home *** CD-ROM | disk | FTP | other *** search
- /*
- HEADER: CUG nnn.nn;
- TITLE: LEX - A Lexical Analyser Generator
- VERSION: 1.1 for IBM-PC
- DATE: Jan 30, 1985
- DESCRIPTION: A Lexical Analyser Generator. From UNIX
- KEYWORDS: Lexical Analyser Generator YACC C PREP
- SYSTEM: IBM-PC and Compatiables
- FILENAME: LEX.Y
- WARNINGS: This program is not for the casual user. It will
- be useful primarily to expert developers.
- This file must be processed by YACC with the -d
- option.
- CRC: N/A
- SEE-ALSO: YACC and PREP
- AUTHORS: Charles H. Forsyth
- Scott Guthery 11100 leafwood lane Austin, TX 78750
- Andrew M. Ward, Jr. Houston, Texas (Modifications)
- COMPILERS: LATTICE C
- REFERENCES: UNIX Systems Manuals -- Lex Manual on distribution disks
- */
- /*
- * Copyright (c) 1978 Charles H. Forsyth
- *
- *
- * Modified 22-Jun-86 Andrew Ward -- Modified code to compile under Lattice C
- * version 3.0h. Corrected several errors
- * from the assumption that pointers and
- * integers are the same size.
- * New debug code for LATTICE C using assert
- * to test for wild pointers. Using a proper
- * YACC this will produce a ytab.c file that
- * does not need hand modification.
- */
-
- /*
- * lex -- grammar/lexical analyser
- */
-
- %{
- #include <stdio.h>
- #include <assert.h>
- #include <stdlib.h>
- #include <string.h>
- #include "lexlex.h"
-
-
- char ccl[ (NCHARS+1) / NBPC ];
- int sz_ccl = (NCHARS+1) / NBPC;
- char *yysterm[] = {
- "error",
- "NAME",
- "CCLASS",
- "STRING",
- "CONCAT",
- 0 };
- extern char *progname;
- extern char *breakc;
- extern char *ignore;
- extern char *illeg;
- extern int nlook;
- char *lalloc();
- struct des {
- struct nfa *d_start;
- struct nfa *d_final;
- };
- struct nlist {
- struct nlist *nl_next;
- struct nfa *nl_base;
- struct nfa *nl_end;
- struct nfa *nl_start;
- struct nfa *nl_final;
- char *nl_name;
- }
- *nlist;
- extern int str_len;
- extern struct nfa *nfap;
- extern struct nfa *elem(int, char *);
- extern struct des *newdp(struct nfa *, struct nfa *);
- extern struct nlist *lookup(char *);
- extern int mapc(int);
- extern int cclass(void);
- extern void name(int);
- extern void action(void);
- extern void skipstr(int);
- extern void copycode(void);
- extern void string(int);
- extern void copynfa(struct nlist *, struct nfa *, struct des *);
- extern void spccl(char *, char *, struct des *, char **);
- extern void unget( int );
- extern void copy(char *, char *, int);
- extern void errmsg(char *);
- extern void yyerror(char *, char *);
- extern void newcase(int);
- extern void llactr(void);
- extern void setline(void);
- extern void cclprint(char *);
- extern int yyline;
- %}
- %union {
- char *buff;
- struct nlist *list;
- struct des *des_ptr;
- struct nfa *elem_ptr;
- }
- %term NAME CCLASS STRING CONCAT
- %token <buff> STRING NAME CCLASS
- %type <buff> NAME CCLASS
- %type <list> namedef name
- %type <des_ptr> pattern regexp
- %left ';'
- %left '='
- %left '/'
- %left '|'
- %left '(' NAME STRING CCLASS
- %left CONCAT
- %left '*'
- %%
- %{
- struct nfa *np, *nbase;
- char *cp;
- struct des *dp, *dp1;
- struct trans *tp;
- struct nlist *nl;
- int i, c;
- %}
- lexfile:
- auxiliary_section translation_section
- |
-
- ;
-
- auxiliary_section:
- auxiliaries '%' '%'
- | '%' '%'
- ;
-
- auxiliaries:
- auxiliaries auxiliary
- | auxiliary
- ;
-
- auxiliary:
- namedef '=' regexp ';' ={
- dp = $3;
- nl = $1;
- np = nl->nl_base;
- nl->nl_start = dp->d_start;
- nl->nl_final = dp->d_final;
- nl->nl_end = nfap;
- #ifdef DEBUG
- printf("NFA for %s\n", nl->nl_name);
- nfaprint(dp->d_start, nl->nl_base);
- #endif
- i = nl->nl_end - nl->nl_base;
- nbase = (struct nfa *)lalloc(i, sizeof(struct nfa), "nfa storage");
- copynfa(nl, nbase, dp);
- nl->nl_start = dp->d_start;
- nl->nl_final = dp->d_final;
- nl->nl_end = nbase+i;
- nl->nl_base = nbase;
- nfap = np;
- spccl(nl->nl_name, "ignore", dp, &ignore);
- spccl(nl->nl_name, "break", dp, &breakc);
- spccl(nl->nl_name, "illegal", dp, &illeg);
- }
- | '%' '{' ={
- copycode();
- }
- ;
-
- namedef:
- NAME ={
- $$ = lookup($1);
- $$->nl_base = nfap;
- if ($$->nl_start)
- yyerror("%s redefined", $$->nl_name);
- }
- ;
-
- name:
- NAME = {
- $$ = lookup($1);
- }
- ;
-
- regexp:
- CCLASS = {
- np = elem(CCL, $1);
- $$ = newdp(np, (np->n_succ[0] = elem(FIN, (char *)NULL ) ) );
- }
- | STRING ={
- cp = $1;
- if (str_len == 0) {
- np = elem(EPSILON, (char *)NULL );
- $$ = newdp(np, (np->n_succ[0] = elem(FIN, (char *)NULL ) ) );
- /* return(0);*/ /* AMW: the return here appears in error */
- }
- else /* AMW: else stmt added 1 May 1986 $$->d_start added */
- {
- $$->d_start = np = elem(*cp++, (char *)NULL);
- while(--str_len > 0)
- np = np->n_succ[0] = elem(*cp++,(char *)NULL);
- $$ = newdp($$->d_start, ( np->n_succ[0] = elem( FIN, (char *)NULL ) ) );
- }
- }
- | name ={
- if ((nl = $1)->nl_end == NULL)
- {
- yyerror("%s not defined", nl->nl_name);
- nl->nl_base = nl->nl_end = elem(FIN, (char *)NULL);
- nl->nl_start = nl->nl_final = nl->nl_base;
- }
- $$ = dp = (struct des *)lalloc(1, sizeof(struct des), "dfa input");
- nbase = nfap;
- i = nl->nl_end - nl->nl_base;
- if ((nfap += i) >= &nfa[MAXNFA]) {
- errmsg("Out of NFA nodes");
- exit(1);
- }
- copynfa(nl, nbase, dp);
- }
- | regexp '*' ={
- $$ = dp = $1;
- dp->d_start = newnfa(EPSILON, (np = dp->d_start), (struct nfa *)NULL);
- dp->d_final->n_char = EPSILON;
- dp->d_final->n_succ[0] = np;
- dp->d_final->n_succ[1] = np = elem(FIN, (char *)NULL);
- dp->d_start->n_succ[1] = np;
- dp->d_final = np;
- }
- | regexp '|' regexp ={
- $$ = dp = $1;
- dp->d_start = newnfa(EPSILON, dp->d_start, $3->d_start);
- dp->d_final->n_char = EPSILON;
- dp->d_final = dp->d_final->n_succ[0] = np = elem(FIN, (char *)NULL );
- dp = $3;
- dp->d_final->n_char = EPSILON;
- dp->d_final->n_succ[0] = np;
- #ifdef DEBUG
- assert( isdata( (char *)$3, sizeof( struct des) ) );
- #endif
- free((char *)$3);
- }
- | regexp regexp %prec CONCAT ={
- $$ = $1;
- dp = $2;
- np = $$->d_final;
- $$->d_final = dp->d_final;
- np->n_char = dp->d_start->n_char;
- np->n_ccl = dp->d_start->n_ccl;
- np->n_succ[0] = dp->d_start->n_succ[0];
- np->n_succ[1] = dp->d_start->n_succ[1];
- #ifdef DEBUG
- assert( isdata( (char *)$2, sizeof( struct nlist) ) );
- #endif
- free((char *)$2);
- }
- | '(' regexp ')' ={
- $$ = $2;
- }
- ;
-
- translation_section:
- translations ={
- ending();
- trans1:
-
- printf("\nNFA for complete syntax\n");
- printf("state 0\n");
- for (tp = trans; tp < transp; tp++)
- printf("\tepsilon\t%d\n", tp->t_start-nfa);
- for (tp = trans; tp < transp; tp++)
- nfaprint(tp->t_start, nfa);
- ;
- }
- | ={
- goto trans1;
- }
- ;
-
- translations:
- translations translation
- | llactr translation
- ;
-
- llactr:
- ={
- llactr();
- }
- ;
-
- translation:
- pattern action ={
- dp = $1;
- newtrans(dp->d_start, dp->d_final);
- }
- | '%' '{' ={
- copycode();
- }
- | '%' '%' ={
- ending();
- while ((c = get()) != EOF)
- putc(c, llout);
- }
- ;
-
- action:
- '{' ={
- action();
- }
- ;
-
- pattern:
- regexp '/' regexp ={
- if (nlook >= NBPW)
- yyerror("More than %d translations with lookahead",NBPW);
- $$ = dp = $1;
- np = dp->d_final;
- np->n_char = EPSILON;
- np->n_flag |= LOOK;
- np->n_succ[0] = $3->d_start;
- dp->d_final = $3->d_final;
- np->n_look = nlook;
- dp->d_final->n_look = nlook++;
- dp->d_final->n_flag |= FLOOK;
- #ifdef DEBUG
- assert( isdata( (char *)$3, sizeof( struct des) ) );
- #endif
- free((char *)$3);
- }
- | regexp
- ;
-
- %%
- /*
- * Lexical analyser
- * (it isn't done with lex...)
- */
- char buffer[150];
-
- int yylex()
- {
-
- int c;
- /* char *cp; */
- int lno;
-
- if (yyline == 0)
- yyline++;
- loop:
- c = get();
- if (isupper(c)) {
- name(c);
- yylval.buff = strlwr( yylval.buff );
- return(STRING);
- }
- else if (islower(c) || c == '_') {
- name(c);
- return(NAME);
- }
- switch (c) {
- case EOF:
- return(0);
-
- case '[':
- return(cclass());
-
- case '(':
- case ')':
- case '{':
- case '}':
- case '*':
- case '|':
- case '=':
- case ';':
- case '%':
- return(c);
-
- case '/':
- if ((c = get()) != '*') {
- unget(c);
- return('/');
- }
- lno = yyline;
- for (; c != EOF; c = get())
- if (c == '*')
- if ((c = get()) == '/')
- goto loop;
- else
- unget(c);
- yyline = lno;
- errmsg("End of file in comment");
-
- case '\'':
- case '"':
- yylval.buff = buffer;
- string(c);
- return(STRING);
-
- case '\n':
- case '\r':
- case ' ':
- case '\t':
- goto loop;
-
- default:
-
- yylval.buff = buffer;
- /* if (c == '\\') {
- unget(c);
- c = mapch(EOF);
- }
- */
- buffer[0] = c;
- buffer[1] = '\0';
- str_len = 1;
- return(STRING);
- }
- }
-
-
- int cclass()
- {
- int c, i, lc;
- int compl;
-
- compl = 0;
- /* Zero the ccl array */
- for (i = 0; i < sz_ccl; i++) ccl[i] = '\0';
- /* Check if exclusion definition */
- if ((c = get()) == '^') compl++;
- else unget(c);
-
- lc = -1;
- while( ( c = mapc(']') ) != EOF)
- {
- if (c == '-' && lc >= 0)
- {
- if((c = mapc(']')) == EOF) break;
- /* Map 'c' into bit pattern */
- for(i = lc; i <= c; i++) ccl[ i / NBPC ] |= 1 << (i % NBPC);
- lc = -1;
- continue;
- }
-
- ccl[ c / NBPC ] |= 1 << ( c % NBPC );
- lc = c;
- }
-
- if (compl) {
- for (i = 0; i < sz_ccl; i++)
- ccl[i] ^= -1;
- if (aflag == 0)
- for (i = 0200; i < ( 1 << NBPC ); i++)
- ccl[i/NBPC] &= ~(1 << (i%NBPC));
- }
- yylval.buff = newccl( ccl );
- return(CCLASS);
- }
-
- void string(ec)
- int ec;
- {
- char *cp;
- int c;
- extern int str_len;
- for(cp = buffer; (c = mapc(ec)) != EOF;)
- *cp++ = c;
- *cp = '\0';
- str_len = strlen( buffer );
- }
-
- int mapc(ec)
- int ec;
- {
-
- int c, v, i;
-
- if((c = get()) == ec) return(EOF);
- switch(c) {
- case EOF:
- errmsg("End of file in string");
- return(c);
-
- case '\\':
- if ((c = get()) >= '0' && c <= '7') {
- i = 0;
- for (v = 0; c>='0' && c<='7' && i++<3; c = get())
- v = v*010 + c-'0';
- unget(c);
- return(v&CMASK);
- }
- switch (c) {
- case 'n':
- return('\n');
-
- case 't':
- return('\t');
-
- case 'b':
- return('\b');
-
- case 'r':
- return('\r');
-
- case '\n':
- yyline++;
- return(mapc(ec));
- }
-
- default:
- return(c);
- }
- }
-
- void name(c)
- int c;
- {
- char *cp;
-
- for (yylval.buff=cp=buffer; isalpha(c) || isdigit(c) || c=='_'; c=get())
- *cp++ = c;
- *cp = '\0';
- str_len = strlen( buffer );
- unget(c);
- }
-
- /*
- * Miscellaneous functions
- * used only by lex.y
- */
- struct nfa *elem(k, v)
- int k;
- char *v;
- {
- struct nfa *fp;
- fp = newnfa(k, (struct nfa *)NULL, (struct nfa *)NULL);
-
- if (k == CCL)
- fp->n_ccl = v;
- return(fp);
- }
-
- struct des *newdp(st, fi)
- struct nfa *st, *fi;
- {
- struct des *dp;
-
- dp = (struct des *)lalloc(1, sizeof(struct des), "dfa input");
- #ifdef DEBUG
- if( st != NULL) assert( isdata( (char *)st, sizeof( struct nfa ) ) );
- if( fi != NULL) assert( isdata( (char *)fi, sizeof( struct nfa ) ) );
- #endif
- dp->d_start = st;
- dp->d_final = fi;
- return(dp);
- }
-
- void action()
- {
- int c;
- int lno, lev;
-
- newcase(transp-trans);
- lno = yyline;
- lev = 0;
- for (; (c = get()) != EOF && (c != '}' || lev); putc(c, llout))
- if (c == '{')
- lev++;
- else if (c == '}')
- lev--;
- else if (c == '\'' || c == '"') {
- putc(c, llout);
- skipstr(c);
- }
- fprintf(llout, "\n\tbreak;\n");
- if (c == EOF) {
- yyline = lno;
- errmsg("End of file in action");
- }
- }
-
- void skipstr(ec)
- int ec;
- {
- int c;
-
- while ((c = get()) != ec && c != EOF) {
- putc(c, llout);
- if (c == '\\' && (c = get()) != EOF)
- putc(c, llout);
- }
- }
-
-
- void copycode()
- {
- int lno;
- int c;
-
- setline();
- lno = yyline;
- for (; (c = get()) != EOF; putc(c, llout))
- if (c == '%')
- {
- if ((c = get()) == '}')
- return;
- unget(c);
- c = '%';
- }
- yyline = lno;
- errmsg("Incomplete %{ declaration");
- exit(1);
- }
-
- struct nlist *lookup(s)
- char *s;
- {
- struct nlist *nl;
- char *cp;
-
- for (nl = nlist; nl; nl = nl->nl_next)
- if (!strcmp(s, nl->nl_name))
- return(nl);
- nl = (struct nlist *)lalloc(1, sizeof(struct nlist), "namelist");
- nl->nl_start = nl->nl_end = nl->nl_base = nl->nl_final = (struct nfa *)NULL;
- nl->nl_next = nlist;
- nlist = nl;
- nl->nl_name = cp = strdup( s );
- return(nl);
- }
-
- void copynfa(nl, nbase, dp)
- struct nlist *nl;
- struct des *dp;
- struct nfa *nbase;
- {
- struct nfa *np, *ob;
- int j;
- int i;
-
- ob = nl->nl_base;
- i = nl->nl_end - ob;
-
- #ifdef DEBUG
- fprintf(stdout, "\nCOPYNFA: index i = %d",i);
- #endif
- np = nbase;
- copy((char *)np, (char *)ob, sizeof(struct nfa)*i);
- for (np = nbase; i-- > 0; np++) {
- np->n_flag &= ~NPRT;
- for (j = 0; j < 2; j++)
- if (np->n_succ[j])
- np->n_succ[j] = nbase + (np->n_succ[j] - ob);
- }
- dp->d_start = nbase + (int)(nl->nl_start-ob);
- dp->d_final = nbase + (int)(nl->nl_final-ob);
- }
-
- void spccl(nm, isit, dp, where)
- char *nm, *isit;
- struct des *dp;
- char **where;
- {
- if (!strcmp(nm, isit))
- {
- if(*where != 0 )
- yyerror("redefinition of %s class", isit);
- if (dp->d_start->n_char == CCL &&
- dp->d_start->n_succ[0] == dp->d_final)
- *where = dp->d_start->n_ccl;
- else
- yyerror("Illegal %s class", isit);
- }
- }
-
- int get()
- {
- int c;
-
- if ((c = getc(lexin)) == '\n')
- yyline++;
- return(c);
- }
-
- void unget(c)
- int c;
- {
- if (c == '\n')
- yyline--;
- ungetc(c, lexin);
- }
-
- void errmsg(s)
- char *s;
- {
- if (yyline)
- fprintf(stderr, "%d: ", yyline);
- fprintf(stderr, "%r", &s);
- if (yychar > 256)
- fprintf(stderr, " near `%s'", yysterm[yychar-256]);
- else if (yychar < 256 && yychar > 0)
- fprintf(stderr, " near `%c'", yychar);
- fprintf(stderr, "\n");
- }
-
- void copy(out, in, count)
- char *out;
- char *in;
- int count;
- {
- while (--count >= 0)
- *out++ = *in++;
- }
-
- void yyerror(format, value)
- char *format, *value;
- {
- if (yyline)
- fprintf(stderr, "%d: ", yyline);
- fprintf(stderr, format, value);
- if (yychar > 256)
- fprintf(stderr, " near '%s'", yysterm[yychar-256]);
- else if (yychar < 256 && yychar > 0)
- fprintf(stderr, " near '%c'", yychar);
- fprintf(stderr, "\n");
- }
- int nterms = 15;
- int nnonter = 13;
- int nstate = 41;
-
- char *yysnter[14] = {
- "$accept",
- "lexfile",
- "auxiliary_section",
- "translation_section",
- "auxiliaries",
- "auxiliary",
- "namedef",
- "regexp",
- "name",
- "translations",
- "translation",
- "llactr",
- "pattern",
- "action" };