3 require_once( "$IP/maintenance/backup.inc" );
6 * Base TestCase for dumps
8 abstract class DumpTestCase
extends MediaWikiTestCase
{
11 * exception to be rethrown once in sound PHPUnit surrounding
13 * As the current MediaWikiTestCase::run is not robust enough to recover
14 * from thrown exceptions directly, we cannot throw frow within
15 * self::addDBData, although it would be appropriate. Hence, we catch the
16 * exception and store it until we are in setUp and may finally rethrow
17 * the exception without crashing the test suite.
21 protected $exceptionFromAddDBData = null;
24 * Holds the xmlreader used for analyzing an xml dump
28 protected $xml = null;
31 * Adds a revision to a page, while returning the resuting revision's id
33 * @param $page WikiPage: page to add the revision to
34 * @param $text string: revisions text
35 * @param $text string: revisions summare
39 protected function addRevision( Page
$page, $text, $summary ) {
40 $status = $page->doEdit( $text, $summary );
41 if ( $status->isGood() ) {
42 $value = $status->getValue();
43 $revision = $value['revision'];
44 $revision_id = $revision->getId();
45 $text_id = $revision->getTextId();
46 if ( ( $revision_id > 0 ) && ( $text_id > 0 ) ) {
47 return array( $revision_id, $text_id );
50 throw new MWException( "Could not determine revision id (" . $status->getXML() . ")" );
55 * gunzips the given file and stores the result in the original file name
57 * @param $fname string: filename to read the gzipped data from and stored
58 * the gunzipped data into
60 protected function gunzip( $fname ) {
61 $gzipped_contents = file_get_contents( $fname );
62 if ( $gzipped_contents === FALSE ) {
63 $this->fail( "Could not get contents of $fname" );
65 // We resort to use gzinflate instead of gzdecode, as gzdecode
66 // need not be available
67 $contents = gzinflate( substr( $gzipped_contents, 10, -8 ) );
68 $this->assertEquals( strlen( $contents ),
69 file_put_contents( $fname, $contents ), "# bytes written" );
73 * Default set up function.
75 * Clears $wgUser, and reports errors from addDBData to PHPUnit
77 protected function setUp() {
82 // Check if any Exception is stored for rethrowing from addDBData
83 // @see self::exceptionFromAddDBData
84 if ( $this->exceptionFromAddDBData
!== null ) {
85 throw $this->exceptionFromAddDBData
;
92 * Checks for test output consisting only of lines containing ETA announcements
94 function expectETAOutput() {
95 // Newer PHPUnits require assertion about the output using PHPUnit's own
96 // expectOutput[...] functions. However, the PHPUnit shipped prediactes
97 // do not allow to check /each/ line of the output using /readable/ REs.
100 // 1. ... add a dummy output checking to make PHPUnit not complain
101 // about unchecked test output
102 $this->expectOutputRegex( '//' );
104 // 2. Do the real output checking on our own.
105 $lines = explode( "\n", $this->getActualOutput() );
106 $this->assertGreaterThan( 1, count( $lines ), "Minimal lines of produced output" );
107 $this->assertEquals( '', array_pop( $lines ), "Output ends in LF" );
108 $timestamp_re = "[0-9]{4}-[01][0-9]-[0-3][0-9] [0-2][0-9]:[0-5][0-9]:[0-6][0-9]";
109 foreach ( $lines as $line ) {
110 $this->assertRegExp( "/$timestamp_re: .* \(ID [0-9]+\) [0-9]* pages .*, [0-9]* revs .*, ETA/", $line );
116 * Step the current XML reader until node end of given name is found.
118 * @param $name string: name of the closing element to look for
119 * (e.g.: "mediawiki" when looking for </mediawiki>)
121 * @return bool: true iff the end node could be found. false otherwise.
123 protected function skipToNodeEnd( $name ) {
124 while ( $this->xml
->read() ) {
125 if ( $this->xml
->nodeType
== XMLReader
::END_ELEMENT
&&
126 $this->xml
->name
== $name ) {
134 * Step the current XML reader to the first element start after the node
135 * end of a given name.
137 * @param $name string: name of the closing element to look for
138 * (e.g.: "mediawiki" when looking for </mediawiki>)
140 * @return bool: true iff new element after the closing of $name could be
141 * found. false otherwise.
143 protected function skipPastNodeEnd( $name ) {
144 $this->assertTrue( $this->skipToNodeEnd( $name ),
145 "Skipping to end of $name" );
146 while ( $this->xml
->read() ) {
147 if ( $this->xml
->nodeType
== XMLReader
::ELEMENT
) {
155 * Opens an XML file to analyze and optionally skips past siteinfo.
157 * @param $fname string: name of file to analyze
158 * @param $skip_siteinfo bool: (optional) If true, step the xml reader
159 * to the first element after </siteinfo>
161 protected function assertDumpStart( $fname, $skip_siteinfo = true ) {
162 $this->xml
= new XMLReader();
163 $this->assertTrue( $this->xml
->open( $fname ),
164 "Opening temporary file $fname via XMLReader failed" );
165 if ( $skip_siteinfo ) {
166 $this->assertTrue( $this->skipPastNodeEnd( "siteinfo" ),
167 "Skipping past end of siteinfo" );
172 * Asserts that the xml reader is at the final closing tag of an xml file and
175 * @param $tag string: (optional) the name of the final tag
176 * (e.g.: "mediawiki" for </mediawiki>)
178 protected function assertDumpEnd( $name = "mediawiki" ) {
179 $this->assertNodeEnd( $name, false );
180 if ( $this->xml
->read() ) {
181 $this->skipWhitespace();
183 $this->assertEquals( $this->xml
->nodeType
, XMLReader
::NONE
,
184 "No proper entity left to parse" );
189 * Steps the xml reader over white space
191 protected function skipWhitespace() {
193 while ( $cont && ( ( $this->xml
->nodeType
== XMLReader
::WHITESPACE
)
194 ||
( $this->xml
->nodeType
== XMLReader
::SIGNIFICANT_WHITESPACE
) ) ) {
195 $cont = $this->xml
->read();
200 * Asserts that the xml reader is at an element of given name, and optionally
203 * @param $name string: the name of the element to check for
204 * (e.g.: "mediawiki" for <mediawiki>)
205 * @param $skip bool: (optional) if true, skip past the found element
207 protected function assertNodeStart( $name, $skip = true ) {
208 $this->assertEquals( $name, $this->xml
->name
, "Node name" );
209 $this->assertEquals( XMLReader
::ELEMENT
, $this->xml
->nodeType
, "Node type" );
211 $this->assertTrue( $this->xml
->read(), "Skipping past start tag" );
216 * Asserts that the xml reader is at an closing element of given name, and optionally
219 * @param $name string: the name of the closing element to check for
220 * (e.g.: "mediawiki" for </mediawiki>)
221 * @param $skip bool: (optional) if true, skip past the found element
223 protected function assertNodeEnd( $name, $skip = true ) {
224 $this->assertEquals( $name, $this->xml
->name
, "Node name" );
225 $this->assertEquals( XMLReader
::END_ELEMENT
, $this->xml
->nodeType
, "Node type" );
227 $this->assertTrue( $this->xml
->read(), "Skipping past end tag" );
233 * Asserts that the xml reader is at an element of given tag that contains a given text,
234 * and skips over the element.
236 * @param $name string: the name of the element to check for
237 * (e.g.: "mediawiki" for <mediawiki>...</mediawiki>)
238 * @param $text string|false: If string, check if it equals the elements text.
239 * If false, ignore the element's text
240 * @param $skip_ws bool: (optional) if true, skip past white spaces that trail the
243 protected function assertTextNode( $name, $text, $skip_ws = true ) {
244 $this->assertNodeStart( $name );
246 if ( $text !== false ) {
247 $this->assertEquals( $text, $this->xml
->value
, "Text of node " . $name );
249 $this->assertTrue( $this->xml
->read(), "Skipping past processed text of " . $name );
250 $this->assertNodeEnd( $name );
253 $this->skipWhitespace();
258 * Asserts that the xml reader is at the start of a page element and skips over the first
259 * tags, after checking them.
261 * Besides the opening page element, this function also checks for and skips over the
262 * title, ns, and id tags. Hence after this function, the xml reader is at the first
263 * revision of the current page.
265 * @param $id int: id of the page to assert
266 * @param $ns int: number of namespage to assert
267 * @param $name string: title of the current page
269 protected function assertPageStart( $id, $ns, $name ) {
271 $this->assertNodeStart( "page" );
272 $this->skipWhitespace();
274 $this->assertTextNode( "title", $name );
275 $this->assertTextNode( "ns", $ns );
276 $this->assertTextNode( "id", $id );
281 * Asserts that the xml reader is at the page's closing element and skips to the next
284 protected function assertPageEnd() {
285 $this->assertNodeEnd( "page" );
286 $this->skipWhitespace();
290 * Asserts that the xml reader is at a revision and checks its representation before
293 * @param $id int: id of the revision
294 * @param $summary string: summary of the revision
295 * @param $text_id int: id of the revision's text
296 * @param $text_bytes int: # of bytes in the revision's text
297 * @param $text_sha1 string: the base36 SHA-1 of the revision's text
298 * @param $text string|false: (optional) The revision's string, or false to check for a
301 protected function assertRevision( $id, $summary, $text_id, $text_bytes, $text_sha1, $text = false ) {
303 $this->assertNodeStart( "revision" );
304 $this->skipWhitespace();
306 $this->assertTextNode( "id", $id );
307 $this->assertTextNode( "timestamp", false );
309 $this->assertNodeStart( "contributor" );
310 $this->skipWhitespace();
311 $this->assertTextNode( "ip", false );
312 $this->assertNodeEnd( "contributor" );
313 $this->skipWhitespace();
315 $this->assertTextNode( "comment", $summary );
317 $this->assertNodeStart( "text", false );
318 if ( $text_bytes !== false ) {
319 $this->assertEquals( $this->xml
->getAttribute( "bytes" ), $text_bytes,
320 "Attribute 'bytes' of revision " . $id );
324 if ( $text === false ) {
325 // Testing for a stub
326 $this->assertEquals( $this->xml
->getAttribute( "id" ), $text_id,
327 "Text id of revision " . $id );
328 $this->assertFalse( $this->xml
->hasValue
, "Revision has text" );
329 $this->assertTrue( $this->xml
->read(), "Skipping text start tag" );
330 if ( ( $this->xml
->nodeType
== XMLReader
::END_ELEMENT
)
331 && ( $this->xml
->name
== "text" ) ) {
335 $this->skipWhitespace();
337 // Testing for a real dump
338 $this->assertTrue( $this->xml
->read(), "Skipping text start tag" );
339 $this->assertEquals( $text, $this->xml
->value
, "Text of revision " . $id );
340 $this->assertTrue( $this->xml
->read(), "Skipping past text" );
341 $this->assertNodeEnd( "text" );
342 $this->skipWhitespace();
345 $this->assertTextNode( "sha1", $text_sha1 );
347 $this->assertNodeEnd( "revision" );
348 $this->skipWhitespace();