1 /*
2  * misc.c
3  *
4  * Manage tokens, regular expressions.
5  * Print methods for debugging
6  * Compute follow lists onto tail ends of rules.
7  *
8  * The following functions are visible:
9  *
10  *		int		addTname(char *);		Add token name
11  *		int		addTexpr(char *);		Add token expression
12  *		int		Tnum(char *);			Get number of expr/token
13  *		void	Tklink(char *, char *);	Link a name with an expression
14  *		int		hasAction(expr);		Does expr already have action assigned?
15  *		void	setHasAction(expr);		Indicate that expr now has an action
16  *		Entry	*newEntry(char *,int);	Create new table entry with certain size
17  *		void	list_add(ListNode **list, char *e)
18  *      void    list_free(ListNode **list, int freeData);   *** MR10 ***
19  *		void	list_apply(ListNode *list, void (*f)())
20  *		void	lexclass(char *m);		switch to new/old lexical class
21  *		void	lexmode(int i);			switch to old lexical class i
22  *
23  * SOFTWARE RIGHTS
24  *
25  * We reserve no LEGAL rights to the Purdue Compiler Construction Tool
26  * Set (PCCTS) -- PCCTS is in the public domain.  An individual or
27  * company may do whatever they wish with source code distributed with
28  * PCCTS or the code generated by PCCTS, including the incorporation of
29  * PCCTS, or its output, into commerical software.
30  *
31  * We encourage users to develop software with PCCTS.  However, we do ask
32  * that credit is given to us for developing PCCTS.  By "credit",
33  * we mean that if you incorporate our source code into one of your
34  * programs (commercial product, research project, or otherwise) that you
35  * acknowledge this fact somewhere in the documentation, research report,
36  * etc...  If you like PCCTS and have developed a nice tool with the
37  * output, please mention that you developed it using PCCTS.  In
38  * addition, we ask that this header remain intact in our source code.
39  * As long as these guidelines are kept, we expect to continue enhancing
40  * this system and expect to make other tools available as they are
41  * completed.
42  *
43  * ANTLR 1.33
44  * Terence Parr
45  * Parr Research Corporation
46  * with Purdue University and AHPCRC, University of Minnesota
47  * 1989-2001
48  */
49 
50 #include <stdio.h>
51 #include "pcctscfg.h"
52 #include "set.h"
53 #include "syn.h"
54 #include "hash.h"
55 #include "generic.h"
56 #include "dlgdef.h"
57 #include <ctype.h>
58 
59 static int tsize=TSChunk;		/* size of token str arrays */
60 
61 static void
62 #ifdef __USE_PROTOS
63 RemapForcedTokensInSyntaxDiagram(Node *);
64 #else
65 RemapForcedTokensInSyntaxDiagram();
66 #endif
67 
68 				/* T o k e n  M a n i p u l a t i o n */
69 
70 /*
71  * add token 't' to the TokenStr/Expr array.  Make more room if necessary.
72  * 't' is either an expression or a token name.
73  *
74  * There is only one TokenStr array, but multiple ExprStr's.  Therefore,
75  * for each lex class (element of lclass) we must extend the ExprStr array.
76  * ExprStr's and TokenStr are always all the same size.
77  *
78  * Also, there is a Texpr hash table for each automaton.
79  */
80 static void
81 #ifdef __USE_PROTOS
Ttrack(char * t)82 Ttrack( char *t )
83 #else
84 Ttrack( t )
85 char *t;
86 #endif
87 {
88 	if ( TokenNum >= tsize )	/* terminal table overflow? */
89 	{
90 		char **p;
91 		int i, more, j;
92 
93 		more = TSChunk * (1 + ((TokenNum-tsize) / TSChunk));
94 		tsize += more;
95 		TokenStr = (char **) realloc((char *)TokenStr, tsize*sizeof(char *));
96 		require(TokenStr != NULL, "Ttrack: can't extend TokenStr");
97 		for (i=0; i<NumLexClasses; i++)
98 		{
99 			lclass[i].exprs = (char **)
100 							  realloc((char *)lclass[i].exprs, tsize*sizeof(char *));
101 			require(lclass[i].exprs != NULL, "Ttrack: can't extend ExprStr");
102 			for (p= &lclass[i].exprs[tsize-more],j=1; j<=more; j++) *p++ = NULL;
103 		}
104 		for (p= &TokenStr[tsize-more],i=1; i<=more; i++) *p++ = NULL;
105 		lexmode( CurrentLexClass ); /* reset ExprStr in case table moved */
106 	}
107 	/* note: we use the actual ExprStr/TokenStr array
108 	 * here as TokenInd doesn't exist yet
109 	 */
110 	if ( *t == '"' ) ExprStr[TokenNum] = t;
111 	else TokenStr[TokenNum] = t;
112 }
113 
114 static Expr *
115 #ifdef __USE_PROTOS
newExpr(char * e)116 newExpr( char *e )
117 #else
118 newExpr( e )
119 char *e;
120 #endif
121 {
122 	Expr *p = (Expr *) calloc(1, sizeof(Expr));
123 	require(p!=NULL, "newExpr: cannot alloc Expr node");
124 
125 	p->expr = e;
126 	p->lclass = CurrentLexClass;
127 	return p;
128 }
129 
130 /* switch to lexical class/mode m.  This amounts to creating a new
131  * lex mode if one does not already exist and making ExprStr point
132  * to the correct char string array.  We must also switch Texpr tables.
133  *
134  * BTW, we need multiple ExprStr arrays because more than one automaton
135  * may have the same label for a token, but with different expressions.
136  * We need to track an expr for each automaton.  If we disallowed this
137  * feature, only one ExprStr would be required.
138  */
139 void
140 #ifdef __USE_PROTOS
lexclass(char * m)141 lexclass( char *m )
142 #else
143 lexclass( m )
144 char *m;
145 #endif
146 {
147 	int i;
148 	TermEntry *p;
149 	static char EOFSTR[] = "\"@\"";
150 
151 	if ( hash_get(Tname, m) != NULL )
152 	{
153 		warn(eMsg1("lexclass name conflicts with token/errclass label '%s'",m));
154 	}
155 	/* does m already exist? */
156 	i = LexClassIndex(m);
157 	if ( i != -1 ) {lexmode(i); return;}
158 	/* must make new one */
159 	NumLexClasses++;
160 	CurrentLexClass = NumLexClasses-1;
161 	require(NumLexClasses<=MaxLexClasses, "number of allowable lexclasses exceeded\nIncrease MaxLexClasses in generic.h and recompile all C files");
162 	lclass[CurrentLexClass].classnum = m;
163 	lclass[CurrentLexClass].exprs = (char **) calloc(tsize, sizeof(char *));
164 	require(lclass[CurrentLexClass].exprs!=NULL,
165 			"lexclass: cannot allocate ExprStr");
166 	lclass[CurrentLexClass].htable = newHashTable();
167 	ExprStr = lclass[CurrentLexClass].exprs;
168 	Texpr = lclass[CurrentLexClass].htable;
169 	/* define EOF for each automaton */
170 	p = newTermEntry( EOFSTR );
171 	p->token = EofToken;	/* couldn't have remapped tokens yet, use EofToken */
172 	hash_add(Texpr, EOFSTR, (Entry *)p);
173 	list_add(&ExprOrder, (void *)newExpr(EOFSTR));
174 	/* note: we use the actual ExprStr array
175 	 * here as TokenInd doesn't exist yet
176 	 */
177 	ExprStr[EofToken] = EOFSTR;
178 }
179 
180 void
181 #ifdef __USE_PROTOS
lexmode(int i)182 lexmode( int i )
183 #else
184 lexmode( i )
185 int i;
186 #endif
187 {
188 	require(i<NumLexClasses, "lexmode: invalid mode");
189 	ExprStr = lclass[i].exprs;
190 	Texpr = lclass[i].htable;
191 	CurrentLexClass = i;
192 }
193 
194 /* return index into lclass array of lexical class. return -1 if nonexistent */
195 int
196 #ifdef __USE_PROTOS
LexClassIndex(char * cl)197 LexClassIndex( char *cl )
198 #else
199 LexClassIndex( cl )
200 char *cl;
201 #endif
202 {
203 	int i;
204 
205 	for (i=0; i<NumLexClasses; i++)
206 	{
207 		if ( strcmp(lclass[i].classnum, cl) == 0 ) return i;
208 	}
209 	return -1;
210 }
211 
212 int
213 #ifdef __USE_PROTOS
hasAction(char * expr)214 hasAction( char *expr )
215 #else
216 hasAction( expr )
217 char *expr;
218 #endif
219 {
220 	TermEntry *p;
221 	require(expr!=NULL, "hasAction: invalid expr");
222 
223 	p = (TermEntry *) hash_get(Texpr, expr);
224 	require(p!=NULL, eMsg1("hasAction: expr '%s' doesn't exist",expr));
225 	return (p->action!=NULL);
226 }
227 
228 void
229 #ifdef __USE_PROTOS
setHasAction(char * expr,char * action)230 setHasAction( char *expr, char *action )
231 #else
232 setHasAction( expr, action )
233 char *expr;
234 char *action;
235 #endif
236 {
237 	TermEntry *p;
238 	require(expr!=NULL, "setHasAction: invalid expr");
239 
240 	p = (TermEntry *) hash_get(Texpr, expr);
241 	require(p!=NULL, eMsg1("setHasAction: expr '%s' doesn't exist",expr));
242 	p->action = action;
243 }
244 
245 ForcedToken *
246 #ifdef __USE_PROTOS
newForcedToken(char * token,int tnum)247 newForcedToken(char *token, int tnum)
248 #else
249 newForcedToken(token, tnum)
250 char *token;
251 int tnum;
252 #endif
253 {
254 	ForcedToken *ft = (ForcedToken *) calloc(1, sizeof(ForcedToken));
255 	require(ft!=NULL, "out of memory");
256 	ft->token = token;
257 	ft->tnum = tnum;
258 	return ft;
259 }
260 
261 /*
262  * Make a token indirection array that remaps token numbers and then walk
263  * the appropriate symbol tables and SynDiag to change token numbers
264  */
265 void
266 #ifdef __USE_PROTOS
RemapForcedTokens(void)267 RemapForcedTokens(void)
268 #else
269 RemapForcedTokens()
270 #endif
271 {
272 	ListNode *p;
273 	ForcedToken *q;
274 	int max_token_number=0;     /* MR9 23-Sep-97 Removed "unsigned" */
275 	int i;
276 
277 	if ( ForcedTokens == NULL ) return;
278 
279 	/* find max token num */
280 	for (p = ForcedTokens->next; p!=NULL; p=p->next)
281 	{
282 		q = (ForcedToken *) p->elem;
283 		if ( q->tnum > max_token_number ) max_token_number = q->tnum;
284 	}
285 	fprintf(stderr, "max token number is %d\n", max_token_number);
286 
287 	/* make token indirection array */
288 	TokenInd = (int *) calloc(max_token_number+1, sizeof(int));
289 	LastTokenCounted = TokenNum;
290 	TokenNum = max_token_number+1;
291 	require(TokenInd!=NULL, "RemapForcedTokens: cannot allocate TokenInd");
292 
293 	/* fill token indirection array and change token id htable ; swap token indices */
294 	for (i=1; i<TokenNum; i++) TokenInd[i] = i;
295 	for (p = ForcedTokens->next; p!=NULL; p=p->next)
296 	{
297 		TermEntry *te;
298 		int old_pos, t;
299 
300 		q = (ForcedToken *) p->elem;
301 		fprintf(stderr, "%s forced to %d\n", q->token, q->tnum);
302 		te = (TermEntry *) hash_get(Tname, q->token);
303 		require(te!=NULL, "RemapForcedTokens: token not in hash table");
304 		old_pos = te->token;
305 		fprintf(stderr, "Before: TokenInd[old_pos==%d] is %d\n", old_pos, TokenInd[old_pos]);
306 		fprintf(stderr, "Before: TokenInd[target==%d] is %d\n", q->tnum, TokenInd[q->tnum]);
307 		q = (ForcedToken *) p->elem;
308 		t = TokenInd[old_pos];
309 		TokenInd[old_pos] = q->tnum;
310 		TokenInd[q->tnum] = t;
311 		te->token = q->tnum;		/* update token type id symbol table */
312 		fprintf(stderr, "After: TokenInd[old_pos==%d] is %d\n", old_pos, TokenInd[old_pos]);
313 		fprintf(stderr, "After: TokenInd[target==%d] is %d\n", q->tnum, TokenInd[q->tnum]);
314 
315 		/* Change the token number in the sym tab entry for the exprs
316 		 * at the old position of the token id and the target position
317 		 */
318 		/* update expr at target (if any) of forced token id */
319 		if ( q->tnum < TokenNum )	/* is it a valid position? */
320 		{
321 			for (i=0; i<NumLexClasses; i++)
322 			{
323 				if ( lclass[i].exprs[q->tnum]!=NULL )
324 				{
325 					/* update the symbol table for this expr */
326 					TermEntry *e = (TermEntry *) hash_get(lclass[i].htable, lclass[i].exprs[q->tnum]);
327 					require(e!=NULL, "RemapForcedTokens: expr not in hash table");
328 					e->token = old_pos;
329 					fprintf(stderr, "found expr '%s' at target %d in lclass[%d]; changed to %d\n",
330 							lclass[i].exprs[q->tnum], q->tnum, i, old_pos);
331 				}
332 			}
333 		}
334 		/* update expr at old position (if any) of forced token id */
335 		for (i=0; i<NumLexClasses; i++)
336 		{
337 			if ( lclass[i].exprs[old_pos]!=NULL )
338 			{
339 				/* update the symbol table for this expr */
340 				TermEntry *e = (TermEntry *) hash_get(lclass[i].htable, lclass[i].exprs[old_pos]);
341 				require(e!=NULL, "RemapForcedTokens: expr not in hash table");
342 				e->token = q->tnum;
343 				fprintf(stderr, "found expr '%s' for id %s in lclass[%d]; changed to %d\n",
344 						lclass[i].exprs[old_pos], q->token, i, q->tnum);
345 			}
346 		}
347 	}
348 
349 	/* Update SynDiag */
350 	RemapForcedTokensInSyntaxDiagram((Node *)SynDiag);
351 }
352 
353 static void
354 #ifdef __USE_PROTOS
RemapForcedTokensInSyntaxDiagram(Node * p)355 RemapForcedTokensInSyntaxDiagram(Node *p)
356 #else
357 RemapForcedTokensInSyntaxDiagram(p)
358 Node *p;
359 #endif
360 {
361 	Junction *j = (Junction *) p;
362 	RuleRefNode *r = (RuleRefNode *) p;
363 	TokNode *t = (TokNode *)p;
364 
365 	if ( p==NULL ) return;
366 	require(p->ntype>=1 && p->ntype<=NumNodeTypes,	"Remap...: invalid diagram node");
367 	switch ( p->ntype )
368 	{
369 		case nJunction :
370 			if ( j->visited ) return;
371 			if ( j->jtype == EndRule ) return;
372 			j->visited = TRUE;
373 			RemapForcedTokensInSyntaxDiagram( j->p1 );
374 			RemapForcedTokensInSyntaxDiagram( j->p2 );
375 			j->visited = FALSE;
376 			return;
377 		case nRuleRef :
378 			RemapForcedTokensInSyntaxDiagram( r->next );
379 			return;
380 		case nToken :
381 			if ( t->remapped ) return;	/* we've been here before */
382 			t->remapped = 1;
383 			fprintf(stderr, "remapping %d to %d\n", t->token, TokenInd[t->token]);
384 			t->token = TokenInd[t->token];
385 			RemapForcedTokensInSyntaxDiagram( t->next );
386 			return;
387 		case nAction :
388 			RemapForcedTokensInSyntaxDiagram( ((ActionNode *)p)->next );
389 			return;
390 		default :
391 			fatal_internal("invalid node type");
392 	}
393 }
394 
395 /*
396  * Add a token name.  Return the token number associated with it.  If it already
397  * exists, then return the token number assigned to it.
398  *
399  * Track the order in which tokens are found so that the DLG output maintains
400  * that order.  It also lets us map token numbers to strings.
401  */
402 int
403 #ifdef __USE_PROTOS
addTname(char * token)404 addTname( char *token )
405 #else
406 addTname( token )
407 char *token;
408 #endif
409 {
410 	TermEntry *p;
411 	require(token!=NULL, "addTname: invalid token name");
412 
413 	if ( (p=(TermEntry *)hash_get(Tname, token)) != NULL ) return p->token;
414 	p = newTermEntry( token );
415 	Ttrack( p->str );
416 	p->token = TokenNum++;
417 	hash_add(Tname, token, (Entry *)p);
418 	return p->token;
419 }
420 
421 /* This is the same as addTname except we force the TokenNum to be tnum.
422  * We don't have to use the Forced token stuff as no tokens will have
423  * been defined with #tokens when this is called.  This is only called
424  * when a #tokdefs meta-op is used.
425  */
426 int
427 #ifdef __USE_PROTOS
addForcedTname(char * token,int tnum)428 addForcedTname( char *token, int tnum )
429 #else
430 addForcedTname( token, tnum )
431 char *token;
432 int tnum;
433 #endif
434 {
435 	TermEntry *p;
436 	require(token!=NULL, "addTname: invalid token name");
437 
438 	if ( (p=(TermEntry *)hash_get(Tname, token)) != NULL ) return p->token;
439 	p = newTermEntry( token );
440 	Ttrack( p->str );
441 	p->token = tnum;
442 	hash_add(Tname, token, (Entry *)p);
443 	return p->token;
444 }
445 
446 /*
447  * Add a token expr.  Return the token number associated with it.  If it already
448  * exists, then return the token number assigned to it.
449  */
450 int
451 #ifdef __USE_PROTOS
addTexpr(char * expr)452 addTexpr( char *expr )
453 #else
454 addTexpr( expr )
455 char *expr;
456 #endif
457 {
458 	TermEntry *p;
459 	require(expr!=NULL, "addTexpr: invalid regular expression");
460 
461 	if ( (p=(TermEntry *)hash_get(Texpr, expr)) != NULL ) return p->token;
462 	p = newTermEntry( expr );
463 	Ttrack( p->str );
464 	/* track the order in which they occur */
465 	list_add(&ExprOrder, (void *)newExpr(p->str));
466 	p->token = TokenNum++;
467 	hash_add(Texpr, expr, (Entry *)p);
468 	return p->token;
469 }
470 
471 /* return the token number of 'term'.  Return 0 if no 'term' exists */
472 int
473 #ifdef __USE_PROTOS
Tnum(char * term)474 Tnum( char *term )
475 #else
476 Tnum( term )
477 char *term;
478 #endif
479 {
480 	TermEntry *p;
481 	require(term!=NULL, "Tnum: invalid terminal");
482 
483 	if ( *term=='"' ) p = (TermEntry *) hash_get(Texpr, term);
484 	else p = (TermEntry *) hash_get(Tname, term);
485 	if ( p == NULL ) return 0;
486 	else return p->token;
487 }
488 
489 /* associate a Name with an expr.  If both have been already assigned
490  * token numbers, then an error is reported.  Add the token or expr
491  * that has not been added if no error.  This 'represents' the #token
492  * ANTLR pseudo-op.  If both have not been defined, define them both
493  * linked to same token number.
494  */
495 void
496 #ifdef __USE_PROTOS
Tklink(char * token,char * expr)497 Tklink( char *token, char *expr )
498 #else
499 Tklink( token, expr )
500 char *token;
501 char *expr;
502 #endif
503 {
504 	TermEntry *p, *q;
505 	require(token!=NULL && expr!=NULL, "Tklink: invalid token name and/or expr");
506 
507 	p = (TermEntry *) hash_get(Tname, token);
508 	q = (TermEntry *) hash_get(Texpr, expr);
509 	if ( p != NULL && q != NULL )	/* both defined */
510 	{
511 		warn( eMsg2("token name %s and rexpr %s already defined; ignored",
512 					token, expr) );
513 		return;
514 	}
515 	if ( p==NULL && q==NULL )		/* both not defined */
516 	{
517 		int t = addTname( token );
518 		q = newTermEntry( expr );
519 		hash_add(Texpr, expr, (Entry *)q);
520 		q->token = t;
521 		/* note: we use the actual ExprStr array
522 		 * here as TokenInd doesn't exist yet
523 		 */
524 		ExprStr[t] = q->str;
525 		/* track the order in which they occur */
526 		list_add(&ExprOrder, (void *)newExpr(q->str));
527 		return;
528 	}
529 	if ( p != NULL )				/* one is defined, one is not */
530 	{
531 		q = newTermEntry( expr );
532 		hash_add(Texpr, expr, (Entry *)q);
533 		q->token = p->token;
534 		ExprStr[p->token] = q->str;	/* both expr and token str defined now */
535 		list_add(&ExprOrder, (void *)newExpr(q->str));
536 	}
537 	else							/* trying to associate name with expr here*/
538 	{
539 		p = newTermEntry( token );
540 		hash_add(Tname, token, (Entry *)p);
541 		p->token = q->token;
542 		TokenStr[p->token] = p->str;/* both expr and token str defined now */
543 	}
544 }
545 
546 /*
547  * Given a string, this function allocates and returns a pointer to a
548  * hash table record of size 'sz' whose "str" pointer is reset to a position
549  * in the string table.
550  */
551 Entry *
552 #ifdef __USE_PROTOS
newEntry(char * text,int sz)553 newEntry( char *text, int sz )
554 #else
555 newEntry( text, sz )
556 char *text;
557 int sz;
558 #endif
559 {
560 	Entry *p;
561 	require(text!=NULL, "new: NULL terminal");
562 
563 	if ( (p = (Entry *) calloc(1,sz)) == 0 )
564 	{
565 		fatal_internal("newEntry: out of memory for terminals\n");
566 		exit(PCCTS_EXIT_FAILURE);
567 	}
568 	p->str = mystrdup(text);
569 
570 	return(p);
571 }
572 
573 /*
574  * add an element to a list.
575  *
576  * Any non-empty list has a sentinel node whose 'elem' pointer is really
577  * a pointer to the last element.  (i.e. length(list) = #elemIn(list)+1).
578  * Elements are appended to the list.
579  */
580 void
581 #ifdef __USE_PROTOS
list_add(ListNode ** list,void * e)582 list_add( ListNode **list, void *e )
583 #else
584 list_add( list, e )
585 ListNode **list;
586 void *e;
587 #endif
588 {
589 	ListNode *p, *tail;
590 	require(e!=NULL, "list_add: attempting to add NULL list element");
591 
592 	p = newListNode;
593 	require(p!=NULL, "list_add: cannot alloc new list node");
594 	p->elem = e;
595 	if ( *list == NULL )
596 	{
597 		ListNode *sentinel = newListNode;
598 		require(sentinel!=NULL, "list_add: cannot alloc sentinel node");
599 		*list=sentinel;
600 		sentinel->next = p;
601 		sentinel->elem = (char *)p;		/* set tail pointer */
602 	}
603 	else								/* find end of list */
604 	{
605 		tail = (ListNode *) (*list)->elem;	/* get tail pointer */
606 		tail->next = p;
607 		(*list)->elem = (char *) p;		/* reset tail */
608 	}
609 }
610 
611 /* MR10 list_free() frees the ListNode elements in the list       */
612 /* MR10   if freeData then free the data elements of the list too */
613 
614 void
615 #ifdef __USE_PROTOS
list_free(ListNode ** list,int freeData)616 list_free(ListNode **list,int freeData)
617 #else
618 list_free(list,freeData)
619   ListNode      **list;
620   int           freeData;
621 #endif
622 {
623 	ListNode *p;
624     ListNode *next;
625 
626 	if (list == NULL) return;
627     if (*list == NULL) return;
628 	for (p=*list; p != NULL; p=next) {
629       next=p->next;
630       if (freeData && p->elem != NULL) {
631         free( (char *) p->elem);
632       };
633       free( (char *) p);
634     };
635     *list=NULL;
636 }
637 
638 void
639 #ifdef __USE_PROTOS
list_apply(ListNode * list,void (* f)(void *))640 list_apply( ListNode *list, void (*f)(void *) )
641 #else
642 list_apply( list, f )
643 ListNode *list;
644 void (*f)();
645 #endif
646 {
647 	ListNode *p;
648 	require(f!=NULL, "list_apply: NULL function to apply");
649 
650 	if ( list == NULL ) return;
651 	for (p = list->next; p!=NULL; p=p->next) (*f)( p->elem );
652 }
653 
654 /* MR27 */
655 
656 #ifdef __USE_PROTOS
list_search_cstring(ListNode * list,char * cstring)657 int list_search_cstring(ListNode *list, char * cstring)
658 #else
659 int list_search_cstring(list, cstring)
660   ListNode * list;
661   char * cstring;
662 #endif
663 {
664 	ListNode *p;
665 	if (list == NULL ) return 0;
666 	for (p = list->next; p!=NULL; p=p->next) {
667 		if (p->elem == NULL) continue;
668 		if (0 == strcmp((char *) p->elem , cstring)) return 1;
669 	}
670 	return 0;
671 }
672 
673 			/* F O L L O W  C y c l e  S t u f f */
674 
675 /* make a key based upon (rulename, computation, k value).
676  * Computation values are 'i'==FIRST, 'o'==FOLLOW.
677  */
678 
679 /* MR10  Make the key all characters so it can be read easily   */
680 /* MR10    by a simple dump program.  Also, separates           */
681 /* MR10   'o' and 'i' from rule name                            */
682 
683 char *
684 #ifdef __USE_PROTOS
Fkey(char * rule,int computation,int k)685 Fkey( char *rule, int computation, int k )
686 #else
687 Fkey( rule, computation, k )
688 char *rule;
689 int computation;
690 int k;
691 #endif
692 {
693 	static char key[MaxRuleName+2+2+1];                                 /* MR10 */
694 	int i;
695 
696 	if ( k > 99 )                                                       /* MR10 */
697 		fatal("k>99 is too big for this implementation of ANTLR!\n");   /* MR10 */
698 	if ( (i=strlen(rule)) > MaxRuleName )                               /* MR10 */
699 		fatal( eMsgd("rule name > max of %d\n", MaxRuleName) );         /* MR10 */
700 	strcpy(key,rule);
701 
702 /* MR10 */     key[i]='*';
703 /* MR10 */     key[i+1] = (char) computation; /* MR20 G. Hobbelt */
704 /* MR10 */     if (k < 10) {
705 /* MR10 */       key[i+2] = (char) ( '0' + k);
706 /* MR10 */  	 key[i+3] = '\0';
707 /* MR10 */     } else {
708 /* MR10 */       key[i+2] = (char) ( '0' + k/10);
709 /* MR10 */       key[i+3] = (char) ( '0' + k % 10);
710 /* MR10 */       key[i+4] = '\0';
711 /* MR10 */     };
712 
713 	return key;
714 }
715 
716 /* Push a rule onto the kth FOLLOW stack */
717 void
718 #ifdef __USE_PROTOS
FoPush(char * rule,int k)719 FoPush( char *rule, int k )
720 #else
721 FoPush( rule, k )
722 char *rule;
723 int k;
724 #endif
725 {
726 	RuleEntry *r;
727 	require(rule!=NULL, "FoPush: tried to push NULL rule");
728 	require(k<=CLL_k,	"FoPush: tried to access non-existent stack");
729 
730 	/*fprintf(stderr, "FoPush(%s)\n", rule);*/
731 	r = (RuleEntry *) hash_get(Rname, rule);
732 	if ( r == NULL ) {fatal_internal( eMsg1("rule %s must be defined but isn't", rule) );}
733 	if ( FoStack[k] == NULL )		/* Does the kth stack exist yet? */
734 	{
735 		/*fprintf(stderr, "allocating FoStack\n");*/
736 		FoStack[k] = (int *) calloc(FoStackSize, sizeof(int));
737 		require(FoStack[k]!=NULL, "FoPush: cannot allocate FOLLOW stack\n");
738 	}
739 	if ( FoTOS[k] == NULL )
740 	{
741 		FoTOS[k]=FoStack[k];
742 		*(FoTOS[k]) = r->rulenum;
743 	}
744 	else
745 	{
746 #ifdef MEMCHK
747 		require(valid(FoStack[k]), "FoPush: invalid FoStack");
748 #endif
749 		if ( FoTOS[k] >= &(FoStack[k][FoStackSize-1]) )
750 			fatal( eMsgd("exceeded max depth of FOLLOW recursion (%d)\n",
751 						FoStackSize) );
752 		require(FoTOS[k]>=FoStack[k],
753 				eMsg1("FoPush: FoStack stack-ptr is playing out of its sandbox",
754 					  rule));
755 		++(FoTOS[k]);
756 		*(FoTOS[k]) = r->rulenum;
757 	}
758 	{
759 		/*
760 ****		int *p;
761 ****		fprintf(stderr, "FoStack[k=%d]:\n", k);
762 ****		for (p=FoStack[k]; p<=FoTOS[k]; p++)
763 ****		{
764 ****			fprintf(stderr, "\t%s\n", RulePtr[*p]->rname);
765 ****		}
766 		*/
767 	}
768 }
769 
770 /* Pop one rule off of the FOLLOW stack.  TOS ptr is NULL if empty. */
771 void
772 #ifdef __USE_PROTOS
FoPop(int k)773 FoPop( int k )
774 #else
775 FoPop( k )
776 int k;
777 #endif
778 {
779 	require(k<=CLL_k, "FoPop: tried to access non-existent stack");
780 	/*fprintf(stderr, "FoPop\n");*/
781 	require(FoTOS[k]>=FoStack[k]&&FoTOS[k]<=&(FoStack[k][FoStackSize-1]),
782 			"FoPop: FoStack stack-ptr is playing out of its sandbox");
783 	if ( FoTOS[k] == FoStack[k] ) FoTOS[k] = NULL;
784 	else (FoTOS[k])--;
785 }
786 
787 /* Compute FOLLOW cycle.
788  * Mark all FOLLOW sets for rules in cycle as incomplete.
789  * Then, save cycle on the cycle list (Cycles) for later resolution.
790  * The Cycle is stored in the form:
791  *		(head of cycle==croot, rest of rules in cycle==cyclicDep)
792  *
793  * e.g. (Fo means "FOLLOW of", "-->" means requires or depends on)
794  *
795  *		Fo(x)-->Fo(a)-->Fo(b)-->Fo(c)-->Fo(x)
796  *										   ^----Infinite recursion (cycle)
797  *
798  * the cycle would be: x -> {a,b,c} or stored as (x,{a,b,c}).  Fo(x) depends
799  * on the FOLLOW of a,b, and c.  The root of a cycle is always complete after
800  * Fo(x) finishes.  Fo(a,b,c) however are not.  It turns out that all rules
801  * in a FOLLOW cycle have the same FOLLOW set.
802  */
803 void
804 #ifdef __USE_PROTOS
RegisterCycle(char * rule,int k)805 RegisterCycle( char *rule, int k )
806 #else
807 RegisterCycle( rule, k )
808 char *rule;
809 int k;
810 #endif
811 {
812 	CacheEntry *f;
813 	Cycle *c;
814 	int *p;
815 	RuleEntry *r;
816 	require(rule!=NULL, "RegisterCycle: tried to register NULL rule");
817 	require(k<=CLL_k,	"RegisterCycle: tried to access non-existent stack");
818 
819 	/*fprintf(stderr, "RegisterCycle(%s)\n", rule);*/
820 	/* Find cycle start */
821 	r = (RuleEntry *) hash_get(Rname, rule);
822 	require(r!=NULL,eMsg1("rule %s must be defined but isn't", rule));
823 	require(FoTOS[k]>=FoStack[k]&&FoTOS[k]<=&(FoStack[k][FoStackSize-1]),
824 			eMsg1("RegisterCycle(%s): FoStack stack-ptr is playing out of its sandbox",
825 				  rule));
826 /***	if ( FoTOS[k]<FoStack[k]||FoTOS[k]>&(FoStack[k][FoStackSize-1]) )
827 ****	{
828 ****		fprintf(stderr, "RegisterCycle(%s): FoStack stack-ptr is playing out of its sandbox\n",
829 ****						rule);
830 ****		fprintf(stderr, "RegisterCycle: sp==0x%x out of bounds 0x%x...0x%x\n",
831 ****						FoTOS[k], FoStack[k], &(FoStack[k][FoStackSize-1]));
832 ****		exit(PCCTS_EXIT_FAILURE);
833 ****	}
834 ****/
835 
836 #ifdef MEMCHK
837 	require(valid(FoStack[k]), "RegisterCycle: invalid FoStack");
838 #endif
839 	for (p=FoTOS[k]; *p != r->rulenum && p >= FoStack[k]; --p) {;}
840 	require(p>=FoStack[k], "RegisterCycle: FoStack is screwed up beyond belief");
841 	if ( p == FoTOS[k] ) return;	/* don't worry about cycles to oneself */
842 
843 	/* compute cyclic dependents (rules in cycle except head) */
844 	c = newCycle;
845 	require(c!=NULL, "RegisterCycle: couldn't alloc new cycle");
846 	c->cyclicDep = empty;
847 	c->croot = *p++;		/* record root of cycle */
848 	for (; p<=FoTOS[k]; p++)
849 	{
850 		/* Mark all dependent rules as incomplete */
851 		f = (CacheEntry *) hash_get(Fcache, Fkey(RulePtr[*p]->rname,'o',k));
852 		if ( f==NULL )
853 		{
854 			f = newCacheEntry( Fkey(RulePtr[*p]->rname,'o',k) );
855 			hash_add(Fcache, Fkey(RulePtr[*p]->rname,'o',k), (Entry *)f);
856 		}
857 		f->incomplete = TRUE;
858 
859 		set_orel(*p, &(c->cyclicDep)); /* mark rule as dependent of croot */
860 	}
861 	list_add(&(Cycles[k]), (void *)c);
862 }
863 
864 /* make all rules in cycle complete
865  *
866  * while ( some set has changed ) do
867  *		for each cycle do
868  *			if degree of FOLLOW set for croot > old degree then
869  *				update all FOLLOW sets for rules in cyclic dependency
870  *				change = TRUE
871  *			endif
872  *		endfor
873  * endwhile
874  */
875 void
876 #ifdef __USE_PROTOS
ResolveFoCycles(int k)877 ResolveFoCycles( int k )
878 #else
879 ResolveFoCycles( k )
880 int k;
881 #endif
882 {
883 	ListNode *p, *q;
884 	Cycle *c;
885 	int changed = 1;
886 	CacheEntry *f,*g;
887 	int r;
888 /*  int i;  */  /* MR10 not useful */
889 	unsigned d;
890 
891     unsigned    *cursor;        /* MR10 */
892     unsigned    *origin;        /* MR10 */
893 
894 	/*fprintf(stderr, "Resolving following cycles for %d\n", k);*/
895 	while ( changed )
896 	{
897 		changed = 0;
898 /* MR10 i = 0;  */
899 		for (p = Cycles[k]->next; p!=NULL; p=p->next)
900 		{
901 			c = (Cycle *) p->elem;
902 			/*fprintf(stderr, "cycle %d: %s -->", i++, RulePtr[c->croot]->rname);*/
903 			/*s_fprT(stderr, c->cyclicDep);*/
904 			/*fprintf(stderr, "\n");*/
905 			f = (CacheEntry *)
906 					hash_get(Fcache, Fkey(RulePtr[c->croot]->rname,'o',k));
907 			require(f!=NULL, eMsg1("FOLLOW(%s) must be in cache but isn't", RulePtr[c->croot]->rname) );
908 			if ( (d=set_deg(f->fset)) > c->deg )
909 			{
910 				/*fprintf(stderr, "Fo(%s) has changed\n", RulePtr[c->croot]->rname);*/
911 				changed = 1;
912 				c->deg = d;		/* update cycle FOLLOW set degree */
913 
914 /* MR10 */      origin=set_pdq(c->cyclicDep);
915 /* MR10 */      for (cursor=origin; *cursor != nil; cursor++) {
916 /* MR10 */         r=*cursor;
917 
918 /********		while ( !set_nil(c->cyclicDep) ) {      *****/
919 /********					r = set_int(c->cyclicDep);  *****/
920 /********					set_rm(r, c->cyclicDep);    *****/
921 
922 					/*fprintf(stderr, "updating Fo(%s)\n", RulePtr[r]->rname);*/
923 					g = (CacheEntry *)
924 							hash_get(Fcache, Fkey(RulePtr[r]->rname,'o',k));
925 					require(g!=NULL, eMsg1("FOLLOW(%s) must be in cache but isn't", RulePtr[r]->rname) );
926 					set_orin(&(g->fset), f->fset);
927 					g->incomplete = FALSE;
928 				}
929 /* MR10 */      free( (char *) origin);
930 /* MR10 */      origin=NULL;
931 			}
932 		}
933 /* MR10 - this if statement appears to be meaningless since i is always 0 */
934 /* MR10		if ( i == 1 ) changed = 0;	*/ /* if only 1 cycle, no need to repeat */
935 	}
936 	/* kill Cycle list */
937 	for (q = Cycles[k]->next; q != NULL; q=p)
938 	{
939 		p = q->next;
940 		set_free( ((Cycle *)q->elem)->cyclicDep );
941 		free((char *)q);
942 	}
943 	free( (char *)Cycles[k] );
944 	Cycles[k] = NULL;
945 }
946 
947 
948 			/* P r i n t i n g  S y n t a x  D i a g r a m s */
949 
950 static void
951 #ifdef __USE_PROTOS
pBlk(Junction * q,int btype)952 pBlk( Junction *q, int btype )
953 #else
954 pBlk( q, btype )
955 Junction *q;
956 int btype;
957 #endif
958 {
959 	int k,a;
960 	Junction *alt, *p;
961 
962 	q->end->pvisited = TRUE;
963 	if ( btype == aLoopBegin )
964 	{
965 		require(q->p2!=NULL, "pBlk: invalid ()* block");
966 		PRINT(q->p1);
967 		alt = (Junction *)q->p2;
968 		PRINT(alt->p1);
969 		if ( PrintAnnotate )
970 		{
971 			printf(" /* Opt ");
972 			k = 1;
973 			while ( !set_nil(alt->fset[k]) )
974 			{
975 				s_fprT(stdout, alt->fset[k]);
976 				if ( k++ == CLL_k ) break;
977 				if ( !set_nil(alt->fset[k]) ) printf(", ");
978 			}
979 			printf(" */\n");
980 		}
981 		return;
982 	}
983 	for (a=1,alt=q; alt != NULL; alt= (Junction *) alt->p2, a++)
984 	{
985 		if ( alt->p1 != NULL ) PRINT(alt->p1);
986 		if ( PrintAnnotate )
987 		{
988 			printf( " /* [%d] ", alt->altnum);
989 			k = 1;
990 			while ( !set_nil(alt->fset[k]) )
991 			{
992 				s_fprT(stdout, alt->fset[k]);
993 				if ( k++ == CLL_k ) break;
994 				if ( !set_nil(alt->fset[k]) ) printf(", ");
995 			}
996 			if ( alt->p2 == NULL && btype == aOptBlk )
997 				printf( " (optional branch) */\n");
998 			else printf( " */\n");
999 		}
1000 
1001 		/* ignore implied empty alt of Plus blocks */
1002 		if ( alt->p2 != NULL && ((Junction *)alt->p2)->ignore ) break;
1003 
1004 		if ( alt->p2 != NULL && !(((Junction *)alt->p2)->p2==NULL && btype == aOptBlk) )
1005 		{
1006 			if ( pLevel == 1 )
1007 			{
1008 				printf("\n");
1009 				if ( a+1==pAlt1 || a+1==pAlt2 ) printf("=>");
1010 				printf("\t");
1011 			}
1012 			else printf(" ");
1013 			printf("|");
1014 			if ( pLevel == 1 )
1015 			{
1016 				p = (Junction *) ((Junction *)alt->p2)->p1;
1017 				while ( p!=NULL )
1018 				{
1019 					if ( p->ntype==nAction )
1020 					{
1021 						p=(Junction *)((ActionNode *)p)->next;
1022 						continue;
1023 					}
1024 					if ( p->ntype!=nJunction )
1025 					{
1026 						break;
1027 					}
1028 					if ( p->jtype==EndBlk || p->jtype==EndRule )
1029 					{
1030 						p = NULL;
1031 						break;
1032 					}
1033 					p = (Junction *)p->p1;
1034 				}
1035 				if ( p==NULL ) printf("\n\t");	/* Empty alt? */
1036 			}
1037 		}
1038 	}
1039 	q->end->pvisited = FALSE;
1040 }
1041 
1042 /* How to print out a junction */
1043 void
1044 #ifdef __USE_PROTOS
pJunc(Junction * q)1045 pJunc( Junction *q )
1046 #else
1047 pJunc( q )
1048 Junction *q;
1049 #endif
1050 {
1051 	int dum_k;
1052 	int doing_rule;
1053 	require(q!=NULL, "pJunc: NULL node");
1054 	require(q->ntype==nJunction, "pJunc: not junction");
1055 
1056 	if ( q->pvisited == TRUE ) return;
1057 	q->pvisited = TRUE;
1058 	switch ( q->jtype )
1059 	{
1060 		case aSubBlk :
1061 			if ( PrintAnnotate ) First(q, 1, q->jtype, &dum_k);
1062 			if ( q->end->p1 != NULL && ((Junction *)q->end->p1)->ntype==nJunction &&
1063 				 ((Junction *)q->end->p1)->jtype == EndRule ) doing_rule = 1;
1064 			else doing_rule = 0;
1065 			pLevel++;
1066 			if ( pLevel==1 )
1067 			{
1068 				if ( pAlt1==1 ) printf("=>");
1069 				printf("\t");
1070 			}
1071 			else printf(" ");
1072 			if ( doing_rule )
1073 			{
1074 				if ( pLevel==1 ) printf(" ");
1075 				pBlk(q,q->jtype);
1076 			}
1077 			else {
1078 				printf("(");
1079 				if ( pLevel==1 ) printf(" ");
1080 				pBlk(q,q->jtype);
1081 				if ( pLevel>1 ) printf(" ");
1082 				printf(")");
1083 			}
1084 			if ( q->guess ) printf("?");
1085 			pLevel--;
1086 			if ( PrintAnnotate ) freeBlkFsets(q);
1087 			if ( q->end->p1 != NULL ) PRINT(q->end->p1);
1088 			break;
1089 		case aOptBlk :
1090 			if ( PrintAnnotate ) First(q, 1, q->jtype, &dum_k);
1091 			pLevel++;
1092 			if ( pLevel==1 )
1093 			{
1094 				if ( pAlt1==1 ) printf("=>");
1095 				printf("\t");
1096 			}
1097 			else printf(" ");
1098 			printf("{");
1099 			if ( pLevel==1 ) printf(" ");
1100 			pBlk(q,q->jtype);
1101 			if ( pLevel>1 ) printf(" ");
1102 			else printf("\n\t");
1103 			printf("}");
1104 			pLevel--;
1105 			if ( PrintAnnotate ) freeBlkFsets(q);
1106 			if ( q->end->p1 != NULL ) PRINT(q->end->p1);
1107 			break;
1108 		case aLoopBegin :
1109 			if ( PrintAnnotate ) First(q, 1, q->jtype, &dum_k);
1110 			pLevel++;
1111 			if ( pLevel==1 )
1112 			{
1113 				if ( pAlt1==1 ) printf("=>");
1114 				printf("\t");
1115 			}
1116 			else printf(" ");
1117 			printf("(");
1118 			if ( pLevel==1 ) printf(" ");
1119 			pBlk(q,q->jtype);
1120 			if ( pLevel>1 ) printf(" ");
1121 			else printf("\n\t");
1122 			printf(")*");
1123 			pLevel--;
1124 			if ( PrintAnnotate ) freeBlkFsets(q);
1125 			if ( q->end->p1 != NULL ) PRINT(q->end->p1);
1126 			break;
1127 		case aLoopBlk :
1128 			if ( PrintAnnotate ) First(q, 1, q->jtype, &dum_k);
1129 			pBlk(q,q->jtype);
1130 			if ( PrintAnnotate ) freeBlkFsets(q);
1131 			break;
1132 		case aPlusBlk :
1133 			if ( PrintAnnotate ) First(q, 1, q->jtype, &dum_k);
1134 			pLevel++;
1135 			if ( pLevel==1 )
1136 			{
1137 				if ( pAlt1==1 ) printf("=>");
1138 				printf("\t");
1139 			}
1140 			else printf(" ");
1141 			printf("(");
1142 			if ( pLevel==1 ) printf(" ");
1143 			pBlk(q,q->jtype);
1144 			if ( pLevel>1 ) printf(" ");
1145 			printf(")+");
1146 			pLevel--;
1147 			if ( PrintAnnotate ) freeBlkFsets(q);
1148 			if ( q->end->p1 != NULL ) PRINT(q->end->p1);
1149 			break;
1150 		case EndBlk :
1151 			break;
1152 		case RuleBlk :
1153 			printf( "\n%s :\n", q->rname);
1154 			PRINT(q->p1);
1155 			if ( q->p2 != NULL ) PRINT(q->p2);
1156 			break;
1157 		case Generic :
1158 			if ( q->p1 != NULL ) PRINT(q->p1);
1159 			q->pvisited = FALSE;
1160 			if ( q->p2 != NULL ) PRINT(q->p2);
1161 			break;
1162 		case EndRule :
1163 			printf( "\n\t;\n");
1164 			break;
1165 	}
1166 	q->pvisited = FALSE;
1167 }
1168 
1169 /* How to print out a rule reference node */
1170 void
1171 #ifdef __USE_PROTOS
pRuleRef(RuleRefNode * p)1172 pRuleRef( RuleRefNode *p )
1173 #else
1174 pRuleRef( p )
1175 RuleRefNode *p;
1176 #endif
1177 {
1178 	require(p!=NULL, "pRuleRef: NULL node");
1179 	require(p->ntype==nRuleRef, "pRuleRef: not rule ref node");
1180 
1181 	printf( " %s", p->text);
1182 	PRINT(p->next);
1183 }
1184 
1185 /* How to print out a terminal node */
1186 void
1187 #ifdef __USE_PROTOS
pToken(TokNode * p)1188 pToken( TokNode *p )
1189 #else
1190 pToken( p )
1191 TokNode *p;
1192 #endif
1193 {
1194 	require(p!=NULL, "pToken: NULL node");
1195 	require(p->ntype==nToken, "pToken: not token node");
1196 
1197 	if ( p->wild_card ) printf(" .");
1198 	printf( " %s", TerminalString(p->token));
1199 	PRINT(p->next);
1200 }
1201 
1202 /* How to print out a terminal node */
1203 void
1204 #ifdef __USE_PROTOS
pAction(ActionNode * p)1205 pAction( ActionNode *p )
1206 #else
1207 pAction( p )
1208 ActionNode *p;
1209 #endif
1210 {
1211 	require(p!=NULL, "pAction: NULL node");
1212 	require(p->ntype==nAction, "pAction: not action node");
1213 
1214 	PRINT(p->next);
1215 }
1216 
1217 					/* F i l l  F o l l o w  L i s t s */
1218 
1219 /*
1220  * Search all rules for all rule reference nodes, q to rule, r.
1221  * Add q->next to follow list dangling off of rule r.
1222  * i.e.
1223  *
1224  *		r: -o-R-o-->o--> Ptr to node following rule r in another rule
1225  *					|
1226  *					o--> Ptr to node following another reference to r.
1227  *
1228  * This is the data structure employed to avoid FOLLOW set computation.  We
1229  * simply compute the FIRST (reach) of the EndRule Node which follows the
1230  * list found at the end of all rules which are referenced elsewhere.  Rules
1231  * not invoked by other rules have no follow list (r->end->p1==NULL).
1232  * Generally, only start symbols are not invoked by another rule.
1233  *
1234  * Note that this mechanism also gives a free cross-reference mechanism.
1235  *
1236  * The entire syntax diagram is layed out like this:
1237  *
1238  * SynDiag
1239  *	|
1240  *	v
1241  *	o-->R1--o
1242  *	|
1243  *	o-->R2--o
1244  *	|
1245  *	...
1246  *	|
1247  *	o-->Rn--o
1248  *
1249  */
1250 void
1251 #ifdef __USE_PROTOS
FoLink(Node * p)1252 FoLink( Node *p )
1253 #else
1254 FoLink( p )
1255 Node *p;
1256 #endif
1257 {
1258 	RuleEntry *q;
1259 	Junction *j = (Junction *) p;
1260 	RuleRefNode *r = (RuleRefNode *) p;
1261 
1262 	if ( p==NULL ) return;
1263 	require(p->ntype>=1 && p->ntype<=NumNodeTypes,
1264 			eMsgd("FoLink: invalid diagram node: ntype==%d",p->ntype));
1265 	switch ( p->ntype )
1266 	{
1267 		case nJunction :
1268 			if ( j->fvisited ) return;
1269 			if ( j->jtype == EndRule ) return;
1270 			j->fvisited = TRUE;
1271 			FoLink( j->p1 );
1272 			FoLink( j->p2 );
1273 /* MR14 */
1274 /* MR14 */  /* Need to determine whether the guess block is an         */
1275 /* MR14 */  /* of the form (alpha)? beta before follow sets are        */
1276 /* MR14 */  /* computed.  This is necessary to solve problem           */
1277 /* MR14 */  /* of doing follow on the alpha of an (alpha)? beta block. */
1278 /* MR14 */
1279 /* MR14 */  /* This is performed by analysis_point as a side-effect.   */
1280 /* MR14 */
1281 /* MR14 */
1282 /* MR14 */  if (j->jtype == aSubBlk && j->guess) {
1283 /* MR14 */    Junction *ignore;
1284 /* MR14 */    ignore=analysis_point(j);
1285 /* MR14 */  }
1286 /* MR14 */
1287 			return;
1288 		case nRuleRef :
1289 			if ( r->linked ) return;
1290 			q = (RuleEntry *) hash_get(Rname, r->text);
1291 			if ( q == NULL )
1292 			{
1293 				warnFL( eMsg1("rule %s not defined",r->text), FileStr[r->file], r->line );
1294 			}
1295 			else
1296 			{
1297 				if ( r->parms!=NULL && RulePtr[q->rulenum]->pdecl==NULL )
1298 				{
1299 					warnFL( eMsg1("rule %s accepts no parameter(s)", r->text),
1300 							FileStr[r->file], r->line );
1301 				}
1302 				if ( r->parms==NULL && RulePtr[q->rulenum]->pdecl!=NULL )
1303 				{
1304 					warnFL( eMsg1("rule %s requires parameter(s)", r->text),
1305 							FileStr[r->file], r->line );
1306 				}
1307 				if ( r->assign!=NULL && RulePtr[q->rulenum]->ret==NULL )
1308 				{
1309 					warnFL( eMsg1("rule %s yields no return value(s)", r->text),
1310 							FileStr[r->file], r->line );
1311 				}
1312 				if ( r->assign==NULL && RulePtr[q->rulenum]->ret!=NULL )
1313 				{
1314 					warnFL( eMsg1("rule %s returns a value(s)", r->text),
1315 							FileStr[r->file], r->line );
1316 				}
1317 				if ( !r->linked )
1318 				{
1319 					addFoLink(	r->next, r->rname, RulePtr[q->rulenum] );
1320 					r->linked = TRUE;
1321 				}
1322 			}
1323 			FoLink( r->next );
1324 			return;
1325 		case nToken :
1326 			FoLink( ((TokNode *)p)->next );
1327 			return;
1328 		case nAction :
1329 			FoLink( ((ActionNode *)p)->next );
1330 			return;
1331 		default :
1332 			fatal_internal("invalid node type");
1333 	}
1334 }
1335 
1336 /*
1337  * Add a reference to the end of a rule.
1338  *
1339  * 'r' points to the RuleBlk node in a rule.  r->end points to the last node
1340  * (EndRule jtype) in a rule.
1341  *
1342  * Initial:
1343  *		r->end --> 	o
1344  *
1345  * After:
1346  *		r->end --> 	o-->o--> Ptr to node following rule r in another rule
1347  *						|
1348  *						o--> Ptr to node following another reference to r.
1349  *
1350  * Note that the links are added to the head of the list so that r->end->p1
1351  * always points to the most recently added follow-link.  At the end, it should
1352  * point to the last reference found in the grammar (starting from the 1st rule).
1353  */
1354 void
1355 #ifdef __USE_PROTOS
addFoLink(Node * p,char * rname,Junction * r)1356 addFoLink( Node *p, char *rname, Junction *r )
1357 #else
1358 addFoLink( p, rname, r )
1359 Node *p;
1360 char *rname;
1361 Junction *r;
1362 #endif
1363 {
1364 	Junction *j;
1365 	require(r!=NULL,				"addFoLink: incorrect rule graph");
1366 	require(r->end!=NULL,			"addFoLink: incorrect rule graph");
1367 	require(r->end->jtype==EndRule,	"addFoLink: incorrect rule graph");
1368 	require(p!=NULL,				"addFoLink: NULL FOLLOW link");
1369 
1370 	j = newJunction();
1371 	j->rname = rname;			/* rname on follow links point to target rule */
1372 	j->p1 = p;					/* link to other rule */
1373 	j->p2 = (Node *) r->end->p1;/* point to head of list */
1374 	r->end->p1 = (Node *) j;	/* reset head to point to new node */
1375 }
1376 
1377 void
1378 #ifdef __USE_PROTOS
GenCrossRef(Junction * p)1379 GenCrossRef( Junction *p )
1380 #else
1381 GenCrossRef( p )
1382 Junction *p;
1383 #endif
1384 {
1385 	set a;
1386 	Junction *j;
1387 	RuleEntry *q;
1388 	unsigned e;
1389 	require(p!=NULL, "GenCrossRef: why are you passing me a null grammar?");
1390 
1391 	printf("Cross Reference:\n\n");
1392 	a = empty;
1393 	for (; p!=NULL; p = (Junction *)p->p2)
1394 	{
1395 		printf("Rule %20s referenced by {", p->rname);
1396 		/* make a set of rules for uniqueness */
1397 		for (j = (Junction *)(p->end)->p1; j!=NULL; j = (Junction *)j->p2)
1398 		{
1399 			q = (RuleEntry *) hash_get(Rname, j->rname);
1400 			require(q!=NULL, "GenCrossRef: FoLinks are screwed up");
1401 			set_orel(q->rulenum, &a);
1402 		}
1403 		for (; !set_nil(a); set_rm(e, a))
1404 		{
1405 			e = set_int(a);
1406 			printf(" %s", RulePtr[e]->rname);
1407 		}
1408 		printf(" }\n");
1409 	}
1410 	set_free( a );
1411 }
1412 
1413 /*
1414    The single argument is a pointer to the start of an element of a
1415    C++ style function prototypet list.  Given a pointer to the start of
1416    an formal we must locate the comma (or the end of the string)
1417    and locate the datatype, formal name, and initial value expression.
1418 
1419    The function returns a pointer to the character following the comma
1420    which terminates the formal declaration, or a pointer to the end of
1421    the string if none was found.
1422 
1423    I thought we were parsing specialists, how come I'm doing this by
1424    hand written code ?
1425 
1426    Examples of input:
1427 
1428         Foo f,
1429         Foo f = Foo(1),
1430         Foo f = Foo(1,2),
1431         Foo f = &farray[1,2],
1432         Foo f = ",",
1433         Foo f = ',',
1434         TFoo<int,char> f = TFoo<int,char>(1,2),
1435 
1436    A non-zero value for nesting indicates a problem matching '(' and ')',
1437    '[' and ']', '<' and '>', '{' and '}', or improperly terminated string
1438    or character literal.
1439 
1440 */
1441 
1442 
1443 /*
1444  *  Don't care if it is a valid string literal or not, just find the end
1445  *  Start with pointer to leading "\""
1446  */
1447 
1448 #ifdef __USE_PROTOS
skipStringLiteral(char * pCurrent)1449 char * skipStringLiteral(char *pCurrent)
1450 #else
1451 char * skipStringLiteral(pCurrent)
1452 char *pCurrent;
1453 #endif
1454 {
1455   char *p = pCurrent;
1456   if (*p == 0) return p;
1457   require (*p == '\"', "skipStringLiteral")
1458   p++;
1459   for (p = p; *p != 0; p++) {
1460     if (*p == '\\') {
1461       p++;
1462       if (*p == 0) break;
1463       p++;
1464     }
1465     if (*p == '\"') {
1466       p++;
1467       break;
1468     }
1469   }
1470   return p;
1471 }
1472 
1473 /*
1474  *  Don't care if it is a valid character literal or not, just find the end
1475  *  Start with pointer to leading "'"
1476  */
1477 
1478 #ifdef __USE_PROTOS
skipCharLiteral(char * pStart)1479 char * skipCharLiteral(char *pStart)
1480 #else
1481 char * skipCharLiteral(pStart)
1482  char *pStart;
1483 #endif
1484 {
1485   char *p = pStart;
1486   if (*p == 0) return p;
1487   require (*p == '\'', "skipCharLiteral")
1488   p++;
1489   for (p = p; *p != 0; p++) {
1490     if (*p == '\\') {
1491       p++;
1492       if (*p == 0) break;
1493       p++;
1494     }
1495     if (*p == '\'') {
1496       p++;
1497       break;
1498     }
1499   }
1500   return p;
1501 }
1502 
1503 #ifdef __USE_PROTOS
skipSpaces(char * pStart)1504 char * skipSpaces(char *pStart)
1505 #else
1506 char * skipSpaces(pStart)
1507 char * pStart;
1508 #endif
1509 {
1510   char *p = pStart;
1511   while (*p != 0 && isspace(*p)) p++;
1512   return p;
1513 }
1514 
1515 #ifdef __USE_PROTOS
skipToSeparatorOrEqualSign(char * pStart,int * pNest)1516 char * skipToSeparatorOrEqualSign(char *pStart, int *pNest)
1517 #else
1518 char * skipToSeparatorOrEqualSign(pStart, pNest)
1519 char *pStart;
1520 int *pNest;
1521 #endif
1522 {
1523   char *p = pStart;
1524 
1525   int nest = 0;
1526 
1527   *pNest = (-1);
1528 
1529   while (*p != 0) {
1530     switch (*p) {
1531 
1532       case '(' :
1533       case '[' :
1534       case '<' :
1535       case '{' :
1536         nest++;
1537         p++;
1538         break;
1539 
1540       case ')' :
1541       case ']' :
1542       case '>' :
1543       case '}' :
1544         nest--;
1545         p++;
1546         break;
1547 
1548       case '"' :
1549         p = skipStringLiteral(p);
1550         break;
1551 
1552       case '\'' :
1553         p = skipCharLiteral(p);
1554         break;
1555 
1556       case '\\':
1557         p++;
1558         if (*p == 0) goto EXIT;
1559         p++;
1560         break;
1561 
1562       case ',':
1563       case '=':
1564         if (nest == 0) goto EXIT;
1565 		p++;
1566         break;
1567 
1568       default:
1569         p++;
1570     }
1571   }
1572 EXIT:
1573   *pNest = nest;
1574   return p;
1575 }
1576 
1577 #ifdef __USE_PROTOS
skipToSeparator(char * pStart,int * pNest)1578 char * skipToSeparator(char *pStart, int *pNest)
1579 #else
1580 char * skipToSeparator(pStart, pNest)
1581 char *pStart;
1582 int *pNest;
1583 #endif
1584 {
1585   char * p = pStart;
1586   for ( ; ; ) {
1587     p = skipToSeparatorOrEqualSign(p, pNest);
1588     if (*pNest != 0) return p;
1589     if (*p == ',') return p;
1590     if (*p == 0) return p;
1591 	p++;
1592   }
1593 }
1594 
1595 /* skip to just past the "=" separating the declaration from the initialization value */
1596 
1597 #ifdef __USE_PROTOS
getInitializer(char * pStart)1598 char * getInitializer(char *pStart)
1599 #else
1600 char * getInitializer(pStart)
1601 char * pStart;
1602 #endif
1603 {
1604 	char *p;
1605 	char *pDataType;
1606 	char *pSymbol;
1607 	char *pEqualSign;
1608 	char *pValue;
1609 	char *pSeparator;
1610 	int nest = 0;
1611 
1612 	require(pStart!=NULL, "getInitializer: invalid string");
1613 
1614 	p = endFormal(pStart,
1615 			      &pDataType,
1616 				  &pSymbol,
1617 				  &pEqualSign,
1618 				  &pValue,
1619 				  &pSeparator,
1620 				  &nest);
1621     if (nest != 0) return NULL;
1622     if (pEqualSign == NULL) return NULL;
1623     if (pValue == NULL) return NULL;
1624 	return strBetween(pValue, NULL, pSeparator);
1625 }
1626 
1627 /*
1628    Examines the string from pStart to pEnd-1.
1629    If the string has 0 length or is entirely white space
1630    returns 1.  Otherwise 0.
1631 */
1632 
1633 #ifdef __USE_PROTOS
isWhiteString(const char * pStart,const char * pEnd)1634 int isWhiteString(const char *pStart, const char *pEnd)
1635 #else
1636 int isWhiteString(pStart, pEnd)
1637 const char *pStart;
1638 const char *pEnd;
1639 #endif
1640 {
1641   const char *p;
1642   for (p = pStart; p < pEnd; p++) {
1643     if (! isspace(*p)) return 0;
1644   }
1645   return 1;
1646 }
1647 
1648 /*
1649    This replaces HasComma() which couldn't distinguish
1650 
1651         foo ["a,b"]
1652 
1653    from:
1654 
1655         foo[a,b]
1656 
1657 */
1658 
1659 #ifdef __USE_PROTOS
hasMultipleOperands(char * pStart)1660 int hasMultipleOperands(char *pStart)
1661 #else
1662 int hasMultipleOperands(pStart)
1663 char *pStart;
1664 #endif
1665 {
1666   char *p = pStart;
1667   int nest = 0;
1668 
1669   p = skipSpaces(p);
1670   if (*p == 0) return 0;
1671   p = skipToSeparator(p, &nest);
1672   if (nest == 0 && *p == ',') return 1;
1673   return 0;
1674 }
1675 
1676 
1677 #define MAX_STR_BETWEEN_WORK_AREA 1000
1678 
1679 static char strBetweenWorkArea[MAX_STR_BETWEEN_WORK_AREA];
1680 
1681 
1682 /*
1683 	strBetween(pStart, pNext, pStop)
1684 
1685     Creates a null terminated string by copying the text between two pointers
1686 	to a work area.  The start of the string is pStart.  The end of the string
1687 	is the character before pNext, or if pNext is null then the character before
1688 	pStop.  Trailing spaces are not included in the copy operation.
1689 
1690 	This is used when a string contains several parts.  The pNext part may be
1691 	optional.  The pStop will stop the scan when the optional part is not present
1692 	(is a null pointer).
1693 */
1694 
1695 #ifdef __USE_PROTOS
strBetween(char * pStart,char * pNext,char * pStop)1696 char *strBetween(char *pStart, char *pNext, char *pStop)
1697 #else
1698 char *strBetween(pStart, pNext, pStop)
1699 char *pStart;
1700 char *pNext;
1701 char *pStop;
1702 #endif
1703 {
1704   char *p;
1705   char *q = strBetweenWorkArea;
1706   const char *pEnd;
1707 
1708   pEnd = (pNext != NULL) ? pNext : pStop;
1709 
1710   require (pEnd != NULL, "pEnd == NULL");
1711   require (pEnd >= pStart, "pEnd < pStart");
1712   for (pEnd--; pEnd >= pStart; pEnd--) { /* MR31 */
1713 	if (! isspace(*pEnd)) break;
1714   }
1715   for (p = pStart;
1716        p <= pEnd && q < &strBetweenWorkArea[MAX_STR_BETWEEN_WORK_AREA-2];
1717 	   p++, q++) {
1718 	 *q = *p;
1719   }
1720   *q = 0;
1721   return strBetweenWorkArea;
1722 }
1723 
1724 /*
1725    function     Returns pointer to character following separator at
1726    value        which to continue search for next formal.  If at the
1727                 end of the string a pointer to the null byte at the
1728                 end of the string is returned.
1729 
1730    pStart       Pointer to the starting position of the formal list
1731 
1732                 This may be the middle of a longer string, for example
1733                 when looking for the end of formal #3 starting from
1734                 the middle of the complete formal list.
1735 
1736    ppDataType   Returns a pointer to the start of the data type in the
1737                 formal. Example: pointer to "Foo".
1738 
1739    ppSymbol     Returns a pointer to the start of the formal symbol.
1740                 Example: pointer to "f".
1741 
1742    ppEqualSign  Returns a pointer to the equal sign separating the
1743                 formal symbol from the initial value.  If there is
1744                 no "=" then this will be NULL.
1745 
1746    ppValue      Returns a pointer to the initial value part of the
1747                 formal declaration.  Example: pointer to "&farray[1,2]"
1748 
1749    ppSeparator  Returns a pointer to the character which terminated the
1750                 scan.  This should be a pointer to a comma or a null
1751                 byte which terminates the string.
1752 
1753    pNest        Returns the nesting level when a separator was found.
1754                 This is non-zero for any kind of error.  This is zero
1755                 for a successful parse of this portion of the formal
1756                 list.
1757 
1758 */
1759 
1760 #ifdef __USE_PROTOS
endFormal(char * pStart,char ** ppDataType,char ** ppSymbol,char ** ppEqualSign,char ** ppValue,char ** ppSeparator,int * pNest)1761 char * endFormal(char *pStart,
1762                  char **ppDataType,
1763                  char **ppSymbol,
1764                  char **ppEqualSign,
1765                  char **ppValue,
1766                  char **ppSeparator,
1767                  int *pNest)
1768 #else
1769 char * endFormal(pStart,
1770 			     ppDataType,
1771 				 ppSymbol,
1772 				 ppEqualSign,
1773 				 ppValue,
1774 				 ppSeparator,
1775 				 pNest)
1776 char *pStart;
1777 char **ppDataType;
1778 char **ppSymbol;
1779 char **ppEqualSign;
1780 char **ppValue;
1781 char **ppSeparator;
1782 int *pNest;
1783 
1784 #endif
1785 {
1786   char *p = pStart;
1787   char *q;
1788 
1789   *ppDataType = NULL;
1790   *ppSymbol = NULL;
1791   *ppEqualSign = NULL;
1792   *ppValue = NULL;
1793   *ppSeparator = NULL;
1794 
1795   *pNest = 0;
1796 
1797   /* The first non-blank is the start of the datatype */
1798 
1799   p = skipSpaces(p);
1800   if (*p == 0) goto EXIT;
1801   *ppDataType = p;
1802 
1803   /* We are not looking for the symbol, we are looking
1804      for the separator that follows the symbol.  Then
1805      we'll back up.
1806 
1807      Search for the ',' or '=" or null terminator.
1808    */
1809 
1810   p = skipToSeparatorOrEqualSign(p, pNest);
1811 
1812   if (*pNest != 0) goto EXIT;
1813 
1814   /*
1815      Work backwards to find start of symbol
1816      Skip spaces between the end of symbol and separator
1817      Assume that there are no spaces in the formal, but
1818      there is a space preceding the formal
1819   */
1820 
1821   for (q = &p[-1]; q >= *ppDataType; q--) {
1822     if (! isspace(*q)) break;
1823   }
1824   if (q < *ppDataType) goto EXIT;
1825 
1826   /*
1827      MR26 Handle things like: IIR_Bool (IIR_Decl::*constraint)()
1828      Backup until we hit the end of a symbol string, then find the
1829      start of the symbol string.  This wont' work for functions
1830      with prototypes, but works for the most common cases.  For
1831      others, use typedef names.
1832    */
1833 
1834 /* MR26 */  for (q = q; q >= *ppDataType; q--) {
1835 /* MR26 */    if (isalpha(*q) || isdigit(*q) || *q == '_' || *q == '$') break;
1836 /* MR26 */  }
1837 /* MR26 */  if (q < *ppDataType) goto EXIT;
1838 
1839   for (q = q; q >= *ppDataType; q--) {
1840     if ( ! (isalpha(*q) || isdigit(*q) || *q == '_' || *q == '$')) break;
1841   }
1842 
1843   *ppSymbol = &q[1];
1844 
1845   if (*p == ',' || *p == 0) {
1846     *ppSeparator = p;
1847     goto EXIT;
1848   }
1849 
1850   *ppEqualSign = p;
1851   p = skipSpaces(++p);
1852   *ppValue = p;
1853   if (*p == 0) goto EXIT;
1854 
1855 
1856   while (*p != 0 && *pNest == 0 && *p != ',') {
1857       p = skipToSeparator(p, pNest);
1858   }
1859   if (*pNest == 0) *ppSeparator = p;
1860 
1861 EXIT:
1862   if (*p == ',') p++;
1863   return p;
1864 }
1865