|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* Helper class for the --prefetch option of dumpTextPass.php |
|
4
|
|
|
* |
|
5
|
|
|
* Copyright © 2005 Brion Vibber <[email protected]> |
|
6
|
|
|
* https://www.mediawiki.org/ |
|
7
|
|
|
* |
|
8
|
|
|
* This program is free software; you can redistribute it and/or modify |
|
9
|
|
|
* it under the terms of the GNU General Public License as published by |
|
10
|
|
|
* the Free Software Foundation; either version 2 of the License, or |
|
11
|
|
|
* (at your option) any later version. |
|
12
|
|
|
* |
|
13
|
|
|
* This program is distributed in the hope that it will be useful, |
|
14
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
15
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
16
|
|
|
* GNU General Public License for more details. |
|
17
|
|
|
* |
|
18
|
|
|
* You should have received a copy of the GNU General Public License along |
|
19
|
|
|
* with this program; if not, write to the Free Software Foundation, Inc., |
|
20
|
|
|
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
|
21
|
|
|
* http://www.gnu.org/copyleft/gpl.html |
|
22
|
|
|
* |
|
23
|
|
|
* @file |
|
24
|
|
|
* @ingroup Maintenance |
|
25
|
|
|
*/ |
|
26
|
|
|
|
|
27
|
|
|
/** |
|
28
|
|
|
* Readahead helper for making large MediaWiki data dumps; |
|
29
|
|
|
* reads in a previous XML dump to sequentially prefetch text |
|
30
|
|
|
* records already normalized and decompressed. |
|
31
|
|
|
* |
|
32
|
|
|
* This can save load on the external database servers, hopefully. |
|
33
|
|
|
* |
|
34
|
|
|
* Assumes that dumps will be recorded in the canonical order: |
|
35
|
|
|
* - ascending by page_id |
|
36
|
|
|
* - ascending by rev_id within each page |
|
37
|
|
|
* - text contents are immutable and should not change once |
|
38
|
|
|
* recorded, so the previous dump is a reliable source |
|
39
|
|
|
* |
|
40
|
|
|
* @ingroup Maintenance |
|
41
|
|
|
*/ |
|
42
|
|
|
class BaseDump { |
|
43
|
|
|
protected $reader = null; |
|
44
|
|
|
protected $atEnd = false; |
|
45
|
|
|
protected $atPageEnd = false; |
|
46
|
|
|
protected $lastPage = 0; |
|
47
|
|
|
protected $lastRev = 0; |
|
48
|
|
|
protected $infiles = null; |
|
49
|
|
|
|
|
50
|
|
|
public function __construct( $infile ) { |
|
51
|
|
|
$this->infiles = explode( ';', $infile ); |
|
52
|
|
|
$this->reader = new XMLReader(); |
|
53
|
|
|
$infile = array_shift( $this->infiles ); |
|
54
|
|
|
if ( defined( 'LIBXML_PARSEHUGE' ) ) { |
|
55
|
|
|
$this->reader->open( $infile, null, LIBXML_PARSEHUGE ); |
|
56
|
|
|
} else { |
|
57
|
|
|
$this->reader->open( $infile ); |
|
58
|
|
|
} |
|
59
|
|
|
} |
|
60
|
|
|
|
|
61
|
|
|
/** |
|
62
|
|
|
* Attempts to fetch the text of a particular page revision |
|
63
|
|
|
* from the dump stream. May return null if the page is |
|
64
|
|
|
* unavailable. |
|
65
|
|
|
* |
|
66
|
|
|
* @param int $page ID number of page to read |
|
67
|
|
|
* @param int $rev ID number of revision to read |
|
68
|
|
|
* @return string|null |
|
69
|
|
|
*/ |
|
70
|
|
|
function prefetch( $page, $rev ) { |
|
71
|
|
|
$page = intval( $page ); |
|
72
|
|
|
$rev = intval( $rev ); |
|
73
|
|
|
while ( $this->lastPage < $page && !$this->atEnd ) { |
|
74
|
|
|
$this->debug( "BaseDump::prefetch at page $this->lastPage, looking for $page" ); |
|
75
|
|
|
$this->nextPage(); |
|
76
|
|
|
} |
|
77
|
|
|
if ( $this->lastPage > $page || $this->atEnd ) { |
|
78
|
|
|
$this->debug( "BaseDump::prefetch already past page $page " |
|
79
|
|
|
. "looking for rev $rev [$this->lastPage, $this->lastRev]" ); |
|
80
|
|
|
|
|
81
|
|
|
return null; |
|
82
|
|
|
} |
|
83
|
|
|
while ( $this->lastRev < $rev && !$this->atEnd && !$this->atPageEnd ) { |
|
84
|
|
|
$this->debug( "BaseDump::prefetch at page $this->lastPage, rev $this->lastRev, " |
|
85
|
|
|
. "looking for $page, $rev" ); |
|
86
|
|
|
$this->nextRev(); |
|
87
|
|
|
} |
|
88
|
|
|
if ( $this->lastRev == $rev && !$this->atEnd ) { |
|
89
|
|
|
$this->debug( "BaseDump::prefetch hit on $page, $rev [$this->lastPage, $this->lastRev]" ); |
|
90
|
|
|
|
|
91
|
|
|
return $this->nextText(); |
|
92
|
|
|
} else { |
|
93
|
|
|
$this->debug( "BaseDump::prefetch already past rev $rev on page $page " |
|
94
|
|
|
. "[$this->lastPage, $this->lastRev]" ); |
|
95
|
|
|
|
|
96
|
|
|
return null; |
|
97
|
|
|
} |
|
98
|
|
|
} |
|
99
|
|
|
|
|
100
|
|
|
function debug( $str ) { |
|
101
|
|
|
wfDebug( $str . "\n" ); |
|
102
|
|
|
// global $dumper; |
|
103
|
|
|
// $dumper->progress( $str ); |
|
104
|
|
|
} |
|
105
|
|
|
|
|
106
|
|
|
/** |
|
107
|
|
|
* @access private |
|
108
|
|
|
*/ |
|
109
|
|
|
function nextPage() { |
|
110
|
|
|
if ( $this->skipTo( 'page', 'mediawiki' ) ) { |
|
111
|
|
View Code Duplication |
if ( $this->skipTo( 'id' ) ) { |
|
112
|
|
|
$this->lastPage = intval( $this->nodeContents() ); |
|
113
|
|
|
$this->lastRev = 0; |
|
114
|
|
|
$this->atPageEnd = false; |
|
115
|
|
|
} |
|
116
|
|
|
} else { |
|
117
|
|
|
$this->close(); |
|
118
|
|
|
if ( count( $this->infiles ) ) { |
|
119
|
|
|
$infile = array_shift( $this->infiles ); |
|
120
|
|
|
$this->reader->open( $infile ); |
|
121
|
|
|
$this->atEnd = false; |
|
122
|
|
|
} |
|
123
|
|
|
} |
|
124
|
|
|
} |
|
125
|
|
|
|
|
126
|
|
|
/** |
|
127
|
|
|
* @access private |
|
128
|
|
|
*/ |
|
129
|
|
|
function nextRev() { |
|
130
|
|
View Code Duplication |
if ( $this->skipTo( 'revision' ) ) { |
|
131
|
|
|
if ( $this->skipTo( 'id' ) ) { |
|
132
|
|
|
$this->lastRev = intval( $this->nodeContents() ); |
|
133
|
|
|
} |
|
134
|
|
|
} else { |
|
135
|
|
|
$this->atPageEnd = true; |
|
136
|
|
|
} |
|
137
|
|
|
} |
|
138
|
|
|
|
|
139
|
|
|
/** |
|
140
|
|
|
* @access private |
|
141
|
|
|
* @return string |
|
142
|
|
|
*/ |
|
143
|
|
|
function nextText() { |
|
144
|
|
|
$this->skipTo( 'text' ); |
|
145
|
|
|
|
|
146
|
|
|
return strval( $this->nodeContents() ); |
|
147
|
|
|
} |
|
148
|
|
|
|
|
149
|
|
|
/** |
|
150
|
|
|
* @access private |
|
151
|
|
|
* @param string $name |
|
152
|
|
|
* @param string $parent |
|
153
|
|
|
* @return bool|null |
|
154
|
|
|
*/ |
|
155
|
|
|
function skipTo( $name, $parent = 'page' ) { |
|
156
|
|
|
if ( $this->atEnd ) { |
|
157
|
|
|
return false; |
|
158
|
|
|
} |
|
159
|
|
|
while ( $this->reader->read() ) { |
|
160
|
|
|
if ( $this->reader->nodeType == XMLReader::ELEMENT |
|
161
|
|
|
&& $this->reader->name == $name |
|
162
|
|
|
) { |
|
163
|
|
|
return true; |
|
164
|
|
|
} |
|
165
|
|
|
if ( $this->reader->nodeType == XMLReader::END_ELEMENT |
|
166
|
|
|
&& $this->reader->name == $parent |
|
167
|
|
|
) { |
|
168
|
|
|
$this->debug( "BaseDump::skipTo found </$parent> searching for <$name>" ); |
|
169
|
|
|
|
|
170
|
|
|
return false; |
|
171
|
|
|
} |
|
172
|
|
|
} |
|
173
|
|
|
|
|
174
|
|
|
return $this->close(); |
|
175
|
|
|
} |
|
176
|
|
|
|
|
177
|
|
|
/** |
|
178
|
|
|
* Shouldn't something like this be built-in to XMLReader? |
|
179
|
|
|
* Fetches text contents of the current element, assuming |
|
180
|
|
|
* no sub-elements or such scary things. |
|
181
|
|
|
* |
|
182
|
|
|
* @return string |
|
183
|
|
|
* @access private |
|
184
|
|
|
*/ |
|
185
|
|
View Code Duplication |
function nodeContents() { |
|
186
|
|
|
if ( $this->atEnd ) { |
|
187
|
|
|
return null; |
|
188
|
|
|
} |
|
189
|
|
|
if ( $this->reader->isEmptyElement ) { |
|
190
|
|
|
return ""; |
|
191
|
|
|
} |
|
192
|
|
|
$buffer = ""; |
|
193
|
|
|
while ( $this->reader->read() ) { |
|
194
|
|
|
switch ( $this->reader->nodeType ) { |
|
195
|
|
|
case XMLReader::TEXT: |
|
196
|
|
|
// case XMLReader::WHITESPACE: |
|
197
|
|
|
case XMLReader::SIGNIFICANT_WHITESPACE: |
|
198
|
|
|
$buffer .= $this->reader->value; |
|
199
|
|
|
break; |
|
200
|
|
|
case XMLReader::END_ELEMENT: |
|
201
|
|
|
return $buffer; |
|
202
|
|
|
} |
|
203
|
|
|
} |
|
204
|
|
|
|
|
205
|
|
|
return $this->close(); |
|
206
|
|
|
} |
|
207
|
|
|
|
|
208
|
|
|
/** |
|
209
|
|
|
* @access private |
|
210
|
|
|
* @return null |
|
211
|
|
|
*/ |
|
212
|
|
|
function close() { |
|
213
|
|
|
$this->reader->close(); |
|
214
|
|
|
$this->atEnd = true; |
|
215
|
|
|
|
|
216
|
|
|
return null; |
|
217
|
|
|
} |
|
218
|
|
|
} |
|
219
|
|
|
|
wikimedia /
mediawiki
