scanner.c

来自「a little DFA compiler.」· C语言 代码 · 共 1,421 行 · 第 1/2 页

/* Generated by re2c */
#line 1 "scanner.re"
/* Id: scanner.re,v 1.37 2006/02/25 12:41:41 helly Exp */
#include <stdlib.h>
#include <string.h>
#include <iostream>
#include <sstream>
#include "scanner.h"
#include "parser.h"
#include "y.tab.h"
#include "globals.h"
#include "dfa.h"

extern YYSTYPE yylval;

#ifndef MAX
#define MAX(a,b) (((a)>(b))?(a):(b))
#endif

#define	BSIZE	8192

#define	YYCTYPE		char
#define	YYCURSOR	cursor
#define	YYLIMIT		lim
#define	YYMARKER	ptr
#define	YYFILL(n)	{cursor = fill(cursor);}

#define	RETURN(i)	{cur = cursor; return i;}

namespace re2c
{

Scanner::Scanner(std::istream& i, std::ostream& o)
	: in(i)
	, out(o)
	, bot(NULL), tok(NULL), ptr(NULL), cur(NULL), pos(NULL), lim(NULL)
	, top(NULL), eof(NULL), tchar(0), tline(0), cline(1), iscfg(0)
{
    ;
}

char *Scanner::fill(char *cursor)
{
	if(!eof)
	{
		uint cnt = tok - bot;
		if(cnt)
		{
			memcpy(bot, tok, lim - tok);
			tok = bot;
			ptr -= cnt;
			cursor -= cnt;
			pos -= cnt;
			lim -= cnt;
		}
		if((top - lim) < BSIZE)
		{
			char *buf = new char[(lim - bot) + BSIZE];
			memcpy(buf, tok, lim - tok);
			tok = buf;
			ptr = &buf[ptr - bot];
			cursor = &buf[cursor - bot];
			pos = &buf[pos - bot];
			lim = &buf[lim - bot];
			top = &lim[BSIZE];
			delete [] bot;
			bot = buf;
		}
		in.read(lim, BSIZE);
		if ((cnt = in.gcount()) != BSIZE )
		{
			eof = &lim[cnt]; *eof++ = '\0';
		}
		lim += cnt;
	}
	return cursor;
}

#line 95 "scanner.re"


int Scanner::echo()
{
    char *cursor = cur;
    bool ignore_eoc = false;

    if (eof && cursor == eof) // Catch EOF
	{
    	return 0;
	}

    tok = cursor;
echo:

#line 96 "<stdout>"
{
	YYCTYPE yych;

	if((YYLIMIT - YYCURSOR) < 11) YYFILL(11);
	yych = *YYCURSOR;
	switch(yych) {
	case 0x00:	goto yy7;
	case 0x0A:	goto yy5;
	case '*':	goto yy4;
	case '/':	goto yy2;
	default:	goto yy9;
	}
yy2:
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case '*':	goto yy12;
	default:	goto yy3;
	}
yy3:
#line 141 "scanner.re"
	{
					goto echo;
				}
#line 120 "<stdout>"
yy4:
	yych = *++YYCURSOR;
	switch(yych) {
	case '/':	goto yy10;
	default:	goto yy3;
	}
yy5:
	++YYCURSOR;
#line 130 "scanner.re"
	{
					out.write((const char*)(tok), (const char*)(cursor) - (const char*)(tok));
					tok = pos = cursor; cline++;
				  	goto echo;
				}
#line 135 "<stdout>"
yy7:
	++YYCURSOR;
#line 135 "scanner.re"
	{
					out.write((const char*)(tok), (const char*)(cursor) - (const char*)(tok) - 1); // -1 so we don't write out the \0
					if(cursor == eof) {
						RETURN(0);
					}
				}
#line 145 "<stdout>"
yy9:
	yych = *++YYCURSOR;
	goto yy3;
yy10:
	++YYCURSOR;
#line 121 "scanner.re"
	{
					if (ignore_eoc) {
						ignore_eoc = false;
					} else {
						out.write((const char*)(tok), (const char*)(cursor) - (const char*)(tok));
					}
					tok = pos = cursor;
					goto echo;
				}
#line 161 "<stdout>"
yy12:
	yych = *++YYCURSOR;
	switch(yych) {
	case '!':	goto yy14;
	default:	goto yy13;
	}
yy13:
	YYCURSOR = YYMARKER;
	goto yy3;
yy14:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'm':	goto yy15;
	case 'r':	goto yy16;
	default:	goto yy13;
	}
yy15:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'a':	goto yy21;
	default:	goto yy13;
	}
yy16:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'e':	goto yy17;
	default:	goto yy13;
	}
yy17:
	yych = *++YYCURSOR;
	switch(yych) {
	case '2':	goto yy18;
	default:	goto yy13;
	}
yy18:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'c':	goto yy19;
	default:	goto yy13;
	}
yy19:
	++YYCURSOR;
#line 110 "scanner.re"
	{ 
					out.write((const char*)(tok), (const char*)(&cursor[-7]) - (const char*)(tok));
					tok = cursor;
					RETURN(1);
				}
#line 210 "<stdout>"
yy21:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'x':	goto yy22;
	default:	goto yy13;
	}
yy22:
	yych = *++YYCURSOR;
	switch(yych) {
	case ':':	goto yy23;
	default:	goto yy13;
	}
yy23:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'r':	goto yy24;
	default:	goto yy13;
	}
yy24:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'e':	goto yy25;
	default:	goto yy13;
	}
yy25:
	yych = *++YYCURSOR;
	switch(yych) {
	case '2':	goto yy26;
	default:	goto yy13;
	}
yy26:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'c':	goto yy27;
	default:	goto yy13;
	}
yy27:
	++YYCURSOR;
#line 115 "scanner.re"
	{
					out << "#define YYMAXFILL " << maxFill << std::endl;
					tok = pos = cursor;
					ignore_eoc = true;
					goto echo;
				}
#line 256 "<stdout>"
}
#line 144 "scanner.re"

}


int Scanner::scan()
{
    char *cursor = cur;
    uint depth;

scan:
    tchar = cursor - pos;
    tline = cline;
    tok = cursor;
	if (iscfg == 1)
	{
		goto config;
	}
	else if (iscfg == 2)
	{
   		goto value;
    }

#line 281 "<stdout>"
{
	YYCTYPE yych;
	unsigned int yyaccept = 0;
	if((YYLIMIT - YYCURSOR) < 5) YYFILL(5);
	yych = *YYCURSOR;
	switch(yych) {
	case 0x09:
	case ' ':	goto yy50;
	case 0x0A:	goto yy52;
	case 0x0D:	goto yy54;
	case '"':	goto yy37;
	case '\'':	goto yy39;
	case '(':
	case ')':
	case ';':
	case '=':
	case '\\':
	case '|':	goto yy43;
	case '*':	goto yy35;
	case '+':
	case '?':	goto yy44;
	case '.':	goto yy48;
	case '/':	goto yy33;
	case 'A':
	case 'B':
	case 'C':
	case 'D':
	case 'E':
	case 'F':
	case 'G':
	case 'H':
	case 'I':
	case 'J':
	case 'K':
	case 'L':
	case 'M':
	case 'N':
	case 'O':
	case 'P':
	case 'Q':
	case 'R':
	case 'S':
	case 'T':
	case 'U':
	case 'V':
	case 'W':
	case 'X':
	case 'Y':
	case 'Z':
	case 'a':
	case 'b':
	case 'c':
	case 'd':
	case 'e':
	case 'f':
	case 'g':
	case 'h':
	case 'i':
	case 'j':
	case 'k':
	case 'l':
	case 'm':
	case 'n':
	case 'o':
	case 'p':
	case 'q':
	case 's':
	case 't':
	case 'u':
	case 'v':
	case 'w':
	case 'x':
	case 'y':
	case 'z':	goto yy47;
	case '[':	goto yy41;
	case 'r':	goto yy45;
	case '{':	goto yy31;
	default:	goto yy56;
	}
yy31:
	yyaccept = 0;
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case ',':	goto yy97;
	case '0':	goto yy94;
	case '1':
	case '2':
	case '3':
	case '4':
	case '5':
	case '6':
	case '7':
	case '8':
	case '9':	goto yy95;
	default:	goto yy32;
	}
yy32:
#line 166 "scanner.re"
	{ depth = 1;
				  goto code;
				}
#line 383 "<stdout>"
yy33:
	++YYCURSOR;
	switch((yych = *YYCURSOR)) {
	case '*':	goto yy92;
	default:	goto yy34;
	}
yy34:
#line 196 "scanner.re"
	{ RETURN(*tok); }
#line 393 "<stdout>"
yy35:
	++YYCURSOR;
	switch((yych = *YYCURSOR)) {
	case '/':	goto yy90;
	default:	goto yy36;
	}
yy36:
#line 198 "scanner.re"
	{ yylval.op = *tok;
				  RETURN(CLOSE); }
#line 404 "<stdout>"
yy37:
	yyaccept = 1;
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case 0x0A:	goto yy38;
	default:	goto yy86;
	}
yy38:
#line 183 "scanner.re"
	{ fatal("unterminated string constant (missing \")"); }
#line 415 "<stdout>"
yy39:
	yyaccept = 2;
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case 0x0A:	goto yy40;
	default:	goto yy81;
	}
yy40:
#line 184 "scanner.re"
	{ fatal("unterminated string constant (missing ')"); }
#line 426 "<stdout>"
yy41:
	yyaccept = 3;
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case 0x0A:	goto yy42;
	case '^':	goto yy72;
	default:	goto yy71;
	}
yy42:
#line 194 "scanner.re"
	{ fatal("unterminated range (missing ])"); }
#line 438 "<stdout>"
yy43:
	yych = *++YYCURSOR;
	goto yy34;
yy44:
	yych = *++YYCURSOR;
	goto yy36;
yy45:
	++YYCURSOR;
	switch((yych = *YYCURSOR)) {
	case 'e':	goto yy62;
	default:	goto yy61;
	}
yy46:
#line 225 "scanner.re"
	{ cur = cursor;
				  yylval.symbol = Symbol::find(token());
				  return ID; }
#line 456 "<stdout>"
yy47:
	yych = *++YYCURSOR;
	goto yy61;
yy48:
	++YYCURSOR;
#line 229 "scanner.re"
	{ cur = cursor;
				  yylval.regexp = mkDot();
				  return RANGE;
				}
#line 467 "<stdout>"
yy50:
	++YYCURSOR;
	yych = *YYCURSOR;
	goto yy59;
yy51:
#line 234 "scanner.re"
	{ goto scan; }
#line 475 "<stdout>"
yy52:
	++YYCURSOR;
yy53:
#line 236 "scanner.re"
	{ if(cursor == eof) RETURN(0);
				  pos = cursor; cline++;
				  goto scan;
	    			}
#line 484 "<stdout>"
yy54:
	++YYCURSOR;
	switch((yych = *YYCURSOR)) {
	case 0x0A:	goto yy57;
	default:	goto yy55;
	}
yy55:
#line 241 "scanner.re"
	{ std::ostringstream msg;
				  msg << "unexpected character: ";
				  prtChOrHex(msg, *tok);
				  fatal(msg.str().c_str());
				  goto scan;
				}
#line 499 "<stdout>"
yy56:
	yych = *++YYCURSOR;
	goto yy55;
yy57:
	yych = *++YYCURSOR;
	goto yy53;
yy58:
	++YYCURSOR;
	if(YYLIMIT == YYCURSOR) YYFILL(1);
	yych = *YYCURSOR;
yy59:
	switch(yych) {
	case 0x09:
	case ' ':	goto yy58;
	default:	goto yy51;
	}
yy60:
	++YYCURSOR;
	if(YYLIMIT == YYCURSOR) YYFILL(1);
	yych = *YYCURSOR;
yy61:
	switch(yych) {
	case '0':
	case '1':
	case '2':
	case '3':
	case '4':
	case '5':
	case '6':
	case '7':
	case '8':
	case '9':
	case 'A':
	case 'B':
	case 'C':
	case 'D':
	case 'E':
	case 'F':
	case 'G':
	case 'H':
	case 'I':
	case 'J':
	case 'K':
	case 'L':
	case 'M':
	case 'N':
	case 'O':
	case 'P':
	case 'Q':
	case 'R':
	case 'S':
	case 'T':
	case 'U':
	case 'V':
	case 'W':
	case 'X':
	case 'Y':
	case 'Z':
	case 'a':
	case 'b':
	case 'c':
	case 'd':
	case 'e':
	case 'f':
	case 'g':
	case 'h':
	case 'i':
	case 'j':
	case 'k':
	case 'l':
	case 'm':
	case 'n':
	case 'o':
	case 'p':
	case 'q':
	case 'r':
	case 's':
	case 't':
	case 'u':
	case 'v':
	case 'w':
	case 'x':
	case 'y':
	case 'z':	goto yy60;
	default:	goto yy46;
	}
yy62:
	yych = *++YYCURSOR;
	switch(yych) {
	case '2':	goto yy63;
	default:	goto yy61;
	}
yy63:
	yych = *++YYCURSOR;
	switch(yych) {
	case 'c':	goto yy64;
	default:	goto yy61;
	}
yy64:
	yyaccept = 4;
	yych = *(YYMARKER = ++YYCURSOR);
	switch(yych) {
	case ':':	goto yy65;
	default:	goto yy61;
	}
yy65:
	++YYCURSOR;
	if(YYLIMIT == YYCURSOR) YYFILL(1);
	yych = *YYCURSOR;
	switch(yych) {
	case 'A':
	case 'B':
	case 'C':
	case 'D':
	case 'E':
	case 'F':
	case 'G':
	case 'H':
	case 'I':
	case 'J':
	case 'K':
	case 'L':
	case 'M':
	case 'N':
	case 'O':
	case 'P':
	case 'Q':
	case 'R':
	case 'S':
	case 'T':
	case 'U':
	case 'V':
	case 'W':
	case 'X':
	case 'Y':
	case 'Z':
	case 'a':
	case 'b':
	case 'c':
	case 'd':
	case 'e':
	case 'f':
	case 'g':
	case 'h':
	case 'i':
	case 'j':
	case 'k':
	case 'l':
	case 'm':
	case 'n':
	case 'o':
	case 'p':
	case 'q':
	case 'r':
	case 's':
	case 't':
	case 'u':
	case 'v':
	case 'w':
	case 'x':
	case 'y':
	case 'z':	goto yy67;
	default:	goto yy66;
	}
yy66:
	YYCURSOR = YYMARKER;
	switch(yyaccept) {
	case 0: 	goto yy32;
	case 1: 	goto yy38;
	case 2: 	goto yy40;
	case 3: 	goto yy42;
	case 4: 	goto yy46;
	case 5: 	goto yy69;
	case 6: 	goto yy98;
	}
yy67:
	yyaccept = 5;
	YYMARKER = ++YYCURSOR;
	if(YYLIMIT == YYCURSOR) YYFILL(1);
	yych = *YYCURSOR;
	switch(yych) {
	case '0':
	case '1':
	case '2':
	case '3':
	case '4':
	case '5':
	case '6':
	case '7':
	case '8':
	case '9':
	case 'A':
	case 'B':
	case 'C':
	case 'D':
	case 'E':
	case 'F':
	case 'G':
	case 'H':
	case 'I':
	case 'J':
	case 'K':
	case 'L':
	case 'M':
	case 'N':
	case 'O':
	case 'P':
	case 'Q':
	case 'R':
	case 'S':
	case 'T':
	case 'U':
	case 'V':