|
a/src/internfile/mh_html.cpp |
|
b/src/internfile/mh_html.cpp |
|
... |
|
... |
| 22 |
* -----END-LICENCE-----
|
22 |
* -----END-LICENCE-----
|
| 23 |
*/
|
23 |
*/
|
| 24 |
|
24 |
|
| 25 |
// This file has code from omindex + an adaptor function for recoll at the end
|
25 |
// This file has code from omindex + an adaptor function for recoll at the end
|
| 26 |
|
26 |
|
| 27 |
#include "htmlparse.h"
|
|
|
| 28 |
#include "mimehandler.h"
|
27 |
#include "mimehandler.h"
|
| 29 |
#include "debuglog.h"
|
28 |
#include "debuglog.h"
|
| 30 |
#include "csguess.h"
|
29 |
#include "csguess.h"
|
| 31 |
#include "readfile.h"
|
30 |
#include "readfile.h"
|
| 32 |
#include "transcode.h"
|
31 |
#include "transcode.h"
|
| 33 |
#include "mimeparse.h"
|
32 |
#include "mimeparse.h"
|
| 34 |
|
33 |
#include "myhtmlparse.h"
|
| 35 |
class MyHtmlParser : public HtmlParser {
|
34 |
#include "indextext.h"
|
| 36 |
public:
|
|
|
| 37 |
bool in_script_tag;
|
|
|
| 38 |
bool in_style_tag;
|
|
|
| 39 |
string title, sample, keywords, dump;
|
|
|
| 40 |
string ocharset; // This is the charset our user thinks the doc was
|
|
|
| 41 |
string charset; // This is the charset it was supposedly converted to
|
|
|
| 42 |
string doccharset; // Set this to value of charset parameter in header
|
|
|
| 43 |
bool indexing_allowed;
|
|
|
| 44 |
void process_text(const string &text);
|
|
|
| 45 |
void opening_tag(const string &tag, const map<string,string> &p);
|
|
|
| 46 |
void closing_tag(const string &tag);
|
|
|
| 47 |
MyHtmlParser() :
|
|
|
| 48 |
in_script_tag(false),
|
|
|
| 49 |
in_style_tag(false),
|
|
|
| 50 |
indexing_allowed(true) { }
|
|
|
| 51 |
};
|
|
|
| 52 |
|
|
|
| 53 |
void
|
|
|
| 54 |
MyHtmlParser::process_text(const string &text)
|
|
|
| 55 |
{
|
|
|
| 56 |
// some tags are meaningful mid-word so this is simplistic at best...
|
|
|
| 57 |
|
|
|
| 58 |
if (!in_script_tag && !in_style_tag) {
|
|
|
| 59 |
string::size_type firstchar = text.find_first_not_of(" \t\n\r");
|
|
|
| 60 |
if (firstchar != string::npos) {
|
|
|
| 61 |
dump += text.substr(firstchar);
|
|
|
| 62 |
dump += " ";
|
|
|
| 63 |
}
|
|
|
| 64 |
}
|
|
|
| 65 |
}
|
|
|
| 66 |
|
|
|
| 67 |
// lets hope that the charset includes ascii values...
|
|
|
| 68 |
static inline void
|
|
|
| 69 |
lowercase_term(string &term)
|
|
|
| 70 |
{
|
|
|
| 71 |
string::iterator i = term.begin();
|
|
|
| 72 |
while (i != term.end()) {
|
|
|
| 73 |
if (*i >= 'A' && *i <= 'Z')
|
|
|
| 74 |
*i = *i + 'a' - 'A';
|
|
|
| 75 |
i++;
|
|
|
| 76 |
}
|
|
|
| 77 |
}
|
|
|
| 78 |
|
35 |
|
| 79 |
#include <iostream>
|
36 |
#include <iostream>
|
| 80 |
using namespace std;
|
37 |
using namespace std;
|
| 81 |
|
38 |
|
| 82 |
|
|
|
| 83 |
void
|
|
|
| 84 |
MyHtmlParser::opening_tag(const string &tag, const map<string,string> &p)
|
|
|
| 85 |
{
|
|
|
| 86 |
#if 0
|
|
|
| 87 |
cout << "TAG: " << tag << ": " << endl;
|
|
|
| 88 |
map<string, string>::const_iterator x;
|
|
|
| 89 |
for (x = p.begin(); x != p.end(); x++) {
|
|
|
| 90 |
cout << " " << x->first << " -> '" << x->second << "'" << endl;
|
|
|
| 91 |
}
|
|
|
| 92 |
#endif
|
|
|
| 93 |
|
|
|
| 94 |
if (tag == "meta") {
|
|
|
| 95 |
map<string, string>::const_iterator i, j;
|
|
|
| 96 |
if ((i = p.find("content")) != p.end()) {
|
|
|
| 97 |
if ((j = p.find("name")) != p.end()) {
|
|
|
| 98 |
string name = j->second;
|
|
|
| 99 |
lowercase_term(name);
|
|
|
| 100 |
if (name == "description") {
|
|
|
| 101 |
if (sample.empty()) {
|
|
|
| 102 |
sample = i->second;
|
|
|
| 103 |
decode_entities(sample);
|
|
|
| 104 |
}
|
|
|
| 105 |
} else if (name == "keywords") {
|
|
|
| 106 |
if (!keywords.empty()) keywords += ' ';
|
|
|
| 107 |
string tmp = i->second;
|
|
|
| 108 |
decode_entities(tmp);
|
|
|
| 109 |
keywords += tmp;
|
|
|
| 110 |
} else if (name == "robots") {
|
|
|
| 111 |
string val = i->second;
|
|
|
| 112 |
decode_entities(val);
|
|
|
| 113 |
lowercase_term(val);
|
|
|
| 114 |
if (val.find("none") != string::npos ||
|
|
|
| 115 |
val.find("noindex") != string::npos) {
|
|
|
| 116 |
indexing_allowed = false;
|
|
|
| 117 |
throw true;
|
|
|
| 118 |
}
|
|
|
| 119 |
}
|
|
|
| 120 |
} else if ((j = p.find("http-equiv")) != p.end()) {
|
|
|
| 121 |
string hequiv = j->second;
|
|
|
| 122 |
lowercase_term(hequiv);
|
|
|
| 123 |
if (hequiv == "content-type") {
|
|
|
| 124 |
string value = i->second;
|
|
|
| 125 |
MimeHeaderValue p = parseMimeHeaderValue(value);
|
|
|
| 126 |
map<string, string>::const_iterator k;
|
|
|
| 127 |
if ((k = p.params.find("charset")) != p.params.end()) {
|
|
|
| 128 |
doccharset = k->second;
|
|
|
| 129 |
if (doccharset != ocharset) {
|
|
|
| 130 |
LOGDEB1(("Doc specified charset '%s' "
|
|
|
| 131 |
"differs from announced '%s'\n",
|
|
|
| 132 |
doccharset.c_str(), ocharset.c_str()));
|
|
|
| 133 |
throw true;
|
|
|
| 134 |
}
|
|
|
| 135 |
}
|
|
|
| 136 |
}
|
|
|
| 137 |
}
|
|
|
| 138 |
}
|
|
|
| 139 |
} else if (tag == "p" || tag == "br" || tag == "li") {
|
|
|
| 140 |
dump += "\n";
|
|
|
| 141 |
} else if (tag == "script") {
|
|
|
| 142 |
in_script_tag = true;
|
|
|
| 143 |
} else if (tag == "style") {
|
|
|
| 144 |
in_style_tag = true;
|
|
|
| 145 |
} else if (tag == "body") {
|
|
|
| 146 |
dump = "";
|
|
|
| 147 |
}
|
|
|
| 148 |
}
|
|
|
| 149 |
|
|
|
| 150 |
void
|
|
|
| 151 |
MyHtmlParser::closing_tag(const string &tag)
|
|
|
| 152 |
{
|
|
|
| 153 |
if (tag == "title") {
|
|
|
| 154 |
title = dump;
|
|
|
| 155 |
dump = "";
|
|
|
| 156 |
} else if (tag == "script") {
|
|
|
| 157 |
in_script_tag = false;
|
|
|
| 158 |
} else if (tag == "style") {
|
|
|
| 159 |
in_style_tag = false;
|
|
|
| 160 |
} else if (tag == "body") {
|
|
|
| 161 |
throw true;
|
|
|
| 162 |
}
|
|
|
| 163 |
}
|
|
|
| 164 |
|
39 |
|
| 165 |
bool textHtmlToDoc(RclConfig *conf, const string &fn,
|
40 |
bool textHtmlToDoc(RclConfig *conf, const string &fn,
|
| 166 |
const string &mtype, Rcl::Doc &docout)
|
41 |
const string &mtype, Rcl::Doc &docout)
|
| 167 |
{
|
42 |
{
|
| 168 |
LOGDEB(("textHtmlToDoc: %s\n", fn.c_str()));
|
43 |
LOGDEB(("textHtmlToDoc: %s\n", fn.c_str()));
|