3 %option outfile="dtd_parse.c"
4 %option prefix="Dtd_Parse_"
5 %option never-interactive
10 * an XML dissector for Wireshark
11 * lexical analyzer for DTDs
13 * Copyright 2004, Luis E. Garcia Ontanon <luis.ontanon@gmail.com>
17 * Wireshark - Network traffic analyzer
18 * By Gerald Combs <gerald@wireshark.org>
19 * Copyright 1998 Gerald Combs
21 * This program is free software; you can redistribute it and/or
22 * modify it under the terms of the GNU General Public License
23 * as published by the Free Software Foundation; either version 2
24 * of the License, or (at your option) any later version.
26 * This program is distributed in the hope that it will be useful,
27 * but WITHOUT ANY WARRANTY; without even the implied warranty of
28 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
29 * GNU General Public License for more details.
31 * You should have received a copy of the GNU General Public License
32 * along with this program; if not, write to the Free Software
33 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
40 #include "dtd_grammar.h"
41 #include "dtd_parse.h"
43 struct _proto_xmlpi_attr {
49 static GString* input_string;
52 static gchar* location;
53 static gchar* attr_name;
55 static int my_yyinput(char* buff,guint size);
57 static dtd_token_data_t* new_token(gchar*);
59 static dtd_build_data_t* build_data;
61 static void set_proto_name (gchar* val) { if(build_data->proto_name) g_free(build_data->proto_name); build_data->proto_name = g_strdup(val); }
62 static void set_media_type (gchar* val) { if(build_data->media_type) g_free(build_data->media_type); build_data->media_type = g_strdup(val); }
63 static void set_proto_root (gchar* val) { if(build_data->proto_root) g_free(build_data->proto_root); build_data->proto_root = g_strdup(val); }
64 static void set_description (gchar* val) { if(build_data->description) g_free(build_data->description); build_data->description = g_strdup(val); }
65 static void set_recursive (gchar* val) { build_data->recursion = ( g_strcasecmp(val,"yes") == 0 ) ? TRUE : FALSE; }
67 static struct _proto_xmlpi_attr proto_attrs[] =
69 { "proto_name", set_proto_name },
70 { "media", set_media_type },
71 { "root", set_proto_root },
72 { "description", set_description },
73 { "hierarchy", set_recursive },
77 #ifdef DEBUG_DTD_PARSER
78 #define DEBUG_DTD_TOKEN fprintf(stderr,"->%s (%i)%s\n",location,token_type,yytext)
80 #define DEBUG_DTD_TOKEN
83 #define DTD_PARSE(token_type) \
85 DtdParse(pParser, (token_type), new_token(yytext), build_data); \
86 if(build_data->error->len > 0) yyterminate(); \
90 #define YY_INPUT(buff,result,max_size) ( (result) = my_yyinput((buff),(max_size)) )
99 location_xmlpi "wireshark:location"
100 protocol_xmlpi "wireshark:protocol"
102 get_attr_quote =[:blank:]*["]
105 get_location_xmlpi [^[:blank:]]+
109 notation_tag "<!"[:blank:]*NOTATION
113 whitespace [[:blank:]\r\n]+
150 name [A-Za-z][-a-zA-Z0-9_]*
154 %START DTD XMLPI LOCATION DONE PROTOCOL GET_ATTR_QUOTE GET_ATTR_VAL GET_ATTR_CLOSE_QUOTE IN_COMMENT IN_NOTATION
160 <DTD>{comment_start} { BEGIN IN_COMMENT; }
163 <IN_COMMENT>{comment_stop} { BEGIN DTD; }
165 <DTD>{notation_tag} { BEGIN IN_NOTATION; }
167 <IN_NOTATION>{special_stop} { BEGIN DTD; }
173 <XMLPI>{location_xmlpi} {
177 <XMLPI>{protocol_xmlpi} {
182 <XMLPI>{stop_xmlpi} BEGIN DTD;
184 <LOCATION>{get_location_xmlpi} {
185 if(location) g_free(location);
186 location = g_strdup(yytext);
190 <DONE>{stop_xmlpi} BEGIN DTD;
193 attr_name = g_strdup(yytext);
194 g_strdown(attr_name);
195 BEGIN GET_ATTR_QUOTE;
198 <GET_ATTR_QUOTE>{get_attr_quote} { BEGIN GET_ATTR_VAL; }
201 g_string_sprintfa(build_data->error,
202 "error in wireshark:protocol xmpli at %s : could not find attribute value!",
207 <GET_ATTR_VAL>[^"]+ {
209 struct _proto_xmlpi_attr* pa;
210 gboolean got_it = FALSE;
212 for(pa = proto_attrs; pa->name; pa++) {
213 if (g_strcasecmp(attr_name,pa->name) == 0) {
221 g_string_sprintfa(build_data->error,
222 "error in wireshark:protocol xmpli at %s : no such parameter %s!",
223 location, attr_name);
230 BEGIN GET_ATTR_CLOSE_QUOTE;
233 <GET_ATTR_CLOSE_QUOTE>{dquote} { BEGIN PROTOCOL;}
235 <PROTOCOL>{stop_xmlpi} BEGIN DTD;
237 <DTD>{special_start} { DTD_PARSE(TOKEN_TAG_START); }
238 <DTD>{special_stop} { DTD_PARSE(TOKEN_TAG_STOP); }
240 <DTD>{attlist_kw} { DTD_PARSE(TOKEN_ATTLIST_KW); }
241 <DTD>{element_kw} { DTD_PARSE(TOKEN_ELEMENT_KW); }
242 <DTD>{doctype_kw} { DTD_PARSE(TOKEN_DOCTYPE_KW); }
244 <DTD>{pcdata} { DTD_PARSE(TOKEN_ELEM_DATA); }
245 <DTD>{any} { DTD_PARSE(TOKEN_ELEM_DATA); }
246 <DTD>{cdata} { DTD_PARSE(TOKEN_ELEM_DATA); }
247 <DTD>{empty} { DTD_PARSE(TOKEN_EMPTY_KW); }
249 <DTD>{iD} { DTD_PARSE(TOKEN_ATT_TYPE); }
250 <DTD>{idref} { DTD_PARSE(TOKEN_ATT_TYPE); }
251 <DTD>{idrefs} { DTD_PARSE(TOKEN_ATT_TYPE); }
252 <DTD>{nmtoken} { DTD_PARSE(TOKEN_ATT_TYPE); }
253 <DTD>{nmtokens} { DTD_PARSE(TOKEN_ATT_TYPE); }
254 <DTD>{entity} { DTD_PARSE(TOKEN_ATT_TYPE); }
255 <DTD>{entities} { DTD_PARSE(TOKEN_ATT_TYPE); }
256 <DTD>{notation} { DTD_PARSE(TOKEN_ATT_TYPE); }
257 <DTD>{cdata_t} { DTD_PARSE(TOKEN_ATT_TYPE); }
258 <DTD>{defaulT} { DTD_PARSE(TOKEN_ATT_DEF_WITH_VALUE); }
259 <DTD>{fixed} { DTD_PARSE(TOKEN_ATT_DEF_WITH_VALUE); }
260 <DTD>{required} { DTD_PARSE(TOKEN_ATT_DEF); }
261 <DTD>{implied} { DTD_PARSE(TOKEN_ATT_DEF); }
263 <DTD>{star} { DTD_PARSE(TOKEN_STAR); }
264 <DTD>{question} { DTD_PARSE(TOKEN_QUESTION); }
265 <DTD>{plus} { DTD_PARSE(TOKEN_PLUS); }
266 <DTD>{comma} { DTD_PARSE(TOKEN_COMMA); }
267 <DTD>{open_parens} { DTD_PARSE(TOKEN_OPEN_PARENS); }
268 <DTD>{close_parens} { DTD_PARSE(TOKEN_CLOSE_PARENS); }
269 <DTD>{open_bracket} { DTD_PARSE(TOKEN_OPEN_BRACKET); }
270 <DTD>{close_bracket} { DTD_PARSE(TOKEN_CLOSE_BRACKET); }
271 <DTD>{pipe} { DTD_PARSE(TOKEN_PIPE); }
274 <DTD>{squoted} { DTD_PARSE(TOKEN_QUOTED); }
275 <DTD>{name} { DTD_PARSE(TOKEN_NAME); }
279 static dtd_token_data_t* new_token(gchar* text) {
280 dtd_token_data_t* t = g_malloc(sizeof(dtd_token_data_t));
282 t->text = g_strdup(text);
283 t->location = g_strdup(location);
289 static int my_yyinput(char* buff, guint size) {
291 if (offset >= len ) {
293 } else if ( offset + size <= len ) {
294 memcpy(buff, input_string->str + offset,size);
299 memcpy(buff, input_string->str + offset,size);
305 extern dtd_build_data_t* dtd_parse(GString* s) {
309 len = input_string->len;
311 pParser = DtdParseAlloc(g_malloc);
313 #ifdef DEBUG_DTD_PARSER
314 DtdParseTrace(stderr, ">>");
317 build_data = g_malloc(sizeof(dtd_build_data_t));
319 build_data->proto_name = NULL;
320 build_data->media_type = NULL;
321 build_data->description = NULL;
322 build_data->proto_root = NULL;
323 build_data->recursion = FALSE;
325 build_data->elements = g_ptr_array_new();
326 build_data->attributes = g_ptr_array_new();
328 build_data->error = g_string_new("");
336 DtdParse(pParser, 0, NULL,build_data);
342 DtdParseFree(pParser, g_free );