1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
|
<?php
// Some smart guy removed XMLReader's global constants from PHP 5.1
// and replaced them with class constants. Breaking source compatibility
// is SUPER awesome, and I love languages which do this constantly!
$xmlReaderConstants = array(
"NONE",
"ELEMENT",
"ATTRIBUTE",
"TEXT",
"CDATA",
"ENTITY_REF",
"ENTITY",
"PI",
"COMMENT",
"DOC",
"DOC_TYPE",
"DOC_FRAGMENT",
"NOTATION",
"WHITESPACE",
"SIGNIFICANT_WHITESPACE",
"END_ELEMENT",
"END_ENTITY",
"XML_DECLARATION",
"LOADDTD",
"DEFAULTATTRS",
"VALIDATE",
"SUBST_ENTITIES" );
foreach( $xmlReaderConstants as $name ) {
$fullName = "XMLREADER_$name";
$newName = "XMLReader::$name";
if( !defined( $fullName ) ) {
if( defined( $newName ) ) {
define( $fullName, constant( $newName ) );
} else {
// broken or missing the extension...
}
}
}
/**
* Readahead helper for making large MediaWiki data dumps;
* reads in a previous XML dump to sequentially prefetch text
* records already normalized and decompressed.
*
* This can save load on the external database servers, hopefully.
*
* Assumes that dumps will be recorded in the canonical order:
* - ascending by page_id
* - ascending by rev_id within each page
* - text contents are immutable and should not change once
* recorded, so the previous dump is a reliable source
*
* Requires PHP 5 and the XMLReader PECL extension.
*/
class BaseDump {
var $reader = null;
var $atEnd = false;
var $atPageEnd = false;
var $lastPage = 0;
var $lastRev = 0;
function BaseDump( $infile ) {
$this->reader = new XMLReader();
$this->reader->open( $infile );
}
/**
* Attempts to fetch the text of a particular page revision
* from the dump stream. May return null if the page is
* unavailable.
*
* @param int $page ID number of page to read
* @param int $rev ID number of revision to read
* @return string or null
*/
function prefetch( $page, $rev ) {
$page = intval( $page );
$rev = intval( $rev );
while( $this->lastPage < $page && !$this->atEnd ) {
$this->debug( "BaseDump::prefetch at page $this->lastPage, looking for $page" );
$this->nextPage();
}
if( $this->lastPage > $page || $this->atEnd ) {
$this->debug( "BaseDump::prefetch already past page $page looking for rev $rev [$this->lastPage, $this->lastRev]" );
return null;
}
while( $this->lastRev < $rev && !$this->atEnd && !$this->atPageEnd ) {
$this->debug( "BaseDump::prefetch at page $this->lastPage, rev $this->lastRev, looking for $page, $rev" );
$this->nextRev();
}
if( $this->lastRev == $rev && !$this->atEnd ) {
$this->debug( "BaseDump::prefetch hit on $page, $rev [$this->lastPage, $this->lastRev]" );
return $this->nextText();
} else {
$this->debug( "BaseDump::prefetch already past rev $rev on page $page [$this->lastPage, $this->lastRev]" );
return null;
}
}
function debug( $str ) {
wfDebug( $str . "\n" );
//global $dumper;
//$dumper->progress( $str );
}
/**
* @access private
*/
function nextPage() {
if( $this->skipTo( 'page', 'mediawiki' ) ) {
if( $this->skipTo( 'id' ) ) {
$this->lastPage = intval( $this->nodeContents() );
$this->lastRev = 0;
$this->atPageEnd = false;
}
} else {
$this->atEnd = true;
}
}
/**
* @access private
*/
function nextRev() {
if( $this->skipTo( 'revision' ) ) {
if( $this->skipTo( 'id' ) ) {
$this->lastRev = intval( $this->nodeContents() );
}
} else {
$this->atPageEnd = true;
}
}
/**
* @access private
*/
function nextText() {
$this->skipTo( 'text' );
return strval( $this->nodeContents() );
}
/**
* @access private
*/
function skipTo( $name, $parent='page' ) {
if( $this->atEnd ) {
return false;
}
while( $this->reader->read() ) {
if( $this->reader->nodeType == XMLREADER_ELEMENT &&
$this->reader->name == $name ) {
return true;
}
if( $this->reader->nodeType == XMLREADER_END_ELEMENT &&
$this->reader->name == $parent ) {
$this->debug( "BaseDump::skipTo found </$parent> searching for <$name>" );
return false;
}
}
return $this->close();
}
/**
* Shouldn't something like this be built-in to XMLReader?
* Fetches text contents of the current element, assuming
* no sub-elements or such scary things.
* @return string
* @access private
*/
function nodeContents() {
if( $this->atEnd ) {
return null;
}
if( $this->reader->isEmptyElement ) {
return "";
}
$buffer = "";
while( $this->reader->read() ) {
switch( $this->reader->nodeType ) {
case XMLREADER_TEXT:
// case XMLREADER_WHITESPACE:
case XMLREADER_SIGNIFICANT_WHITESPACE:
$buffer .= $this->reader->value;
break;
case XMLREADER_END_ELEMENT:
return $buffer;
}
}
return $this->close();
}
/**
* @access private
*/
function close() {
$this->reader->close();
$this->atEnd = true;
return null;
}
}
?>
|
