PHPUnit preparations for tests of maintenance
authorChristian Aistleitner <christian@quelltextlich.at>
Tue, 3 Apr 2012 10:23:03 +0000 (12:23 +0200)
committerChristian Aistleitner <christian@quelltextlich.at>
Tue, 3 Apr 2012 10:57:10 +0000 (12:57 +0200)
Change-Id: Id1ac2f8b940e5417e0facaacb7d5ec30b48a9efb

tests/TestsAutoLoader.php
tests/phpunit/StructureTest.php
tests/phpunit/maintenance/DumpTestCase.php [new file with mode: 0644]
tests/phpunit/suite.xml

index 0c97daa..8301558 100644 (file)
@@ -25,6 +25,9 @@ $wgAutoloadClasses += array(
        //Selenium
        'SeleniumTestConstants' => "$testFolder/selenium/SeleniumTestConstants.php",
 
+       //maintenance
+       'DumpTestCase' => "$testFolder/phpunit/maintenance/DumpTestCase.php",
+
        //Generic providers
        'MediaWikiProvide' => "$testFolder/phpunit/includes/Providers.php",
 );
index 6953ab6..17ea06c 100644 (file)
@@ -20,6 +20,7 @@ class StructureTest extends MediaWikiTestCase {
                        'MediaWikiLangTestCase',
                        'MediaWikiTestCase',
                        'PHPUnit_Framework_TestCase',
+                       'DumpTestCase',
                ) );
                $testClassRegex = "^class .* extends ($testClassRegex)";
                $finder = "find $rootPath -name '*.php' '!' -name '*Test.php'" .
diff --git a/tests/phpunit/maintenance/DumpTestCase.php b/tests/phpunit/maintenance/DumpTestCase.php
new file mode 100644 (file)
index 0000000..a6fe549
--- /dev/null
@@ -0,0 +1,359 @@
+<?php
+global $IP;
+require_once( "$IP/maintenance/backup.inc" );
+
+/**
+ * Base TestCase for dumps
+ */
+abstract class DumpTestCase extends MediaWikiTestCase {
+
+       /**
+        * exception to be rethrown once in sound PHPUnit surrounding
+        *
+        * As the current MediaWikiTestCase::run is not robust enough to recover
+        * from thrown exceptions directly, we cannot throw frow within
+        * self::addDBData, although it would be appropriate. Hence, we catch the
+        * exception and store it until we are in setUp and may finally rethrow
+        * the exception without crashing the test suite.
+        *
+        * @var Exception|null
+        */
+       protected $exceptionFromAddDBData = null;
+
+       /**
+        * Holds the xmlreader used for analyzing an xml dump
+        *
+        * @var XMLReader|null
+        */
+       protected $xml = null;
+
+       /**
+        * Adds a revision to a page, while returning the resuting revision's id
+        *
+        * @param $page WikiPage: page to add the revision to
+        * @param $text string: revisions text
+        * @param $text string: revisions summare
+        *
+        * @throws MWExcepion
+        */
+       protected function addRevision( Page $page, $text, $summary ) {
+               $status = $page->doEdit( $text, $summary );
+               if ( $status->isGood() ) {
+                       $value = $status->getValue();
+                       $revision = $value['revision'];
+                       $revision_id = $revision->getId();
+                       $text_id = $revision->getTextId();
+                       if ( ( $revision_id > 0 ) && ( $text_id > 0 ) ) {
+                               return array( $revision_id, $text_id );
+                       }
+               }
+               throw new MWException( "Could not determine revision id (" . $status->getXML() . ")" );
+       }
+
+
+       /**
+        * gunzips the given file and stores the result in the original file name
+        *
+        * @param $fname string: filename to read the gzipped data from and stored
+        *             the gunzipped data into
+        */
+       protected function gunzip( $fname ) {
+               $gzipped_contents = file_get_contents( $fname );
+               if ( $gzipped_contents === FALSE ) {
+                       $this->fail( "Could not get contents of $fname" );
+               }
+               // We resort to use gzinflate instead of gzdecode, as gzdecode
+               // need not be available
+               $contents = gzinflate( substr( $gzipped_contents, 10, -8 ) );
+               $this->assertEquals( strlen( $contents ),
+                       file_put_contents( $fname, $contents ), "# bytes written" );
+       }
+
+       /**
+        * obtains a new temporary file name
+        *
+        * The obtained filename is enlisted to be removed upon tearDown
+        *
+        * @returns string: absolute name of the temporary file
+        */
+       protected function getNewTempFile() {
+               $fname = tempnam( wfTempDir(), 'MW_PHPUnit_' . get_class( $this ) . '_' );
+               $this->tmpfiles[] = $fname;
+               return $fname;
+       }
+
+       /**
+        * Default set up function.
+        *
+        * Clears $wgUser, and reports errors from addDBData to PHPUnit
+        */
+       function setUp() {
+               global $wgUser;
+
+               parent::setUp();
+
+               // Check if any Exception is stored for rethrowing from addDBData
+               // @see self::exceptionFromAddDBData
+               if ( $this->exceptionFromAddDBData !== null ) {
+                       throw $this->exceptionFromAddDBData;
+               }
+               $this->tmpfiles = array();
+
+               $wgUser = new User();
+       }
+
+       /**
+        * Default tear down function
+        *
+        * Removes all files that have been allocated via self::getNewTempFile, even if
+        * they turn out to be (empty or non-empty) directories now.
+        */
+       function tearDown() {
+               foreach ( $this->tmpfiles as $fname ) {
+                       if ( is_file( $fname ) || ( is_link( $fname ) ) ) {
+                               unlink( $fname );
+                       } elseif ( is_dir( $fname ) ) {
+                               wfRecursiveRemoveDir( $fname );
+                       }
+               }
+               parent::tearDown();
+       }
+
+
+       /**
+        * Step the current XML reader until node end of given name is found.
+        *
+        * @param $name string: name of the closing element to look for
+        *           (e.g.: "mediawiki" when looking for </mediawiki>)
+        *
+        * @return bool: true iff the end node could be found. false otherwise.
+        */
+       protected function skipToNodeEnd( $name ) {
+               while ( $this->xml->read() ) {
+                       if ( $this->xml->nodeType == XMLReader::END_ELEMENT &&
+                               $this->xml->name == $name ) {
+                               return true;
+                       }
+               }
+               return false;
+       }
+
+       /**
+        * Step the current XML reader to the first element start after the node
+        * end of a given name.
+        *
+        * @param $name string: name of the closing element to look for
+        *           (e.g.: "mediawiki" when looking for </mediawiki>)
+        *
+        * @return bool: true iff new element after the closing of $name could be
+        *           found. false otherwise.
+        */
+       protected function skipPastNodeEnd( $name ) {
+               $this->assertTrue( $this->skipToNodeEnd( $name ),
+                       "Skipping to end of $name" );
+               while ( $this->xml->read() ) {
+                       if ( $this->xml->nodeType == XMLReader::ELEMENT ) {
+                               return true;
+                       }
+               }
+               return false;
+       }
+
+       /**
+        * Opens an XML file to analyze and optionally skips past siteinfo.
+        *
+        * @param $fname string: name of file to analyze
+        * @param $skip_siteinfo bool: (optional) If true, step the xml reader
+        *           to the first element after </siteinfo>
+        */
+       protected function assertDumpStart( $fname, $skip_siteinfo = true ) {
+               $this->xml = new XMLReader();
+               $this->assertTrue( $this->xml->open( $fname ),
+                       "Opening temporary file $fname via XMLReader failed" );
+               if ( $skip_siteinfo ) {
+                       $this->assertTrue( $this->skipPastNodeEnd( "siteinfo" ),
+                               "Skipping past end of siteinfo" );
+               }
+       }
+
+       /**
+        * Asserts that the xml reader is at the final closing tag of an xml file and
+        * closes the reader.
+        *
+        * @param $tag string: (optional) the name of the final tag
+        *           (e.g.: "mediawiki" for </mediawiki>)
+        */
+       protected function assertDumpEnd( $name = "mediawiki" ) {
+               $this->assertNodeEnd( $name, false );
+               if ( $this->xml->read() ) {
+                       $this->skipWhitespace();
+               }
+               $this->assertEquals( $this->xml->nodeType, XMLReader::NONE,
+                       "No proper entity left to parse" );
+               $this->xml->close();
+       }
+
+       /**
+        * Steps the xml reader over white space
+        */
+       protected function skipWhitespace() {
+               $cont = true;
+               while ( $cont && ( ( $this->xml->nodeType == XMLReader::WHITESPACE )
+                               || ( $this->xml->nodeType == XMLReader::SIGNIFICANT_WHITESPACE ) ) ) {
+                       $cont = $this->xml->read();
+               }
+       }
+
+       /**
+        * Asserts that the xml reader is at an element of given name, and optionally
+        * skips past it.
+        *
+        * @param $name string: the name of the element to check for
+        *           (e.g.: "mediawiki" for <mediawiki>)
+        * @param $skip bool: (optional) if true, skip past the found element
+        */
+       protected function assertNodeStart( $name, $skip = true ) {
+               $this->assertEquals( $name, $this->xml->name, "Node name" );
+               $this->assertEquals( XMLReader::ELEMENT, $this->xml->nodeType, "Node type" );
+               if ( $skip ) {
+                       $this->assertTrue( $this->xml->read(), "Skipping past start tag" );
+               }
+       }
+
+       /**
+        * Asserts that the xml reader is at an closing element of given name, and optionally
+        * skips past it.
+        *
+        * @param $name string: the name of the closing element to check for
+        *           (e.g.: "mediawiki" for </mediawiki>)
+        * @param $skip bool: (optional) if true, skip past the found element
+        */
+       protected function assertNodeEnd( $name, $skip = true ) {
+               $this->assertEquals( $name, $this->xml->name, "Node name" );
+               $this->assertEquals( XMLReader::END_ELEMENT, $this->xml->nodeType, "Node type" );
+               if ( $skip ) {
+                       $this->assertTrue( $this->xml->read(), "Skipping past end tag" );
+               }
+       }
+
+
+       /**
+        * Asserts that the xml reader is at an element of given tag that contains a given text,
+        * and skips over the element.
+        *
+        * @param $name string: the name of the element to check for
+        *           (e.g.: "mediawiki" for <mediawiki>...</mediawiki>)
+        * @param $text string|false: If string, check if it equals the elements text.
+        *           If false, ignore the element's text
+        * @param $skip_ws bool: (optional) if true, skip past white spaces that trail the
+        *           closing element.
+        */
+       protected function assertTextNode( $name, $text, $skip_ws = true ) {
+               $this->assertNodeStart( $name );
+
+               if ( $text !== false ) {
+                       $this->assertEquals( $text, $this->xml->value, "Text of node " . $name );
+               }
+               $this->assertTrue( $this->xml->read(), "Skipping past processed text of " . $name );
+               $this->assertNodeEnd( $name );
+
+               if ( $skip_ws ) {
+                       $this->skipWhitespace();
+               }
+       }
+
+       /**
+        * Asserts that the xml reader is at the start of a page element and skips over the first
+        * tags, after checking them.
+        *
+        * Besides the opening page element, this function also checks for and skips over the
+        * title, ns, and id tags. Hence after this function, the xml reader is at the first
+        * revision of the current page.
+        *
+        * @param $id int: id of the page to assert
+        * @param $ns int: number of namespage to assert
+        * @param $name string: title of the current page
+        */
+       protected function assertPageStart( $id, $ns, $name ) {
+
+               $this->assertNodeStart( "page" );
+               $this->skipWhitespace();
+
+               $this->assertTextNode( "title", $name );
+               $this->assertTextNode( "ns", $ns );
+               $this->assertTextNode( "id", $id );
+
+       }
+
+       /**
+        * Asserts that the xml reader is at the page's closing element and skips to the next
+        * element.
+        */
+       protected function assertPageEnd() {
+               $this->assertNodeEnd( "page" );
+               $this->skipWhitespace();
+       }
+
+       /**
+        * Asserts that the xml reader is at a revision and checks its representation before
+        * skipping over it.
+        *
+        * @param $id int: id of the revision
+        * @param $summary string: summary of the revision
+        * @param $text_id int: id of the revision's text
+        * @param $text_bytes int: # of bytes in the revision's text
+        * @param $text_sha1 string: the base36 SHA-1 of the revision's text
+        * @param $text string|false: (optional) The revision's string, or false to check for a
+        *            revision stub
+        */
+       protected function assertRevision( $id, $summary, $text_id, $text_bytes, $text_sha1, $text = false ) {
+
+               $this->assertNodeStart( "revision" );
+               $this->skipWhitespace();
+
+               $this->assertTextNode( "id", $id );
+               $this->assertTextNode( "timestamp", false );
+
+               $this->assertNodeStart( "contributor" );
+               $this->skipWhitespace();
+               $this->assertTextNode( "ip", false );
+               $this->assertNodeEnd( "contributor" );
+               $this->skipWhitespace();
+
+               $this->assertTextNode( "comment", $summary );
+
+               $this->assertNodeStart( "text", false );
+               if ( $text_bytes !== false ) {
+                       $this->assertEquals( $this->xml->getAttribute( "bytes" ), $text_bytes,
+                               "Attribute 'bytes' of revision " . $id );
+               }
+
+
+               if ( $text === false ) {
+                       // Testing for a stub
+                       $this->assertEquals( $this->xml->getAttribute( "id" ), $text_id,
+                               "Text id of revision " . $id );
+                       $this->assertFalse( $this->xml->hasValue, "Revision has text" );
+                       $this->assertTrue( $this->xml->read(), "Skipping text start tag" );
+                       if ( ( $this->xml->nodeType == XMLReader::END_ELEMENT )
+                               && ( $this->xml->name == "text" ) ) {
+
+                               $this->xml->read();
+                       }
+                       $this->skipWhitespace();
+               } else {
+                       // Testing for a real dump
+                       $this->assertTrue( $this->xml->read(), "Skipping text start tag" );
+                       $this->assertEquals( $text, $this->xml->value, "Text of revision " . $id );
+                       $this->assertTrue( $this->xml->read(), "Skipping past text" );
+                       $this->assertNodeEnd( "text" );
+                       $this->skipWhitespace();
+               }
+
+               $this->assertTextNode( "sha1", $text_sha1 );
+
+               $this->assertNodeEnd( "revision" );
+               $this->skipWhitespace();
+       }
+
+}
index 8c942a4..a03f392 100644 (file)
                <testsuite name="skins">
                        <directory>skins</directory>
                </testsuite>
+               <!-- As there is a class Maintenance, we cannot use the
+                    name "maintenance" directly -->
+               <testsuite name="maintenance_suite">
+                       <directory>maintenance</directory>
+               </testsuite>
                <testsuite name="structure">
                        <file>StructureTest.php</file>
                </testsuite>