a2py.c

早期freebsd实现

/* $RCSfile: a2py.c,v $$Revision: 4.0.1.2 $$Date: 92/06/08 16:15:16 $
 *
 *    Copyright (c) 1991, Larry Wall
 *
 *    You may distribute under the terms of either the GNU General Public
 *    License or the Artistic License, as specified in the README file.
 *
 * $Log:	a2py.c,v $
 * Revision 4.0.1.2  92/06/08  16:15:16  lwall
 * patch20: in a2p, now warns about spurious backslashes
 * patch20: in a2p, now allows [ to be backslashed in pattern
 * patch20: in a2p, now allows numbers of the form 2.
 * 
 * Revision 4.0.1.1  91/06/07  12:12:59  lwall
 * patch4: new copyright notice
 * 
 * Revision 4.0  91/03/20  01:57:26  lwall
 * 4.0 baseline.
 * 
 */

#ifdef OS2
#include "../patchlevel.h"
#endif
#include "util.h"
char *index();

char *filename;
char *myname;

int checkers = 0;
STR *walk();

#ifdef OS2
usage()
{
    printf("\nThis is the AWK to PERL translator, version 4.0, patchlevel %d\n", PATCHLEVEL);
    printf("\nUsage: %s [-D<number>] [-F<char>] [-n<fieldlist>] [-<number>] filename\n", myname);
    printf("\n  -D<number>      sets debugging flags."
           "\n  -F<character>   the awk script to translate is always invoked with"
           "\n                  this -F switch."
           "\n  -n<fieldlist>   specifies the names of the input fields if input does"
           "\n                  not have to be split into an array."
           "\n  -<number>       causes a2p to assume that input will always have that"
           "\n                  many fields.\n");
    exit(1);
}
#endif
main(argc,argv,env)
register int argc;
register char **argv;
register char **env;
{
    register STR *str;
    register char *s;
    int i;
    STR *tmpstr;

    myname = argv[0];
    linestr = str_new(80);
    str = str_new(0);		/* first used for -I flags */
    for (argc--,argv++; argc; argc--,argv++) {
	if (argv[0][0] != '-' || !argv[0][1])
	    break;
      reswitch:
	switch (argv[0][1]) {
#ifdef DEBUGGING
	case 'D':
	    debug = atoi(argv[0]+2);
#ifdef YYDEBUG
	    yydebug = (debug & 1);
#endif
	    break;
#endif
	case '0': case '1': case '2': case '3': case '4':
	case '5': case '6': case '7': case '8': case '9':
	    maxfld = atoi(argv[0]+1);
	    absmaxfld = TRUE;
	    break;
	case 'F':
	    fswitch = argv[0][2];
	    break;
	case 'n':
	    namelist = savestr(argv[0]+2);
	    break;
	case '-':
	    argc--,argv++;
	    goto switch_end;
	case 0:
	    break;
	default:
	    fatal("Unrecognized switch: %s\n",argv[0]);
#ifdef OS2
            usage();
#endif
	}
    }
  switch_end:

    /* open script */

    if (argv[0] == Nullch) {
#ifdef OS2
	if ( isatty(fileno(stdin)) )
	    usage();
#endif
        argv[0] = "-";
    }
    filename = savestr(argv[0]);

    filename = savestr(argv[0]);
    if (strEQ(filename,"-"))
	argv[0] = "";
    if (!*argv[0])
	rsfp = stdin;
    else
	rsfp = fopen(argv[0],"r");
    if (rsfp == Nullfp)
	fatal("Awk script \"%s\" doesn't seem to exist.\n",filename);

    /* init tokener */

    bufptr = str_get(linestr);
    symtab = hnew();
    curarghash = hnew();

    /* now parse the report spec */

    if (yyparse())
	fatal("Translation aborted due to syntax errors.\n");

#ifdef DEBUGGING
    if (debug & 2) {
	int type, len;

	for (i=1; i<mop;) {
	    type = ops[i].ival;
	    len = type >> 8;
	    type &= 255;
	    printf("%d\t%d\t%d\t%-10s",i++,type,len,opname[type]);
	    if (type == OSTRING)
		printf("\t\"%s\"\n",ops[i].cval),i++;
	    else {
		while (len--) {
		    printf("\t%d",ops[i].ival),i++;
		}
		putchar('\n');
	    }
	}
    }
    if (debug & 8)
	dump(root);
#endif

    /* first pass to look for numeric variables */

    prewalk(0,0,root,&i);

    /* second pass to produce new program */

    tmpstr = walk(0,0,root,&i,P_MIN);
    str = str_make("#!");
    str_cat(str, BIN);
    str_cat(str, "/perl\neval \"exec ");
    str_cat(str, BIN);
    str_cat(str, "/perl -S $0 $*\"\n\
    if $running_under_some_shell;\n\
			# this emulates #! processing on NIH machines.\n\
			# (remove #! line above if indigestible)\n\n");
    str_cat(str,
      "eval '$'.$1.'$2;' while $ARGV[0] =~ /^([A-Za-z_]+=)(.*)/ && shift;\n");
    str_cat(str,
      "			# process any FOO=bar switches\n\n");
    if (do_opens && opens) {
	str_scat(str,opens);
	str_free(opens);
	str_cat(str,"\n");
    }
    str_scat(str,tmpstr);
    str_free(tmpstr);
#ifdef DEBUGGING
    if (!(debug & 16))
#endif
    fixup(str);
    putlines(str);
    if (checkers) {
	fprintf(stderr,
	  "Please check my work on the %d line%s I've marked with \"#???\".\n",
		checkers, checkers == 1 ? "" : "s" );
	fprintf(stderr,
	  "The operation I've selected may be wrong for the operand types.\n");
    }
    exit(0);
}

#define RETURN(retval) return (bufptr = s,retval)
#define XTERM(retval) return (expectterm = TRUE,bufptr = s,retval)
#define XOP(retval) return (expectterm = FALSE,bufptr = s,retval)
#define ID(x) return (yylval=string(x,0),expectterm = FALSE,bufptr = s,idtype)

int idtype;

yylex()
{
    register char *s = bufptr;
    register char *d;
    register int tmp;

  retry:
#ifdef YYDEBUG
    if (yydebug)
	if (index(s,'\n'))
	    fprintf(stderr,"Tokener at %s",s);
	else
	    fprintf(stderr,"Tokener at %s\n",s);
#endif
    switch (*s) {
    default:
	fprintf(stderr,
	    "Unrecognized character %c in file %s line %d--ignoring.\n",
	     *s++,filename,line);
	goto retry;
    case '\\':
	s++;
	if (*s && *s != '\n') {
	    yyerror("Ignoring spurious backslash");
	    goto retry;
	}
	/*FALLSTHROUGH*/
    case 0:
	s = str_get(linestr);
	*s = '\0';
	if (!rsfp)
	    RETURN(0);
	line++;
	if ((s = str_gets(linestr, rsfp)) == Nullch) {
	    if (rsfp != stdin)
		fclose(rsfp);
	    rsfp = Nullfp;
	    s = str_get(linestr);
	    RETURN(0);
	}
	goto retry;
    case ' ': case '\t':
	s++;
	goto retry;
    case '\n':
	*s = '\0';
	XTERM(NEWLINE);
    case '#':
	yylval = string(s,0);
	*s = '\0';
	XTERM(COMMENT);
    case ';':
	tmp = *s++;
	if (*s == '\n') {
	    s++;
	    XTERM(SEMINEW);
	}
	XTERM(tmp);
    case '(':
	tmp = *s++;
	XTERM(tmp);
    case '{':
    case '[':
    case ')':
    case ']':
    case '?':
    case ':':
	tmp = *s++;
	XOP(tmp);
    case 127:
	s++;
	XTERM('}');
    case '}':
	for (d = s + 1; isspace(*d); d++) ;
	if (!*d)
	    s = d - 1;
	*s = 127;
	XTERM(';');
    case ',':
	tmp = *s++;
	XTERM(tmp);
    case '~':
	s++;
	yylval = string("~",1);
	XTERM(MATCHOP);
    case '+':
    case '-':
	if (s[1] == *s) {
	    s++;
	    if (*s++ == '+')
		XTERM(INCR);
	    else
		XTERM(DECR);
	}
	/* FALL THROUGH */
    case '*':
    case '%':
    case '^':
	tmp = *s++;
	if (*s == '=') {
	    if (tmp == '^')
		yylval = string("**=",3);
	    else
		yylval = string(s-1,2);
	    s++;
	    XTERM(ASGNOP);
	}
	XTERM(tmp);
    case '&':
	s++;
	tmp = *s++;
	if (tmp == '&')
	    XTERM(ANDAND);
	s--;
	XTERM('&');
    case '|':
	s++;
	tmp = *s++;
	if (tmp == '|')
	    XTERM(OROR);
	s--;
	while (*s == ' ' || *s == '\t')
	    s++;
	if (strnEQ(s,"getline",7))
	    XTERM('p');
	else
	    XTERM('|');
    case '=':
	s++;
	tmp = *s++;
	if (tmp == '=') {
	    yylval = string("==",2);
	    XTERM(RELOP);
	}
	s--;
	yylval = string("=",1);
	XTERM(ASGNOP);
    case '!':
	s++;
	tmp = *s++;
	if (tmp == '=') {
	    yylval = string("!=",2);
	    XTERM(RELOP);
	}
	if (tmp == '~') {
	    yylval = string("!~",2);
	    XTERM(MATCHOP);
	}
	s--;
	XTERM(NOT);
    case '<':
	s++;
	tmp = *s++;
	if (tmp == '=') {
	    yylval = string("<=",2);
	    XTERM(RELOP);
	}
	s--;
	XTERM('<');
    case '>':
	s++;
	tmp = *s++;
	if (tmp == '>') {
	    yylval = string(">>",2);
	    XTERM(GRGR);
	}
	if (tmp == '=') {
	    yylval = string(">=",2);
	    XTERM(RELOP);
	}
	s--;
	XTERM('>');

#define SNARFWORD \
	d = tokenbuf; \
	while (isalpha(*s) || isdigit(*s) || *s == '_') \
	    *d++ = *s++; \
	*d = '\0'; \
	d = tokenbuf; \
	if (*s == '(') \
	    idtype = USERFUN; \
	else \
	    idtype = VAR;

    case '$':
	s++;
	if (*s == '0') {
	    s++;
	    do_chop = TRUE;
	    need_entire = TRUE;
	    idtype = VAR;
	    ID("0");
	}
	do_split = TRUE;
	if (isdigit(*s)) {
	    for (d = s; isdigit(*s); s++) ;
	    yylval = string(d,s-d);
	    tmp = atoi(d);
	    if (tmp > maxfld)
		maxfld = tmp;
	    XOP(FIELD);
	}
	split_to_array = set_array_base = TRUE;
	XOP(VFIELD);

    case '/':			/* may either be division or pattern */
	if (expectterm) {
	    s = scanpat(s);
	    XTERM(REGEX);
	}
	tmp = *s++;
	if (*s == '=') {
	    yylval = string("/=",2);
	    s++;
	    XTERM(ASGNOP);
	}
	XTERM(tmp);

    case '0': case '1': case '2': case '3': case '4':
    case '5': case '6': case '7': case '8': case '9': case '.':
	s = scannum(s);
	XOP(NUMBER);
    case '"':
	s++;
	s = cpy2(tokenbuf,s,s[-1]);
	if (!*s)
	    fatal("String not terminated:\n%s",str_get(linestr));
	s++;
	yylval = string(tokenbuf,0);
	XOP(STRING);

    case 'a': case 'A':
	SNARFWORD;
	if (strEQ(d,"ARGC"))
	    set_array_base = TRUE;
	if (strEQ(d,"ARGV")) {
	    yylval=numary(string("ARGV",0));
	    XOP(VAR);
	}
	if (strEQ(d,"atan2")) {
	    yylval = OATAN2;
	    XTERM(FUNN);
	}
	ID(d);
    case 'b': case 'B':
	SNARFWORD;
	if (strEQ(d,"break"))
	    XTERM(BREAK);
	if (strEQ(d,"BEGIN"))
	    XTERM(BEGIN);
	ID(d);
    case 'c': case 'C':
	SNARFWORD;
	if (strEQ(d,"continue"))
	    XTERM(CONTINUE);
	if (strEQ(d,"cos")) {
	    yylval = OCOS;
	    XTERM(FUN1);
	}
	if (strEQ(d,"close")) {
	    do_fancy_opens = 1;
	    yylval = OCLOSE;
	    XTERM(FUN1);
	}
	if (strEQ(d,"chdir"))
	    *d = toupper(*d);
	else if (strEQ(d,"crypt"))
	    *d = toupper(*d);
	else if (strEQ(d,"chop"))
	    *d = toupper(*d);
	else if (strEQ(d,"chmod"))
	    *d = toupper(*d);
	else if (strEQ(d,"chown"))
	    *d = toupper(*d);
	ID(d);
    case 'd': case 'D':
	SNARFWORD;
	if (strEQ(d,"do"))
	    XTERM(DO);
	if (strEQ(d,"delete"))
	    XTERM(DELETE);
	if (strEQ(d,"die"))
	    *d = toupper(*d);
	ID(d);
    case 'e': case 'E':
	SNARFWORD;
	if (strEQ(d,"END"))
	    XTERM(END);
	if (strEQ(d,"else"))
	    XTERM(ELSE);
	if (strEQ(d,"exit")) {
	    saw_line_op = TRUE;
	    XTERM(EXIT);
	}
	if (strEQ(d,"exp")) {
	    yylval = OEXP;
	    XTERM(FUN1);
	}
	if (strEQ(d,"elsif"))
	    *d = toupper(*d);
	else if (strEQ(d,"eq"))
	    *d = toupper(*d);
	else if (strEQ(d,"eval"))
	    *d = toupper(*d);
	else if (strEQ(d,"eof"))
	    *d = toupper(*d);
	else if (strEQ(d,"each"))
	    *d = toupper(*d);
	else if (strEQ(d,"exec"))
	    *d = toupper(*d);
	ID(d);
    case 'f': case 'F':
	SNARFWORD;
	if (strEQ(d,"FS")) {
	    saw_FS++;
	    if (saw_FS == 1 && in_begin) {
		for (d = s; *d && isspace(*d); d++) ;
		if (*d == '=') {
		    for (d++; *d && isspace(*d); d++) ;
		    if (*d == '"' && d[2] == '"')
			const_FS = d[1];
		}
	    }
	    ID(tokenbuf);
	}
	if (strEQ(d,"for"))
	    XTERM(FOR);
	else if (strEQ(d,"function"))
	    XTERM(FUNCTION);
	if (strEQ(d,"FILENAME"))
	    d = "ARGV";
	if (strEQ(d,"foreach"))
	    *d = toupper(*d);
	else if (strEQ(d,"format"))
	    *d = toupper(*d);
	else if (strEQ(d,"fork"))
	    *d = toupper(*d);
	else if (strEQ(d,"fh"))
	    *d = toupper(*d);
	ID(d);
    case 'g': case 'G':
	SNARFWORD;
	if (strEQ(d,"getline"))
	    XTERM(GETLINE);
	if (strEQ(d,"gsub"))
	    XTERM(GSUB);
	if (strEQ(d,"ge"))
	    *d = toupper(*d);
	else if (strEQ(d,"gt"))
	    *d = toupper(*d);
	else if (strEQ(d,"goto"))
	    *d = toupper(*d);
	else if (strEQ(d,"gmtime"))
	    *d = toupper(*d);
	ID(d);
    case 'h': case 'H':
	SNARFWORD;
	if (strEQ(d,"hex"))
	    *d = toupper(*d);
	ID(d);
    case 'i': case 'I':
	SNARFWORD;
	if (strEQ(d,"if"))
	    XTERM(IF);
	if (strEQ(d,"in"))
	    XTERM(IN);
	if (strEQ(d,"index")) {
	    set_array_base = TRUE;
	    XTERM(INDEX);
	}
	if (strEQ(d,"int")) {
	    yylval = OINT;
	    XTERM(FUN1);
	}
	ID(d);
    case 'j': case 'J':
	SNARFWORD;
	if (strEQ(d,"join"))
	    *d = toupper(*d);
	ID(d);
    case 'k': case 'K':
	SNARFWORD;
	if (strEQ(d,"keys"))
	    *d = toupper(*d);
	else if (strEQ(d,"kill"))
	    *d = toupper(*d);
	ID(d);
    case 'l': case 'L':
	SNARFWORD;
	if (strEQ(d,"length")) {
	    yylval = OLENGTH;
	    XTERM(FUN1);
	}
	if (strEQ(d,"log")) {
	    yylval = OLOG;
	    XTERM(FUN1);
	}
	if (strEQ(d,"last"))
	    *d = toupper(*d);
	else if (strEQ(d,"local"))
	    *d = toupper(*d);
	else if (strEQ(d,"lt"))
	    *d = toupper(*d);
	else if (strEQ(d,"le"))
	    *d = toupper(*d);
	else if (strEQ(d,"locatime"))
	    *d = toupper(*d);
	else if (strEQ(d,"link"))
	    *d = toupper(*d);
	ID(d);
    case 'm': case 'M':
	SNARFWORD;
	if (strEQ(d,"match")) {
	    set_array_base = TRUE;
	    XTERM(MATCH);
	}
	if (strEQ(d,"m"))
	    *d = toupper(*d);
	ID(d);
    case 'n': case 'N':
	SNARFWORD;
	if (strEQ(d,"NF"))
	    do_chop = do_split = split_to_array = set_array_base = TRUE;
	if (strEQ(d,"next")) {
	    saw_line_op = TRUE;
	    XTERM(NEXT);
	}
	if (strEQ(d,"ne"))
	    *d = toupper(*d);
	ID(d);
    case 'o': case 'O':
	SNARFWORD;
	if (strEQ(d,"ORS")) {
	    saw_ORS = TRUE;
	    d = "\\";
	}
	if (strEQ(d,"OFS")) {
	    saw_OFS = TRUE;
	    d = ",";
	}
	if (strEQ(d,"OFMT")) {
	    d = "#";
	}
	if (strEQ(d,"open"))
	    *d = toupper(*d);
	else if (strEQ(d,"ord"))
	    *d = toupper(*d);
	else if (strEQ(d,"oct"))