Context Navigation

parser.y @ 2083

Last change on this file since 2083 was 407, checked in by ylafon@…, 12 years ago
unixism
File size: 10.6 KB

Rev	Line
[407]	1	%{
	2	/*
	3	* Bill's ABNF parser.
	4	* Copyright 2002-2004 William C. Fenner <fenner@fenron.com>
	5	* All rights reserved.
	6	*
	7	* Redistribution and use in source and binary forms, with or without
	8	* modification, are permitted provided that the following conditions
	9	* are met:
	10	* 1. Redistributions of source code must retain the above copyright
	11	* notice, this list of conditions and the following disclaimer.
	12	* 2. Redistributions in binary form must reproduce the above copyright
	13	* notice, this list of conditions and the following disclaimer in the
	14	* documentation and/or other materials provided with the distribution.
	15	* 3. Neither the name of the author nor the names of contributors
	16	* may be used to endorse or promote products derived from this software
	17	* without specific prior written permission.
	18	*
	19	* THIS SOFTWARE IS PROVIDED BY WILLIAM C. FENNER ``AS IS'' AND
	20	* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
	21	* THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
	22	* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WILLIAM C. FENNER OR HIS
	23	* BROTHER B1FF BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
	24	* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
	25	* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
	26	* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
	27	* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
	28	* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
	29	* USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
	30	* DAMAGE.
	31	*/
	32
	33	#include "config.h"
	34	#include <stdio.h>
	35	#include <stdlib.h>
	36	#include <stdarg.h>
	37	#include <string.h>
	38
	39	#include "common.h"
	40
	41	static const char rcsid[] =
	42	"$Id: parser.y,v 1.1 2008-05-28 12:01:58 jre Exp $";
	43
	44	extern int yylineno, yycolumn, yyerrors;
	45
	46	int defline;
	47	extern struct rule *rules;
	48	extern char *input_file;
	49
	50	#if defined(YYERROR_VERBOSE) && defined(YYBISON)
	51	#define MYERROR_VERBOSE
	52	#endif
	53	#ifdef MYERROR_VERBOSE
	54	/* HACK-O-ROONIE for yyerror verbosity */
	55	int *yystatep = NULL;
	56	int *yychar1p = NULL;
	57	#endif
	58
	59	int pipewarn = 0;
	60
	61	object *newobj(int);
	62	int yyerror(char *);
	63	int yylex(void);
	64	%}
	65
	66	%union {
	67	char *string;
	68	struct range range;
	69	object *object;
	70	int retval;
	71	}
	72
	73	%token <string> CHARVAL PROSEVAL BINVAL DECVAL HEXVAL RULENAME
	74	%token <range> BINVALRANGE DECVALRANGE HEXVALRANGE REPEAT LIST
	75	%token CWSP EQSLASH CRLF
	76
	77	%type <string> numval
	78	%type <range> numvalrange
	79	%type <object> element group option repetition list elements
	80	%type <object> rulerest
	81	%type <retval> definedas
	82
	83	%%
	84
	85	begin: {
	86	#ifdef MYERROR_VERBOSE
	87	/* HACK-O-RAMA */ yystatep = &yystate; yychar1p = &yychar1;
	88	#endif
	89	} rulelist
	90	;
	91
	92	rulelist: rules
	93	\| rulelist rules
	94	;
	95
	96	rules: rule
	97	\| starcwsp CRLF
	98	\| cwsp RULENAME { mywarn(MYERROR, "Indented rules are Right Out."); YYERROR; }
	99	;
	100
	101	recover:
	102	\| error CRLF
	103	;
	104
	105	rule: recover RULENAME { defline = yylineno; } definedas rulerest {
	106	struct rule *r;
	107
	108	r = findrule($2);
	109
	110	if ($4 == 0 \|\| r->name == NULL \|\| r->rule == NULL) { /* = */
	111	if ($4) {
	112	mywarn(MYERROR, "Rule %s does not yet exist; treating /= as =", $2);
	113	}
	114	if (r->name && strcmp(r->name, $2))
	115	if (r->rule)
	116	mywarn(MYERROR, "Rule %s previously defined as %s on line %d",
	117	$2, r->name, r->line);
	118	else
	119	mywarn(MYWARNING, "rule %s previously referred to as %s",
	120	$2, r->name);
	121	if (r->rule)
	122	mywarn(MYERROR, "Rule %s was already defined on line %d of %s", $2,
	123	r->line, (r->file? r->file : "stdin"));
	124	else {
	125	r->name = $2;
	126	r->line = defline;
	127	r->file = input_file;
	128	r->rule = $5;
	129	if (r->next != rules) {
	130	/* unlink r from where it is and move to the end */
	131	r->prev->next = r->next;
	132	r->next->prev = r->prev;
	133	if (rules == r)
	134	rules = r->next;
	135	r->prev = rules->prev;
	136	rules->prev = r;
	137	r->prev->next = r;
	138	r->next = rules;
	139	}
	140	}
	141	} else { /* =/ */
	142	object *tmp;
	143
	144	tmp = newobj(T_ALTERNATION);
	145	tmp->u.alternation.left = r->rule;
	146	tmp->u.alternation.right = $5;
	147	r->rule = tmp;
	148	}
	149	}
	150	;
	151
	152	definedas: starcwsp EQSLASH starcwsp { $$ = 1; }
	153	\| starcwsp '=' starcwsp { $$ = 0; }
	154	\| starcwsp '=' starcwsp CRLF { mywarn(MYERROR, "Empty rule"); YYERROR; /* XXX this CRLF may be the one required to recover from the error, but we've already used it... */}
	155	\| starcwsp repetition { mywarn(MYERROR, "Got definitions, expecting '=' or '=/'"); YYERROR; }
	156	\| starcwsp CRLF { mywarn(MYERROR, "Got EOL, expecting '=' or '=/'"); YYERROR; /* XXX this CRLF may be the one required to recover from the error, but we've already used it... */}
	157	;
	158
	159	cwsp: CWSP
	160	;
	161
	162	starcwsp:
	163	\| CWSP
	164	/* \| CRLF { mywarn(MYERROR, "Blank lines are not permitted inside rules"); YYERROR; } */
	165	;
	166
	167	rulerest: elements starcwsp CRLF { $$ = $1; }
	168	\| elements ')' { mywarn(MYERROR, "Extra ')'? Missing '('?"); YYERROR; }
	169	\| elements ']' { mywarn(MYERROR, "Extra ']'? Missing '['?"); YYERROR; }
	170	;
	171
	172	elements:
	173	repetition
	174	\| list
	175	\| elements cwsp repetition {
	176	/* concatenation */
	177	object *o = $1;
	178
	179	$$ = $1;
	180	if (o->type == T_ALTERNATION)
	181	o = o->u.alternation.right;
	182	while (o->next) /* n*2, do this better /
	183	o = o->next;
	184	o->next = $3;
	185	}
	186	\| elements cwsp list {
	187	/* concatenation */
	188	object *o = $1;
	189
	190	$$ = $1;
	191	if (o->type == T_ALTERNATION)
	192	o = o->u.alternation.right;
	193	while (o->next) /* n*2, do this better /
	194	o = o->next;
	195	o->next = $3;
	196	}
	197	\| elements starcwsp '/' starcwsp repetition {
	198	$$ = newobj(T_ALTERNATION);
	199	$$->u.alternation.left = $1;
	200	$$->u.alternation.right = $5;
	201	}
	202	\| elements starcwsp '\|' starcwsp repetition {
	203	if (!pipewarn) {
	204	mywarn(MYERROR, "'/' is the alternation character in ABNF");
	205	pipewarn = 1;
	206	}
	207	$$ = newobj(T_ALTERNATION);
	208	$$->u.alternation.left = $1;
	209	$$->u.alternation.right = $5;
	210	}
	211	\| elements starcwsp '/' starcwsp list {
	212	$$ = newobj(T_ALTERNATION);
	213	$$->u.alternation.left = $1;
	214	$$->u.alternation.right = $5;
	215	}
	216	\| elements starcwsp '\|' starcwsp list {
	217	if (!pipewarn) {
	218	mywarn(MYERROR, "'/' is the alternation character in ABNF");
	219	pipewarn = 1;
	220	}
	221	$$ = newobj(T_ALTERNATION);
	222	$$->u.alternation.left = $1;
	223	$$->u.alternation.right = $5;
	224	}
	225	\| elements repetition {
	226	object *o = $1;
	227	mywarn(MYERROR, "Concatenation of adjacent elements is not allowed (missing whitespace?)");
	228	printf("; line %d ... trying to concatenate ", yylineno);
	229	if (o->type == T_ALTERNATION)
	230	o = o->u.alternation.right;
	231	while (o->next) /* n*2, do this better /
	232	o = o->next;
	233	printobj(o, 1);
	234	printf("\n; ... with ");
	235	printobj($2, 1);
	236	printf("\n");
	237	YYERROR;
	238	}
	239	\| elements starcwsp '=' {
	240	mywarn(MYERROR, "Encountered definition while parsing rule (Indented rule?)");
	241	YYERROR;
	242	}
	243	\| elements starcwsp EQSLASH {
	244	mywarn(MYERROR, "Encountered definition while parsing rule (Indented rule?)");
	245	YYERROR;
	246	}
	247	;
	248
	249	repetition:
	250	element
	251	\| REPEAT element {
	252	$$ = $2;
	253	/* 510[foo] is really 10(foo), so leave
	254	* the zero that [ put there. */
	255	if ($$->u.e.repetition.lo != 0)
	256	$$->u.e.repetition.lo = $1.lo;
	257	$$->u.e.repetition.hi = $1.hi;
	258	if ($1.hi < $1.lo && $1.hi != -1)
	259	mywarn(MYERROR, "Repeat range swapped, should be min*max");
	260	if ($1.hi == 0)
	261	mywarn(MYFYI, "absolute repeat count of zero means this element may not occur at all");
	262	}
	263	\| REPEAT cwsp {
	264	mywarn(MYERROR, "No whitespace allowed between repeat and element.");
	265	YYERROR;
	266	}
	267	;
	268
	269	list:
	270	LIST element {
	271	$$ = $2;
	272	if ($$->u.e.repetition.lo != 0)
	273	$$->u.e.repetition.lo = $1.lo;
	274	$$->u.e.repetition.hi = $1.hi;
	275	if ($1.hi < $1.lo && $1.hi != -1)
	276	mywarn(MYERROR, "List range swapped, should be min*max");
	277	if ($1.hi == 0)
	278	mywarn(MYFYI, "absolute list count of zero means this element may not occur at all");
	279	if ($1.lo != 0 && $1.lo != 1 && $1.lo != -1)
	280	mywarn(MYERROR, "List range min must be 0 or 1");
	281	if ($1.hi != -1)
	282	mywarn(MYERROR, "List range max must be infinity");
	283	$$->u.e.islist = 1;
	284	}
	285	\| LIST cwsp {
	286	mywarn(MYERROR, "No whitespace allowed between list and element.");
	287	YYERROR;
	288	}
	289	;
	290
	291	numval: BINVAL
	292	\| DECVAL
	293	\| HEXVAL
	294	;
	295
	296	numvalrange:
	297	BINVALRANGE
	298	\| DECVALRANGE
	299	\| HEXVALRANGE
	300	;
	301
	302	element:
	303	RULENAME {
	304	$$ = newobj(T_RULE);
	305	$$->u.e.e.rule.name = $1;
	306	$$->u.e.e.rule.rule = findrule($1);
	307	if (strcmp($1, $$->u.e.e.rule.rule->name))
	308	mywarn(MYWARNING, "rule %s defined on line %d referred to as %s", $$->u.e.e.rule.rule->name, $$->u.e.e.rule.rule->line, $1);
	309	}
	310	\| group
	311	\| option
	312	\| CHARVAL {
	313	char *p = $1;
	314	if (*$1)
	315	p += strlen($1) - 1;
	316	if (p == '\n' \|\| p == '\r') {
	317	mywarn(MYERROR, "unterminated quoted-string");
	318	YYERROR;
	319	}
	320	$$ = newobj(T_TERMSTR);
	321	$$->u.e.e.termstr.str = $1;
	322	$$->u.e.e.termstr.flags = 0;
	323	}
	324	\| numval {
	325	$$ = newobj(T_TERMSTR);
	326	$$->u.e.e.termstr.str = $1;
	327	$$->u.e.e.termstr.flags = F_CASESENSITIVE;
	328	}
	329	\| numvalrange {
	330	$$ = newobj(T_TERMRANGE);
	331	$$->u.e.e.termrange.lo = $1.lo;
	332	$$->u.e.e.termrange.hi = $1.hi;
	333	}
	334	\| PROSEVAL {
	335	$$ = newobj(T_PROSE);
	336	$$->u.e.e.proseval = $1;
	337	if (strcmp($1, "\"") == 0) {
	338	mywarn(MYFYI, "suggest DQUOTE or %%x22 instead of <\">.");
	339	}
	340	}
	341	;
	342
	343	group: '(' starcwsp elements starcwsp ')' {
	344	$$ = newobj(T_GROUP);
	345	$$->u.e.e.group = $3;
	346	}
	347	\| '(' starcwsp elements starcwsp CRLF {
	348	mywarn(MYERROR, "Missing ')'? Extra '('?");
	349	YYERROR;
	350	}
	351	;
	352
	353	option: '[' starcwsp elements starcwsp ']' {
	354	$$ = newobj(T_GROUP);
	355	$$->u.e.e.group = $3;
	356	$$->u.e.repetition.lo = 0;
	357	}
	358	\| '[' starcwsp elements starcwsp CRLF {
	359	mywarn(MYERROR, "Missing ']'? Extra '['?");
	360	YYERROR;
	361	}
	362	;
	363
	364	%%
	365
	366	void
	367	mywarn(int level, const char *fmt, ...)
	368	{
	369	va_list ap;
	370
	371	/* file name */
	372	fprintf(stderr, "%s(%d:%d): ", input_file, yylineno, yycolumn);
	373	switch (level) {
	374	case MYFYI: fprintf(stderr, "fyi: "); break;
	375	case MYWARNING: fprintf(stderr, "warning: "); break;
	376	case MYERROR: ++yyerrors; /* fall through */
	377	default: fprintf(stderr, "error: "); break;
	378	}
	379	va_start(ap, fmt);
	380	vfprintf(stderr, fmt, ap);
	381	fprintf(stderr, "\n");
	382	}
	383
	384	int
	385	yyerror(char *s)
	386	{
	387	#ifdef MYERROR_VERBOSE
	388	mywarn(MYERROR, "state %d, token %s: %s",
	389	*yystatep,
	390	(yychar1p && (yychar1p >= 0 && yychar1p <= (sizeof(yytname)/sizeof(yytname[0])))) ? yytname[*yychar1p] : "?",
	391	s);
	392	#else
	393	mywarn(MYERROR, "%s\n", s);
	394	#endif
	395	return 0;
	396	}
	397
	398	object *
	399	newobj(int type)
	400	{
	401	object *o;
	402
	403	o = calloc(sizeof(object), 1);
	404	if (o == NULL) {
	405	mywarn(MYERROR, "out of memory");
	406	exit(1);
	407	}
	408	o->type = type;
	409	o->next = NULL;
	410	switch (type) {
	411	case T_RULE:
	412	case T_GROUP:
	413	case T_TERMSTR:
	414	case T_TERMRANGE:
	415	case T_PROSE:
	416	o->u.e.repetition.lo = 1;
	417	o->u.e.repetition.hi = 1;
	418	break;
	419	}
	420	return o;
	421	}

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

Original Format