1 /* This file is part of Metaproxy.
2 Copyright (C) 2005-2013 Index Data
4 Metaproxy is free software; you can redistribute it and/or modify it under
5 the terms of the GNU General Public License as published by the Free
6 Software Foundation; either version 2, or (at your option) any later
9 Metaproxy is distributed in the hope that it will be useful, but WITHOUT ANY
10 WARRANTY; without even the implied warranty of MERCHANTABILITY or
11 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
20 #include "html_parser.hpp"
28 #define SPACECHR " \t\r\n\f"
31 namespace metaproxy_1 {
32 class HTMLParser::Rep {
33 friend class HTMLParser;
35 void parse_str(HTMLParserEvent &event, const char *cp);
36 void tagText(HTMLParserEvent &event,
37 const char *text_start, const char *text_end);
38 int tagEnd(HTMLParserEvent &event,
39 const char *tag, int tag_len, const char *cp);
40 int tagAttrs(HTMLParserEvent &event,
41 const char *name, int len,
43 int skipAttribute(HTMLParserEvent &event,
44 const char *cp, int *attr_len,
45 const char **value, int *val_len, int *tr);
52 namespace mp = metaproxy_1;
54 mp::HTMLParser::Rep::Rep()
59 mp::HTMLParser::Rep::~Rep()
63 mp::HTMLParser::HTMLParser() : m_p(new Rep)
67 mp::HTMLParser::~HTMLParser()
71 void mp::HTMLParser::set_verbose(int v)
77 void mp::HTMLParser::parse(mp::HTMLParserEvent & event, const char *str) const
79 m_p->parse_str(event, str);
82 static int isAlpha(int c)
84 return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z');
87 static int skipSpace(const char *cp)
90 while (cp[i] && strchr(SPACECHR, cp[i]))
95 static int skipName(const char *cp)
98 for (i = 0; cp[i] && !strchr(SPACECHR "/>=<", cp[i]); i++)
103 int mp::HTMLParser::Rep::skipAttribute(HTMLParserEvent &event,
104 const char *cp, int *attr_len,
105 const char **value, int *val_len,
109 int i = skipName(cp);
113 return skipSpace(cp);
114 i += skipSpace(cp + i);
118 i += skipSpace(cp + i);
119 if (cp[i] == '\"' || cp[i] == '\'')
123 while (cp[i] != *tr && cp[i])
133 while (cp[i] && !strchr(SPACECHR ">", cp[i]))
139 i += skipSpace(cp + i);
144 int mp::HTMLParser::Rep::tagAttrs(HTMLParserEvent &event,
145 const char *name, int len,
148 int i = skipSpace(cp);
149 while (cp[i] && cp[i] != '>' && cp[i] != '/' && cp[i] != '<')
151 const char *attr_name = cp + i;
157 int nor = skipAttribute(event, cp+i, &attr_len, &value, &val_len, &tr);
165 printf ("------ attr %.*s=%.*s\n", attr_len, attr_name,
167 event.attribute(name, len, attr_name, attr_len, value, val_len, x);
172 int mp::HTMLParser::Rep::tagEnd(HTMLParserEvent &event,
173 const char *tag, int tag_len, const char *cp)
177 for (; cp[i] && cp[i] != '/' && cp[i] != '>' && cp[i] != '<'; i++)
182 printf("------ text %.*s\n", i, cp);
193 printf("------ any tag %s %.*s\n",
194 close_it ? " close" : "end", tag_len, tag);
195 event.anyTagEnd(tag, tag_len, close_it);
201 void mp::HTMLParser::Rep::tagText(HTMLParserEvent &event,
202 const char *text_start, const char *text_end)
204 if (text_end - text_start) //got text to flush
207 printf("------ text %.*s\n",
208 (int) (text_end - text_start), text_start);
209 event.text(text_start, text_end-text_start);
213 void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
215 const char *text_start = cp;
224 tagText(event, text_start, cp - 1);
225 if (cp[1] == '-' && cp[2] == '-')
227 for (i = 3; cp[i]; i++)
228 if (cp[i] == '-' && cp[i+1] == '-' && cp[i+2] == '>')
231 event.openTagStart(cp, i);
237 for (i = 1; cp[i] && cp[i] != '>'; i++)
239 event.openTagStart(cp, i);
242 printf("------ dtd %.*s\n", i, cp);
243 i += tagEnd(event, cp, i, cp + i);
250 tagText(event, text_start, cp - 1);
251 for (i = 1; cp[i] && cp[i] != '>'; i++)
253 event.openTagStart(cp, i);
255 printf("------ pi %.*s\n", i, cp);
256 i += tagEnd(event, cp, i, cp + i);
260 else if (*cp == '/' && isAlpha(cp[1]))
263 tagText(event, text_start, cp - 1);
266 event.closeTag(cp, i);
268 printf("------ tag close %.*s\n", i, cp);
269 i += tagEnd(event, cp, i, cp + i);
273 else if (isAlpha(*cp))
276 tagText(event, text_start, cp - 1);
278 event.openTagStart(cp, i);
280 printf("------ tag open %.*s\n", i, cp);
281 j = tagAttrs(event, cp, i, cp + i);
282 j += tagEnd(event, cp, i, cp + i + j);
287 tagText(event, text_start, cp);
290 mp::HTMLParserEvent::~HTMLParserEvent()
297 * c-file-style: "Stroustrup"
298 * indent-tabs-mode: nil
300 * vim: shiftwidth=4 tabstop=8 expandtab