {
char CurChar = *CurPtr++;
switch (CurChar) {
default:
return CurChar;
case 0:
// A nul character in the stream is either the end of the current buffer or
// a random nul in the file. Disambiguate that here.
if (CurPtr-1 != CurBuf->getBufferEnd())
return 0; // Just whitespace.
// If this is the end of an included file, pop the parent file off the
// include stack.
if (!IncludeStack.empty()) {
delete CurBuf;
CurBuf = IncludeStack.back().Buffer;
CurLineNo = IncludeStack.back().LineNo;
CurPtr = IncludeStack.back().CurPtr;
IncludeStack.pop_back();
return getNextChar();
}
// Otherwise, return end of file.
--CurPtr; // Another call to lex will return EOF again.
return EOF;
case '\n':
case '\r':
// Handle the newline character by ignoring it and incrementing the line
// count. However, be careful about 'dos style' files with \n\r in them.
// Only treat a \n\r or \r\n as a single line.
if ((*CurPtr == '\n' || (*CurPtr == '\r')) &&
*CurPtr != CurChar)
++CurPtr; // Each the two char newline sequence.
++CurLineNo;
return '\n';
}
}
int TGLexer::LexToken() {
// This always consumes at least one character.
int CurChar = getNextChar();
switch (CurChar) {
default:
// Handle letters: [a-zA-Z_]
if (isalpha(CurChar) || CurChar == '_')
return LexIdentifier();
// Unknown character, return the char itself.
return (unsigned char)CurChar;
case EOF: return YYEOF;
case 0:
case ' ':
case '\t':
case '\n':
case '\r':
// Ignore whitespace.
return LexToken();
case '/':
// If this is the start of a // comment, skip until the end of the line or
// the end of the buffer.
if (*CurPtr == '/')
SkipBCPLComment();
else if (*CurPtr == '*') {
if (SkipCComment())
return YYERROR;
} else // Otherwise, return this / as a token.
return CurChar;
return LexToken();
case '-': case '+':
case '0': case '1': case '2': case '3': case '4': case '5': case '6':
case '7': case '8': case '9':
return LexNumber();
case '"': return LexString();
case '$': return LexVarName();
case '[': return LexBracket();
case '!': return LexExclaim();
}
}
/// LexString - Lex "[^"]*"
int TGLexer::LexString() {
const char *StrStart = CurPtr;
while (*CurPtr != '"') {
// If we hit the end of the buffer, report an error.
if (*CurPtr == 0 && CurPtr == CurBuf->getBufferEnd()) {
TheError = "End of file in string literal";
return YYERROR;
} else if (*CurPtr == '\n' || *CurPtr == '\r') {
TheError = "End of line in string literal";
return YYERROR;
}
++CurPtr;
}
Filelval.StrVal = new std::string(StrStart, CurPtr);
++CurPtr;
return STRVAL;
}
int TGLexer::LexVarName() {
if (!isalpha(CurPtr[0]) && CurPtr[0] != '_')
return '$'; // Invalid varname.
// Otherwise, we're ok, consume the rest of the characters.
const char *VarNameStart = CurPtr++;
while (isalpha(*CurPtr) || isdigit(*CurPtr) || *CurPtr == '_')
++CurPtr;
Filelval.StrVal = new std::string(VarNameStart, CurPtr);
return VARNAME;
}
int TGLexer::LexIdentifier() {
// The first letter is [a-zA-Z_].
const char *IdentStart = CurPtr-1;
// Match the rest of the identifier regex: [0-9a-zA-Z_]*
while (isalpha(*CurPtr) || isdigit(*CurPtr) || *CurPtr == '_')
++CurPtr;
// Check to see if this identifier is a keyword.
unsigned Len = CurPtr-IdentStart;
if (Len == 3 && !memcmp(IdentStart, "int", 3)) return INT;
if (Len == 3 && !memcmp(IdentStart, "bit", 3)) return BIT;
if (Len == 4 && !memcmp(IdentStart, "bits", 4)) return BITS;
if (Len == 6 && !memcmp(IdentStart, "string", 6)) return STRING;
if (Len == 4 && !memcmp(IdentStart, "list", 4)) return LIST;
if (Len == 4 && !memcmp(IdentStart, "code", 4)) return CODE;
if (Len == 3 && !memcmp(IdentStart, "dag", 3)) return DAG;
if (Len == 5 && !memcmp(IdentStart, "class", 5)) return CLASS;
if (Len == 3 && !memcmp(IdentStart, "def", 3)) return DEF;
if (Len == 4 && !memcmp(IdentStart, "defm", 4)) return DEFM;
if (Len == 10 && !memcmp(IdentStart, "multiclass", 10)) return MULTICLASS;
if (Len == 5 && !memcmp(IdentStart, "field", 5)) return FIELD;
if (Len == 3 && !memcmp(IdentStart, "let", 3)) return LET;
if (Len == 2 && !memcmp(IdentStart, "in", 2)) return IN;
if (Len == 7 && !memcmp(IdentStart, "include", 7)) {
if (LexInclude()) return YYERROR;
return LexToken();
}
Filelval.StrVal = new std::string(IdentStart, CurPtr);
return ID;
}
/// LexInclude - We just read the "include" token. Get the string token that
/// comes next and enter the include.
bool TGLexer::LexInclude() {
// The token after the include must be a string.
int Tok = LexToken();
if (Tok == YYERROR) return true;
if (Tok != STRVAL) {
TheError = "Expected filename after include";
return true;
}
// Get the string.
std::string Filename = *Filelval.StrVal;
delete Filelval.StrVal;
// Try to find the file.
MemoryBuffer *NewBuf = MemoryBuffer::getFile(&Filename[0], Filename.size());
// If the file didn't exist directly, see if it's in an include path.
for (unsigned i = 0, e = IncludeDirectories.size(); i != e && !NewBuf; ++i) {
std::string IncFile = IncludeDirectories[i] + "/" + Filename;
NewBuf = MemoryBuffer::getFile(&IncFile[0], IncFile.size());
}
if (NewBuf == 0) {
TheError = "Could not find include file '" + Filename + "'";
return true;
}
// Save the line number and lex buffer of the includer.
IncludeStack.push_back(IncludeRec(CurBuf, CurPtr, CurLineNo));
CurLineNo = 1; // Reset line numbering.
CurBuf = NewBuf;
CurPtr = CurBuf->getBufferStart();
return false;
}
void TGLexer::SkipBCPLComment() {
++CurPtr; // skip the second slash.
while (1) {
switch (*CurPtr) {
case '\n':
case '\r':
return; // Newline is end of comment.
case 0:
// If this is the end of the buffer, end the comment.
if (CurPtr == CurBuf->getBufferEnd())
return;
break;
}
// Otherwise, skip the character.
++CurPtr;
}
}
/// SkipCComment - This skips C-style /**/ comments. The only difference from C
/// is that we allow nesting.
bool TGLexer::SkipCComment() {
++CurPtr; // skip the star.
unsigned CommentDepth = 1;
while (1) {
int CurChar = getNextChar();
switch (CurChar) {
case EOF:
TheError = "Unterminated comment!";
return true;
case '*':
// End of the comment?
if (CurPtr[0] != '/') break;
++CurPtr; // End the */.
if (--CommentDepth == 0)
return false;
break;
case '/':
// Start of a nested comment?
if (CurPtr[0] != '*') break;
++CurPtr;
++CommentDepth;
break;
}
}
}
/// LexNumber - Lex:
/// [-+]?[0-9]+
/// 0x[0-9a-fA-F]+
/// 0b[01]+
int TGLexer::LexNumber() {
const char *NumStart = CurPtr-1;
if (CurPtr[-1] == '0') {
if (CurPtr[0] == 'x') {
++CurPtr;
NumStart = CurPtr;
while (isxdigit(CurPtr[0]))
++CurPtr;
if (CurPtr == NumStart) {
TheError = "Invalid hexadecimal number";
return YYERROR;
}
Filelval.IntVal = strtoll(NumStart, 0, 16);
return INTVAL;
} else if (CurPtr[0] == 'b') {
++CurPtr;
NumStart = CurPtr;
while (CurPtr[0] == '0' || CurPtr[0] == '1')
++CurPtr;
if (CurPtr == NumStart) {
TheError = "Invalid binary number";
return YYERROR;
}
Filelval.IntVal = strtoll(NumStart, 0, 2);
return INTVAL;
}
}
// Check for a sign without a digit.
if (CurPtr[-1] == '-' || CurPtr[-1] == '+') {
if (!isdigit(CurPtr[0]))
return CurPtr[-1];
}
while (isdigit(CurPtr[0]))
++CurPtr;
Filelval.IntVal = strtoll(NumStart, 0, 10);
return INTVAL;
}
/// LexBracket - We just read '['. If this is a code block, return it,
/// otherwise return the bracket. Match: '[' and '[{ ( [^}]+ | }[^]] )* }]'
int TGLexer::LexBracket() {
if (CurPtr[0] != '{')
return '[';
++CurPtr;
const char *CodeStart = CurPtr;
while (1) {
int Char = getNextChar();
if (Char == EOF) break;
if (Char != '}') continue;
Char = getNextChar();
if (Char == EOF) break;
if (Char == ']') {
Filelval.StrVal = new std::string(CodeStart, CurPtr-2);
return CODEFRAGMENT;
}
}
TheError = "Invalid Code Block";
return YYERROR;
}
/// LexExclaim - Lex '!' and '![a-zA-Z]+'.
int TGLexer::LexExclaim() {
if (!isalpha(*CurPtr))
return '!';
const char *Start = CurPtr++;
while (isalpha(*CurPtr))
++CurPtr;
// Check to see which operator this is.
unsigned Len = CurPtr-Start;
if (Len == 3 && !memcmp(Start, "con", 3)) return CONCATTOK;
if (Len == 3 && !memcmp(Start, "sra", 3)) return SRATOK;
if (Len == 3 && !memcmp(Start, "srl", 3)) return SRLTOK;
if (Len == 3 && !memcmp(Start, "shl", 3)) return SHLTOK;
if (Len == 9 && !memcmp(Start, "strconcat", 9)) return STRCONCATTOK;
TheError = "Unknown operator";
return YYERROR;
}
//===----------------------------------------------------------------------===//
// Interfaces used by the Bison parser.
//===----------------------------------------------------------------------===//
int Fileparse();
static TGLexer *TheLexer;
namespace llvm {
std::ostream &err() {
return TheLexer->err();
}
/// ParseFile - this function begins the parsing of the specified tablegen
/// file.
///
void ParseFile(const std::string &Filename,
const std::vector<std::string> &IncludeDirs) {
std::string ErrorStr;
MemoryBuffer *F = MemoryBuffer::getFileOrSTDIN(&Filename[0], Filename.size(),
&ErrorStr);
if (F == 0) {
cerr << "Could not open input file '" + Filename + "': " << ErrorStr <<"\n";
exit(1);
}
assert(!TheLexer && "Lexer isn't reentrant yet!");
TheLexer = new TGLexer(F);
// Record the location of the include directory so that the lexer can find
// it later.
TheLexer->setIncludeDirs(IncludeDirs);
Fileparse();
// Cleanup
delete TheLexer;
TheLexer = 0;
}
} // End llvm namespace
int Filelex() {
assert(TheLexer && "No lexer setup yet!");
int Tok = TheLexer->LexToken();
if (Tok == YYERROR) {
err() << TheLexer->getError() << "\n";
exit(1);
}
return Tok;
}

74
utils/TableGen/TGLexer.h Normal file
View File

@ -0,0 +1,74 @@
//===- TGLexer.h - Lexer for TableGen Files ---------------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by Chris Lattner and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This class represents the Lexer for tablegen files.
//
//===----------------------------------------------------------------------===//
#ifndef TGLEXER_H
#define TGLEXER_H
#include <vector>
#include <string>
#include <iosfwd>
namespace llvm {
class MemoryBuffer;
class TGLexer {
const char *CurPtr;
unsigned CurLineNo;
MemoryBuffer *CurBuf;
/// IncludeRec / IncludeStack - This captures the current set of include
/// directives we are nested within.
struct IncludeRec {
MemoryBuffer *Buffer;
const char *CurPtr;
unsigned LineNo;
IncludeRec(MemoryBuffer *buffer, const char *curPtr, unsigned lineNo)
: Buffer(buffer), CurPtr(curPtr), LineNo(lineNo) {}
};
std::vector<IncludeRec> IncludeStack;
// IncludeDirectories - This is the list of directories we should search for
// include files in.
std::vector<std::string> IncludeDirectories;
std::string TheError;
public:
TGLexer(MemoryBuffer *StartBuf);
~TGLexer();
void setIncludeDirs(const std::vector<std::string> &Dirs) {
IncludeDirectories = Dirs;
}
int LexToken();
const std::string getError() const { return TheError; }
std::ostream &err();
void PrintIncludeStack(std::ostream &OS);
private:
int getNextChar();
void SkipBCPLComment();
bool SkipCComment();
int LexIdentifier();
bool LexInclude();
int LexString();
int LexVarName();
int LexNumber();
int LexBracket();
int LexExclaim();
};
} // end namespace llvm
#endif