<?php
-# See diff.doc
+/**
+ * See diff.doc
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
+
+/** */
+define( 'MAX_DIFF_LINE', 10000 );
+define( 'MAX_DIFF_XREF_LENGTH', 10000 );
+/**
+ * @todo document
+ * @public
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class DifferenceEngine {
- /* private */ var $mOldid, $mNewid;
- /* private */ var $mOldtitle, $mNewtitle;
- /* private */ var $mOldtext, $mNewtext;
- /* private */ var $mOldUser, $mNewUser;
- /* private */ var $mOldComment, $mNewComment;
- /* private */ var $mOldPage, $mNewPage;
-
- function DifferenceEngine( $old, $new )
- {
- $this->mOldid = $old;
- $this->mNewid = $new;
- }
-
- function showDiffPage()
- {
- global $wgUser, $wgTitle, $wgOut, $wgLang;
- $fname = "DifferenceEngine::showDiffPage";
+ /**#@+
+ * @private
+ */
+ var $mOldid, $mNewid, $mTitle;
+ var $mOldtitle, $mNewtitle, $mPagetitle;
+ var $mOldtext, $mNewtext;
+ var $mOldPage, $mNewPage;
+ var $mRcidMarkPatrolled;
+ var $mOldRev, $mNewRev;
+ var $mRevisionsLoaded = false; // Have the revisions been loaded
+ var $mTextLoaded = 0; // How many text blobs have been loaded, 0, 1 or 2?
+ /**#@-*/
+
+ /**
+ * Constructor
+ * @param $titleObj Title object that the diff is associated with
+ * @param $old Integer: old ID we want to show and diff with.
+ * @param $new String: either 'prev' or 'next'.
+ * @param $rcid Integer: ??? FIXME (default 0)
+ */
+ function DifferenceEngine( $titleObj = null, $old = 0, $new = 0, $rcid = 0 ) {
+ $this->mTitle = $titleObj;
+ wfDebug("DifferenceEngine old '$old' new '$new' rcid '$rcid'\n");
+
+ if ( 'prev' === $new ) {
+ # Show diff between revision $old and the previous one.
+ # Get previous one from DB.
+ #
+ $this->mNewid = intval($old);
+
+ $this->mOldid = $this->mTitle->getPreviousRevisionID( $this->mNewid );
+
+ } elseif ( 'next' === $new ) {
+ # Show diff between revision $old and the previous one.
+ # Get previous one from DB.
+ #
+ $this->mOldid = intval($old);
+ $this->mNewid = $this->mTitle->getNextRevisionID( $this->mOldid );
+ if ( false === $this->mNewid ) {
+ # if no result, NewId points to the newest old revision. The only newer
+ # revision is cur, which is "0".
+ $this->mNewid = 0;
+ }
+
+ } else {
+ $this->mOldid = intval($old);
+ $this->mNewid = intval($new);
+ }
+ $this->mRcidMarkPatrolled = intval($rcid); # force it to be an integer
+ }
+
+ function showDiffPage() {
+ global $wgUser, $wgOut, $wgContLang, $wgUseExternalEditor, $wgUseRCPatrol;
+ $fname = 'DifferenceEngine::showDiffPage';
wfProfileIn( $fname );
-
- $t = $wgTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
+
+ # If external diffs are enabled both globally and for the user,
+ # we'll use the application/x-external-editor interface to call
+ # an external diff tool like kompare, kdiff3, etc.
+ if($wgUseExternalEditor && $wgUser->getOption('externaldiff')) {
+ global $wgInputEncoding,$wgServer,$wgScript,$wgLang;
+ $wgOut->disable();
+ header ( "Content-type: application/x-external-editor; charset=".$wgInputEncoding );
+ $url1=$this->mTitle->getFullURL("action=raw&oldid=".$this->mOldid);
+ $url2=$this->mTitle->getFullURL("action=raw&oldid=".$this->mNewid);
+ $special=$wgLang->getNsText(NS_SPECIAL);
+ $control=<<<CONTROL
+[Process]
+Type=Diff text
+Engine=MediaWiki
+Script={$wgServer}{$wgScript}
+Special namespace={$special}
+
+[File]
+Extension=wiki
+URL=$url1
+
+[File 2]
+Extension=wiki
+URL=$url2
+CONTROL;
+ echo($control);
+ return;
+ }
+
+ $t = $this->mTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
"{$this->mNewid})";
- $mtext = wfMsg( "missingarticle", $t );
+ $mtext = wfMsg( 'missingarticle', "<nowiki>$t</nowiki>" );
$wgOut->setArticleFlag( false );
- if ( ! $this->loadText() ) {
- $wgOut->setPagetitle( wfMsg( "errorpagetitle" ) );
- $wgOut->addHTML( $mtext );
+ if ( ! $this->loadRevisionData() ) {
+ $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) );
+ $wgOut->addWikitext( $mtext );
wfProfileOut( $fname );
return;
}
- $wgOut->suppressQuickbar();
+ wfRunHooks( 'DiffViewHeader', array( $this, $this->mOldRev, $this->mNewRev ) );
+
+ if ( $this->mNewRev->isCurrent() ) {
+ $wgOut->setArticleFlag( true );
+ }
+
+ # mOldid is false if the difference engine is called with a "vague" query for
+ # a diff between a version V and its previous version V' AND the version V
+ # is the first version of that article. In that case, V' does not exist.
+ if ( $this->mOldid === false ) {
+ $this->showFirstRevision();
+ wfProfileOut( $fname );
+ return;
+ }
+
+ $wgOut->suppressQuickbar();
+
$oldTitle = $this->mOldPage->getPrefixedText();
$newTitle = $this->mNewPage->getPrefixedText();
if( $oldTitle == $newTitle ) {
$wgOut->setPageTitle( $newTitle );
} else {
- $wgOut->setPageTitle( $oldTitle . ", " . $newTitle );
+ $wgOut->setPageTitle( $oldTitle . ', ' . $newTitle );
}
- $wgOut->setSubtitle( wfMsg( "difference" ) );
- $wgOut->setRobotpolicy( "noindex,follow" );
-
+ $wgOut->setSubtitle( wfMsg( 'difference' ) );
+ $wgOut->setRobotpolicy( 'noindex,nofollow' );
+
if ( !( $this->mOldPage->userCanRead() && $this->mNewPage->userCanRead() ) ) {
$wgOut->loginToUse();
$wgOut->output();
}
$sk = $wgUser->getSkin();
- $talk = $wgLang->getNsText( NS_TALK );
- $contribs = wfMsg( "contribslink" );
+ $talk = $wgContLang->getNsText( NS_TALK );
+ $contribs = wfMsg( 'contribslink' );
+
+ if ( $this->mNewRev->isCurrent() && $wgUser->isAllowed('rollback') ) {
+ $username = $this->mNewRev->getUserText();
+ $rollback = ' <strong>[' . $sk->makeKnownLinkObj( $this->mTitle, wfMsg( 'rollbacklink' ),
+ 'action=rollback&from=' . urlencode( $username ) .
+ '&token=' . urlencode( $wgUser->editToken( array( $this->mTitle->getPrefixedText(), $username ) ) ) ) .
+ ']</strong>';
+ } else {
+ $rollback = '';
+ }
+ if( $wgUseRCPatrol && $this->mRcidMarkPatrolled != 0 && $wgUser->isAllowed( 'patrol' ) ) {
+ $patrol = ' [' . $sk->makeKnownLinkObj( $this->mTitle, wfMsg( 'markaspatrolleddiff' ), "action=markpatrolled&rcid={$this->mRcidMarkPatrolled}" ) . ']';
+ } else {
+ $patrol = '';
+ }
+
+ $prevlink = $sk->makeKnownLinkObj( $this->mTitle, wfMsgHtml( 'previousdiff' ),
+ 'diff=prev&oldid='.$this->mOldid, '', '', 'id="differences-prevlink"' );
+ if ( $this->mNewRev->isCurrent() ) {
+ $nextlink = '';
+ } else {
+ $nextlink = $sk->makeKnownLinkObj( $this->mTitle, wfMsgHtml( 'nextdiff' ),
+ 'diff=next&oldid='.$this->mNewid, '', '', 'id="differences-nextlink"' );
+ }
+ $oldHeader = "<strong>{$this->mOldtitle}</strong><br />" .
+ $sk->revUserTools( $this->mOldRev ) . "<br />" .
+ $sk->revComment( $this->mOldRev ) . "<br />" .
+ $prevlink;
+ $newHeader = "<strong>{$this->mNewtitle}</strong><br />" .
+ $sk->revUserTools( $this->mNewRev ) . " $rollback<br />" .
+ $sk->revComment( $this->mNewRev ) . "<br />" .
+ $nextlink . $patrol;
- $this->mOldComment = $sk->formatComment($this->mOldComment);
- $this->mNewComment = $sk->formatComment($this->mNewComment);
+ $this->showDiff( $oldHeader, $newHeader );
+ $wgOut->addHTML( "<hr /><h2>{$this->mPagetitle}</h2>\n" );
+ if( !$this->mNewRev->isCurrent() ) {
+ $oldEditSectionSetting = $wgOut->mParserOptions->setEditSection( false );
+ }
+
+ $this->loadNewText();
+ if( is_object( $this->mNewRev ) ) {
+ $wgOut->setRevisionId( $this->mNewRev->getId() );
+ }
+ $wgOut->addSecondaryWikiText( $this->mNewtext );
+
+ if( !$this->mNewRev->isCurrent() ) {
+ $wgOut->mParserOptions->setEditSection( $oldEditSectionSetting );
+ }
+
+ wfProfileOut( $fname );
+ }
+
+ /**
+ * Show the first revision of an article. Uses normal diff headers in
+ * contrast to normal "old revision" display style.
+ */
+ function showFirstRevision() {
+ global $wgOut, $wgUser;
+
+ $fname = 'DifferenceEngine::showFirstRevision';
+ wfProfileIn( $fname );
+
+ # Get article text from the DB
+ #
+ if ( ! $this->loadNewText() ) {
+ $t = $this->mTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
+ "{$this->mNewid})";
+ $mtext = wfMsg( 'missingarticle', "<nowiki>$t</nowiki>" );
+ $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) );
+ $wgOut->addWikitext( $mtext );
+ wfProfileOut( $fname );
+ return;
+ }
+ if ( $this->mNewRev->isCurrent() ) {
+ $wgOut->setArticleFlag( true );
+ }
+
+ # Check if user is allowed to look at this page. If not, bail out.
+ #
+ if ( !( $this->mTitle->userCanRead() ) ) {
+ $wgOut->loginToUse();
+ $wgOut->output();
+ wfProfileOut( $fname );
+ exit;
+ }
+
+ # Prepare the header box
+ #
+ $sk = $wgUser->getSkin();
+
+ $nextlink = $sk->makeKnownLinkObj( $this->mTitle, wfMsgHtml( 'nextdiff' ), 'diff=next&oldid='.$this->mNewid, '', '', 'id="differences-nextlink"' );
+ $header = "<div class=\"firstrevisionheader\" style=\"text-align: center\"><strong>{$this->mOldtitle}</strong><br />" .
+ $sk->revUserTools( $this->mNewRev ) . "<br />" .
+ $sk->revComment( $this->mNewRev ) . "<br />" .
+ $nextlink . "</div>\n";
+
+ $wgOut->addHTML( $header );
+
+ $wgOut->setSubtitle( wfMsg( 'difference' ) );
+ $wgOut->setRobotpolicy( 'noindex,nofollow' );
+
+
+ # Show current revision
+ #
+ $wgOut->addHTML( "<hr /><h2>{$this->mPagetitle}</h2>\n" );
+ if( is_object( $this->mNewRev ) ) {
+ $wgOut->setRevisionId( $this->mNewRev->getId() );
+ }
+ $wgOut->addSecondaryWikiText( $this->mNewtext );
+
+ wfProfileOut( $fname );
+ }
+
+ /**
+ * Get the diff text, send it to $wgOut
+ * Returns false if the diff could not be generated, otherwise returns true
+ */
+ function showDiff( $otitle, $ntitle ) {
+ global $wgOut;
+ $diff = $this->getDiff( $otitle, $ntitle );
+ if ( $diff === false ) {
+ $wgOut->addWikitext( wfMsg( 'missingarticle', "<nowiki>(fixme, bug)</nowiki>" ) );
+ return false;
+ } else {
+ $wgOut->addHTML( $diff );
+ return true;
+ }
+ }
- $oldUserLink = $sk->makeLinkObj( Title::makeTitle( NS_USER, $this->mOldUser ), $this->mOldUser );
- $newUserLink = $sk->makeLinkObj( Title::makeTitle( NS_USER, $this->mNewUser ), $this->mNewUser );
- $oldUTLink = $sk->makeLinkObj( Title::makeTitle( NS_USER_TALK, $this->mOldUser ), $talk );
- $newUTLink = $sk->makeLinkObj( Title::makeTitle( NS_USER_TALK, $this->mNewUser ), $talk );
- $oldContribs = $sk->makeKnownLinkObj( Title::makeTitle( NS_SPECIAL, "Contributions" ), $contribs,
- "target=" . urlencode($this->mOldUser) );
- $newContribs = $sk->makeKnownLinkObj( Title::makeTitle( NS_SPECIAL, "Contributions" ), $contribs,
- "target=" . urlencode($this->mNewUser) );
- if ( !$this->mNewid && $wgUser->isSysop() ) {
- $rollback = " <strong>[" . $sk->makeKnownLinkObj( $wgTitle, wfMsg( "rollbacklink" ),
- "action=rollback&from=" . urlencode($this->mNewUser) ) . "]</strong>";
+ /**
+ * Get diff table, including header
+ * Note that the interface has changed, it's no longer static.
+ * Returns false on error
+ */
+ function getDiff( $otitle, $ntitle ) {
+ $body = $this->getDiffBody();
+ if ( $body === false ) {
+ return false;
} else {
- $rollback = "";
+ return $this->addHeader( $body, $otitle, $ntitle );
+ }
+ }
+
+ /**
+ * Get the diff table body, without header
+ * Results are cached
+ * Returns false on error
+ */
+ function getDiffBody() {
+ global $wgMemc, $wgDBname;
+ $fname = 'DifferenceEngine::getDiffBody';
+ wfProfileIn( $fname );
+
+ // Cacheable?
+ $key = false;
+ if ( $this->mOldid && $this->mNewid ) {
+ // Try cache
+ $key = "$wgDBname:diff:oldid:{$this->mOldid}:newid:{$this->mNewid}";
+ $difftext = $wgMemc->get( $key );
+ if ( $difftext ) {
+ wfIncrStats( 'diff_cache_hit' );
+ $difftext = $this->localiseLineNumbers( $difftext );
+ $difftext .= "\n<!-- diff cache key $key -->\n";
+ wfProfileOut( $fname );
+ return $difftext;
+ }
}
- $oldHeader = "<strong>{$this->mOldtitle}</strong><br />$oldUserLink ($oldUTLink | $oldContribs)<br />".
- $this->mOldComment;
- $newHeader = "<strong>{$this->mNewtitle}</strong><br />$newUserLink ($newUTLink | $newContribs) $rollback<br />".
- $this->mNewComment;
+ if ( !$this->loadText() ) {
+ wfProfileOut( $fname );
+ return false;
+ }
- DifferenceEngine::showDiff( $this->mOldtext, $this->mNewtext,
- $oldHeader, $newHeader );
- $wgOut->addHTML( "<hr /><h2>{$this->mNewtitle}</h2>\n" );
- $wgOut->addWikiText( $this->mNewtext );
+ $difftext = $this->generateDiffBody( $this->mOldtext, $this->mNewtext );
+ // Save to cache for 7 days
+ if ( $key !== false && $difftext !== false ) {
+ wfIncrStats( 'diff_cache_miss' );
+ $wgMemc->set( $key, $difftext, 7*86400 );
+ } else {
+ wfIncrStats( 'diff_uncacheable' );
+ }
+ // Replace line numbers with the text in the user's language
+ if ( $difftext !== false ) {
+ $difftext = $this->localiseLineNumbers( $difftext );
+ }
wfProfileOut( $fname );
+ return $difftext;
}
- function showDiff( $otext, $ntext, $otitle, $ntitle )
- {
- global $wgOut;
+ /**
+ * Generate a diff, no caching
+ * $otext and $ntext must be already segmented
+ */
+ function generateDiffBody( $otext, $ntext ) {
+ global $wgExternalDiffEngine, $wgContLang;
+ $fname = 'DifferenceEngine::generateDiffBody';
- $ota = explode( "\n", str_replace( "\r\n", "\n",
- htmlspecialchars( $otext ) ) );
- $nta = explode( "\n", str_replace( "\r\n", "\n",
- htmlspecialchars( $ntext ) ) );
-
- $wgOut->addHTML( "<table border='0' width='98%'
-cellpadding='0' cellspacing='4px' class='diff'><tr>
-<td colspan='2' width='50%' align='center' class='diff-otitle'>
-{$otitle}</td>
-<td colspan='2' width='50%' align='center' class='diff-ntitle'>
-{$ntitle}</td>
-</tr>\n" );
-
- $diffs = new Diff( $ota, $nta );
- $formatter = new TableDiffFormatter();
- $formatter->format( $diffs );
- $wgOut->addHTML( "</table>\n" );
- }
-
- # Load the text of the articles to compare. If newid is 0, then compare
- # the old article in oldid to the current article; if oldid is 0, then
- # compare the current article to the immediately previous one (ignoring
- # the value of newid).
- #
- function loadText()
- {
- global $wgTitle, $wgOut, $wgLang, $wgIsMySQL, $wgIsPg;
- $fname = "DifferenceEngine::loadText";
+ $otext = str_replace( "\r\n", "\n", $otext );
+ $ntext = str_replace( "\r\n", "\n", $ntext );
- $oldtable=$wgIsPg?'"old"':'old';
- if ( 0 == $this->mNewid || 0 == $this->mOldid ) {
- $wgOut->setArticleFlag( true );
- $this->mNewtitle = wfMsg( "currentrev" );
- $id = $wgTitle->getArticleID();
-
- $sql = "SELECT cur_text, cur_user_text, cur_comment FROM cur WHERE cur_id={$id}";
- $res = wfQuery( $sql, DB_READ, $fname );
- if ( 0 == wfNumRows( $res ) ) { return false; }
-
- $s = wfFetchObject( $res );
- $this->mNewPage = &$wgTitle;
- $this->mNewtext = $s->cur_text;
- $this->mNewUser = $s->cur_user_text;
- $this->mNewComment = $s->cur_comment;
+ if ( $wgExternalDiffEngine == 'wikidiff' ) {
+ # For historical reasons, external diff engine expects
+ # input text to be HTML-escaped already
+ $otext = htmlspecialchars ( $wgContLang->segmentForDiff( $otext ) );
+ $ntext = htmlspecialchars ( $wgContLang->segmentForDiff( $ntext ) );
+ if( !function_exists( 'wikidiff_do_diff' ) ) {
+ dl('php_wikidiff.so');
+ }
+ return $wgContLang->unsegementForDiff( wikidiff_do_diff( $otext, $ntext, 2 ) );
+ }
+
+ if ( $wgExternalDiffEngine == 'wikidiff2' ) {
+ # Better external diff engine, the 2 may some day be dropped
+ # This one does the escaping and segmenting itself
+ if ( !function_exists( 'wikidiff2_do_diff' ) ) {
+ wfProfileIn( "$fname-dl" );
+ @dl('php_wikidiff2.so');
+ wfProfileOut( "$fname-dl" );
+ }
+ if ( function_exists( 'wikidiff2_do_diff' ) ) {
+ wfProfileIn( 'wikidiff2_do_diff' );
+ $text = wikidiff2_do_diff( $otext, $ntext, 2 );
+ wfProfileOut( 'wikidiff2_do_diff' );
+ return $text;
+ }
+ }
+ if ( $wgExternalDiffEngine !== false ) {
+ # Diff via the shell
+ global $wgTmpDirectory;
+ $tempName1 = tempnam( $wgTmpDirectory, 'diff_' );
+ $tempName2 = tempnam( $wgTmpDirectory, 'diff_' );
+
+ $tempFile1 = fopen( $tempName1, "w" );
+ if ( !$tempFile1 ) {
+ wfProfileOut( $fname );
+ return false;
+ }
+ $tempFile2 = fopen( $tempName2, "w" );
+ if ( !$tempFile2 ) {
+ wfProfileOut( $fname );
+ return false;
+ }
+ fwrite( $tempFile1, $otext );
+ fwrite( $tempFile2, $ntext );
+ fclose( $tempFile1 );
+ fclose( $tempFile2 );
+ $cmd = wfEscapeShellArg( $wgExternalDiffEngine, $tempName1, $tempName2 );
+ wfProfileIn( "$fname-shellexec" );
+ $difftext = wfShellExec( $cmd );
+ wfProfileOut( "$fname-shellexec" );
+ unlink( $tempName1 );
+ unlink( $tempName2 );
+ return $difftext;
+ }
+
+ # Native PHP diff
+ $ota = explode( "\n", $wgContLang->segmentForDiff( $otext ) );
+ $nta = explode( "\n", $wgContLang->segmentForDiff( $ntext ) );
+ $diffs =& new Diff( $ota, $nta );
+ $formatter =& new TableDiffFormatter();
+ return $wgContLang->unsegmentForDiff( $formatter->format( $diffs ) );
+ }
+
+
+ /**
+ * Replace line numbers with the text in the user's language
+ */
+ function localiseLineNumbers( $text ) {
+ return preg_replace_callback( '/<!--LINE (\d+)-->/',
+ array( &$this, 'localiseLineNumbersCb' ), $text );
+ }
+
+ function localiseLineNumbersCb( $matches ) {
+ global $wgLang;
+ return wfMsgExt( 'lineno', array('parseinline'), $wgLang->formatNum( $matches[1] ) );
+ }
+
+ /**
+ * Add the header to a diff body
+ */
+ function addHeader( $diff, $otitle, $ntitle ) {
+ $out = "
+ <table border='0' width='98%' cellpadding='0' cellspacing='4' class='diff'>
+ <tr>
+ <td colspan='2' width='50%' align='center' class='diff-otitle'>{$otitle}</td>
+ <td colspan='2' width='50%' align='center' class='diff-ntitle'>{$ntitle}</td>
+ </tr>
+ $diff
+ </table>
+ ";
+ return $out;
+ }
+
+ /**
+ * Use specified text instead of loading from the database
+ */
+ function setText( $oldText, $newText ) {
+ $this->mOldtext = $oldText;
+ $this->mNewtext = $newText;
+ $this->mTextLoaded = 2;
+ }
+
+ /**
+ * Load revision metadata for the specified articles. If newid is 0, then compare
+ * the old article in oldid to the current article; if oldid is 0, then
+ * compare the current article to the immediately previous one (ignoring the
+ * value of newid).
+ *
+ * If oldid is false, leave the corresponding revision object set
+ * to false. This is impossible via ordinary user input, and is provided for
+ * API convenience.
+ */
+ function loadRevisionData() {
+ global $wgLang;
+ if ( $this->mRevisionsLoaded ) {
+ return true;
} else {
- $sql = "SELECT old_namespace,old_title,old_timestamp,old_text,old_flags,old_user_text,old_comment FROM $oldtable WHERE " .
- "old_id={$this->mNewid}";
-
- $res = wfQuery( $sql, DB_READ, $fname );
- if ( 0 == wfNumRows( $res ) ) { return false; }
-
- $s = wfFetchObject( $res );
- $this->mNewtext = Article::getRevisionText( $s );
-
- $t = $wgLang->timeanddate( $s->old_timestamp, true );
- $this->mNewPage = Title::MakeTitle( $s->old_namespace, $s->old_title );
- $this->mNewtitle = wfMsg( "revisionasof", $t );
- $this->mNewUser = $s->old_user_text;
- $this->mNewComment = $s->old_comment;
- }
- if ( 0 == $this->mOldid ) {
- $use_index=$wgIsMySQL?"USE INDEX (name_title_timestamp)":"";
- $sql = "SELECT old_namespace,old_title,old_timestamp,old_text,old_flags,old_user_text,old_comment " .
- "FROM $oldtable $use_index WHERE " .
- "old_namespace=" . $this->mNewPage->getNamespace() . " AND " .
- "old_title='" . wfStrencode( $this->mNewPage->getDBkey() ) .
- "' ORDER BY inverse_timestamp LIMIT 1";
- $res = wfQuery( $sql, DB_READ, $fname );
+ // Whether it succeeds or fails, we don't want to try again
+ $this->mRevisionsLoaded = true;
+ }
+
+ // Load the new revision object
+ if( $this->mNewid ) {
+ $this->mNewRev = Revision::newFromId( $this->mNewid );
} else {
- $sql = "SELECT old_namespace,old_title,old_timestamp,old_text,old_flags,old_user_text,old_comment FROM $oldtable WHERE " .
- "old_id={$this->mOldid}";
- $res = wfQuery( $sql, DB_READ, $fname );
+ $this->mNewRev = Revision::newFromTitle( $this->mTitle );
}
- if ( 0 == wfNumRows( $res ) ) { return false; }
- $s = wfFetchObject( $res );
- $this->mOldPage = Title::MakeTitle( $s->old_namespace, $s->old_title );
- $this->mOldtext = Article::getRevisionText( $s );
+ if( is_null( $this->mNewRev ) ) {
+ return false;
+ }
- $t = $wgLang->timeanddate( $s->old_timestamp, true );
- $this->mOldtitle = wfMsg( "revisionasof", $t );
- $this->mOldUser = $s->old_user_text;
- $this->mOldComment = $s->old_comment;
-
+ // Set assorted variables
+ if( $this->mNewRev->isCurrent() ) {
+ $this->mNewPage = $this->mTitle;
+ $newLink = $this->mNewPage->escapeLocalUrl();
+ $this->mPagetitle = htmlspecialchars( wfMsg( 'currentrev' ) );
+ $newEdit = $this->mNewPage->escapeLocalUrl( 'action=edit' );
+ } else {
+ $this->mNewPage = $this->mNewRev->getTitle();
+ $newLink = $this->mNewPage->escapeLocalUrl( 'oldid=' . $this->mNewid );
+ $newEdit = $this->mNewPage->escapeLocalUrl( 'action=edit&oldid=' . $this->mNewid );
+ $t = $wgLang->timeanddate( $this->mNewRev->getTimestamp(), true );
+ $this->mPagetitle = htmlspecialchars( wfMsg( 'revisionasof', $t ) );
+ }
+ $this->mNewtitle = "<strong><a href='$newLink'>{$this->mPagetitle}</a></strong>"
+ . " (<a href='$newEdit'>" . htmlspecialchars( wfMsg( 'editold' ) ) . "</a>)";
+
+ // Load the old revision object
+ $this->mOldRev = false;
+ if( $this->mOldid ) {
+ $this->mOldRev = Revision::newFromId( $this->mOldid );
+ } elseif ( $this->mOldid === 0 ) {
+ $rev = $this->mNewRev->getPrevious();
+ if( $rev ) {
+ $this->mOldid = $rev->getId();
+ $this->mOldRev = $rev;
+ } else {
+ // No previous revision; mark to show as first-version only.
+ $this->mOldid = false;
+ $this->mOldRev = false;
+ }
+ }/* elseif ( $this->mOldid === false ) leave mOldRev false; */
+
+ if( is_null( $this->mOldRev ) ) {
+ return false;
+ }
+
+ if ( $this->mOldRev ) {
+ $this->mOldPage = $this->mOldRev->getTitle();
+
+ $t = $wgLang->timeanddate( $this->mOldRev->getTimestamp(), true );
+ $oldLink = $this->mOldPage->escapeLocalUrl( 'oldid=' . $this->mOldid );
+ $oldEdit = $this->mOldPage->escapeLocalUrl( 'action=edit&oldid=' . $this->mOldid );
+ $this->mOldtitle = "<strong><a href='$oldLink'>" . htmlspecialchars( wfMsg( 'revisionasof', $t ) )
+ . "</a></strong> (<a href='$oldEdit'>" . htmlspecialchars( wfMsg( 'editold' ) ) . "</a>)";
+ }
+
+ return true;
+ }
+
+ /**
+ * Load the text of the revisions, as well as revision data.
+ */
+ function loadText() {
+ if ( $this->mTextLoaded == 2 ) {
+ return true;
+ } else {
+ // Whether it succeeds or fails, we don't want to try again
+ $this->mTextLoaded = 2;
+ }
+
+ if ( !$this->loadRevisionData() ) {
+ return false;
+ }
+ if ( $this->mOldRev ) {
+ // FIXME: permission tests
+ $this->mOldtext = $this->mOldRev->getText();
+ if ( $this->mOldtext === false ) {
+ return false;
+ }
+ }
+ if ( $this->mNewRev ) {
+ $this->mNewtext = $this->mNewRev->getText();
+ if ( $this->mNewtext === false ) {
+ return false;
+ }
+ }
+ return true;
+ }
+
+ /**
+ * Load the text of the new revision, not the old one
+ */
+ function loadNewText() {
+ if ( $this->mTextLoaded >= 1 ) {
+ return true;
+ } else {
+ $this->mTextLoaded = 1;
+ }
+ if ( !$this->loadRevisionData() ) {
+ return false;
+ }
+ $this->mNewtext = $this->mNewRev->getText();
return true;
}
+
+
}
// A PHP diff engine for phpwiki. (Taken from phpwiki-1.3.3)
define('USE_ASSERTS', function_exists('assert'));
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _DiffOp {
var $type;
var $orig;
var $closing;
-
+
function reverse() {
- trigger_error("pure virtual", E_USER_ERROR);
+ trigger_error('pure virtual', E_USER_ERROR);
}
function norig() {
}
}
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _DiffOp_Copy extends _DiffOp {
var $type = 'copy';
-
+
function _DiffOp_Copy ($orig, $closing = false) {
if (!is_array($closing))
$closing = $orig;
}
}
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _DiffOp_Delete extends _DiffOp {
var $type = 'delete';
-
+
function _DiffOp_Delete ($lines) {
$this->orig = $lines;
$this->closing = false;
}
}
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _DiffOp_Add extends _DiffOp {
var $type = 'add';
-
+
function _DiffOp_Add ($lines) {
$this->closing = $lines;
$this->orig = false;
}
}
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _DiffOp_Change extends _DiffOp {
var $type = 'change';
-
+
function _DiffOp_Change ($orig, $closing) {
$this->orig = $orig;
$this->closing = $closing;
return new _DiffOp_Change($this->closing, $this->orig);
}
}
-
-
+
+
/**
* Class used internally by Diff to actually compute the diffs.
*
* closingly, some ideas (subdivision by NCHUNKS > 2, and some optimizations)
* are my own.
*
- * @author Geoffrey T. Dairiki
- * @access private
+ * Line length limits for robustness added by Tim Starling, 2005-08-31
+ *
+ * @author Geoffrey T. Dairiki, Tim Starling
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
*/
class _DiffEngine
{
function diff ($from_lines, $to_lines) {
+ $fname = '_DiffEngine::diff';
+ wfProfileIn( $fname );
+
$n_from = sizeof($from_lines);
$n_to = sizeof($to_lines);
unset($this->seq);
unset($this->in_seq);
unset($this->lcs);
-
+
// Skip leading common lines.
for ($skip = 0; $skip < $n_from && $skip < $n_to; $skip++) {
- if ($from_lines[$skip] != $to_lines[$skip])
+ if ($from_lines[$skip] !== $to_lines[$skip])
break;
$this->xchanged[$skip] = $this->ychanged[$skip] = false;
}
// Skip trailing common lines.
$xi = $n_from; $yi = $n_to;
for ($endskip = 0; --$xi > $skip && --$yi > $skip; $endskip++) {
- if ($from_lines[$xi] != $to_lines[$yi])
+ if ($from_lines[$xi] !== $to_lines[$yi])
break;
$this->xchanged[$xi] = $this->ychanged[$yi] = false;
}
-
+
// Ignore lines which do not exist in both files.
- for ($xi = $skip; $xi < $n_from - $endskip; $xi++)
- $xhash[$from_lines[$xi]] = 1;
+ for ($xi = $skip; $xi < $n_from - $endskip; $xi++) {
+ $xhash[$this->_line_hash($from_lines[$xi])] = 1;
+ }
+
for ($yi = $skip; $yi < $n_to - $endskip; $yi++) {
$line = $to_lines[$yi];
- if ( ($this->ychanged[$yi] = empty($xhash[$line])) )
+ if ( ($this->ychanged[$yi] = empty($xhash[$this->_line_hash($line)])) )
continue;
- $yhash[$line] = 1;
+ $yhash[$this->_line_hash($line)] = 1;
$this->yv[] = $line;
$this->yind[] = $yi;
}
for ($xi = $skip; $xi < $n_from - $endskip; $xi++) {
$line = $from_lines[$xi];
- if ( ($this->xchanged[$xi] = empty($yhash[$line])) )
+ if ( ($this->xchanged[$xi] = empty($yhash[$this->_line_hash($line)])) )
continue;
$this->xv[] = $line;
$this->xind[] = $xi;
$add = array();
while ($yi < $n_to && $this->ychanged[$yi])
$add[] = $to_lines[$yi++];
-
+
if ($delete && $add)
$edits[] = new _DiffOp_Change($delete, $add);
elseif ($delete)
elseif ($add)
$edits[] = new _DiffOp_Add($add);
}
+ wfProfileOut( $fname );
return $edits;
}
-
+
+ /**
+ * Returns the whole line if it's small enough, or the MD5 hash otherwise
+ */
+ function _line_hash( $line ) {
+ if ( strlen( $line ) > MAX_DIFF_XREF_LENGTH ) {
+ return md5( $line );
+ } else {
+ return $line;
+ }
+ }
+
/* Divide the Largest Common Subsequence (LCS) of the sequences
* [XOFF, XLIM) and [YOFF, YLIM) into NCHUNKS approximately equally
* of the portions it is going to specify.
*/
function _diag ($xoff, $xlim, $yoff, $ylim, $nchunks) {
- $flip = false;
-
- if ($xlim - $xoff > $ylim - $yoff) {
- // Things seems faster (I'm not sure I understand why)
- // when the shortest sequence in X.
- $flip = true;
- list ($xoff, $xlim, $yoff, $ylim)
- = array( $yoff, $ylim, $xoff, $xlim);
- }
-
- if ($flip)
- for ($i = $ylim - 1; $i >= $yoff; $i--)
- $ymatches[$this->xv[$i]][] = $i;
- else
- for ($i = $ylim - 1; $i >= $yoff; $i--)
- $ymatches[$this->yv[$i]][] = $i;
-
- $this->lcs = 0;
- $this->seq[0]= $yoff - 1;
- $this->in_seq = array();
- $ymids[0] = array();
-
- $numer = $xlim - $xoff + $nchunks - 1;
- $x = $xoff;
- for ($chunk = 0; $chunk < $nchunks; $chunk++) {
- if ($chunk > 0)
- for ($i = 0; $i <= $this->lcs; $i++)
- $ymids[$i][$chunk-1] = $this->seq[$i];
-
- $x1 = $xoff + (int)(($numer + ($xlim-$xoff)*$chunk) / $nchunks);
- for ( ; $x < $x1; $x++) {
+ $fname = '_DiffEngine::_diag';
+ wfProfileIn( $fname );
+ $flip = false;
+
+ if ($xlim - $xoff > $ylim - $yoff) {
+ // Things seems faster (I'm not sure I understand why)
+ // when the shortest sequence in X.
+ $flip = true;
+ list ($xoff, $xlim, $yoff, $ylim)
+ = array( $yoff, $ylim, $xoff, $xlim);
+ }
+
+ if ($flip)
+ for ($i = $ylim - 1; $i >= $yoff; $i--)
+ $ymatches[$this->xv[$i]][] = $i;
+ else
+ for ($i = $ylim - 1; $i >= $yoff; $i--)
+ $ymatches[$this->yv[$i]][] = $i;
+
+ $this->lcs = 0;
+ $this->seq[0]= $yoff - 1;
+ $this->in_seq = array();
+ $ymids[0] = array();
+
+ $numer = $xlim - $xoff + $nchunks - 1;
+ $x = $xoff;
+ for ($chunk = 0; $chunk < $nchunks; $chunk++) {
+ wfProfileIn( "$fname-chunk" );
+ if ($chunk > 0)
+ for ($i = 0; $i <= $this->lcs; $i++)
+ $ymids[$i][$chunk-1] = $this->seq[$i];
+
+ $x1 = $xoff + (int)(($numer + ($xlim-$xoff)*$chunk) / $nchunks);
+ for ( ; $x < $x1; $x++) {
$line = $flip ? $this->yv[$x] : $this->xv[$x];
- if (empty($ymatches[$line]))
- continue;
- $matches = $ymatches[$line];
+ if (empty($ymatches[$line]))
+ continue;
+ $matches = $ymatches[$line];
reset($matches);
- while (list ($junk, $y) = each($matches))
- if (empty($this->in_seq[$y])) {
- $k = $this->_lcs_pos($y);
- USE_ASSERTS && assert($k > 0);
- $ymids[$k] = $ymids[$k-1];
- break;
+ while (list ($junk, $y) = each($matches))
+ if (empty($this->in_seq[$y])) {
+ $k = $this->_lcs_pos($y);
+ USE_ASSERTS && assert($k > 0);
+ $ymids[$k] = $ymids[$k-1];
+ break;
}
- while (list ($junk, $y) = each($matches)) {
- if ($y > $this->seq[$k-1]) {
- USE_ASSERTS && assert($y < $this->seq[$k]);
- // Optimization: this is a common case:
- // next match is just replacing previous match.
- $this->in_seq[$this->seq[$k]] = false;
- $this->seq[$k] = $y;
- $this->in_seq[$y] = 1;
- }
- else if (empty($this->in_seq[$y])) {
- $k = $this->_lcs_pos($y);
- USE_ASSERTS && assert($k > 0);
- $ymids[$k] = $ymids[$k-1];
+ while (list ($junk, $y) = each($matches)) {
+ if ($y > $this->seq[$k-1]) {
+ USE_ASSERTS && assert($y < $this->seq[$k]);
+ // Optimization: this is a common case:
+ // next match is just replacing previous match.
+ $this->in_seq[$this->seq[$k]] = false;
+ $this->seq[$k] = $y;
+ $this->in_seq[$y] = 1;
+ } else if (empty($this->in_seq[$y])) {
+ $k = $this->_lcs_pos($y);
+ USE_ASSERTS && assert($k > 0);
+ $ymids[$k] = $ymids[$k-1];
}
}
}
+ wfProfileOut( "$fname-chunk" );
}
- $seps[] = $flip ? array($yoff, $xoff) : array($xoff, $yoff);
- $ymid = $ymids[$this->lcs];
- for ($n = 0; $n < $nchunks - 1; $n++) {
- $x1 = $xoff + (int)(($numer + ($xlim - $xoff) * $n) / $nchunks);
- $y1 = $ymid[$n] + 1;
- $seps[] = $flip ? array($y1, $x1) : array($x1, $y1);
+ $seps[] = $flip ? array($yoff, $xoff) : array($xoff, $yoff);
+ $ymid = $ymids[$this->lcs];
+ for ($n = 0; $n < $nchunks - 1; $n++) {
+ $x1 = $xoff + (int)(($numer + ($xlim - $xoff) * $n) / $nchunks);
+ $y1 = $ymid[$n] + 1;
+ $seps[] = $flip ? array($y1, $x1) : array($x1, $y1);
}
- $seps[] = $flip ? array($ylim, $xlim) : array($xlim, $ylim);
+ $seps[] = $flip ? array($ylim, $xlim) : array($xlim, $ylim);
- return array($this->lcs, $seps);
+ wfProfileOut( $fname );
+ return array($this->lcs, $seps);
}
function _lcs_pos ($ypos) {
- $end = $this->lcs;
- if ($end == 0 || $ypos > $this->seq[$end]) {
- $this->seq[++$this->lcs] = $ypos;
- $this->in_seq[$ypos] = 1;
- return $this->lcs;
+ $fname = '_DiffEngine::_lcs_pos';
+ wfProfileIn( $fname );
+
+ $end = $this->lcs;
+ if ($end == 0 || $ypos > $this->seq[$end]) {
+ $this->seq[++$this->lcs] = $ypos;
+ $this->in_seq[$ypos] = 1;
+ wfProfileOut( $fname );
+ return $this->lcs;
}
- $beg = 1;
- while ($beg < $end) {
- $mid = (int)(($beg + $end) / 2);
- if ( $ypos > $this->seq[$mid] )
- $beg = $mid + 1;
- else
- $end = $mid;
+ $beg = 1;
+ while ($beg < $end) {
+ $mid = (int)(($beg + $end) / 2);
+ if ( $ypos > $this->seq[$mid] )
+ $beg = $mid + 1;
+ else
+ $end = $mid;
}
- USE_ASSERTS && assert($ypos != $this->seq[$end]);
+ USE_ASSERTS && assert($ypos != $this->seq[$end]);
- $this->in_seq[$this->seq[$end]] = false;
- $this->seq[$end] = $ypos;
- $this->in_seq[$ypos] = 1;
- return $end;
+ $this->in_seq[$this->seq[$end]] = false;
+ $this->seq[$end] = $ypos;
+ $this->in_seq[$ypos] = 1;
+ wfProfileOut( $fname );
+ return $end;
}
/* Find LCS of two sequences.
* All line numbers are origin-0 and discarded lines are not counted.
*/
function _compareseq ($xoff, $xlim, $yoff, $ylim) {
- // Slide down the bottom initial diagonal.
- while ($xoff < $xlim && $yoff < $ylim
+ $fname = '_DiffEngine::_compareseq';
+ wfProfileIn( $fname );
+
+ // Slide down the bottom initial diagonal.
+ while ($xoff < $xlim && $yoff < $ylim
&& $this->xv[$xoff] == $this->yv[$yoff]) {
- ++$xoff;
- ++$yoff;
+ ++$xoff;
+ ++$yoff;
}
- // Slide up the top initial diagonal.
- while ($xlim > $xoff && $ylim > $yoff
+ // Slide up the top initial diagonal.
+ while ($xlim > $xoff && $ylim > $yoff
&& $this->xv[$xlim - 1] == $this->yv[$ylim - 1]) {
- --$xlim;
- --$ylim;
+ --$xlim;
+ --$ylim;
}
- if ($xoff == $xlim || $yoff == $ylim)
- $lcs = 0;
- else {
- // This is ad hoc but seems to work well.
- //$nchunks = sqrt(min($xlim - $xoff, $ylim - $yoff) / 2.5);
- //$nchunks = max(2,min(8,(int)$nchunks));
- $nchunks = min(7, $xlim - $xoff, $ylim - $yoff) + 1;
- list ($lcs, $seps)
- = $this->_diag($xoff,$xlim,$yoff, $ylim,$nchunks);
- }
-
- if ($lcs == 0) {
- // X and Y sequences have no common subsequence:
- // mark all changed.
- while ($yoff < $ylim)
- $this->ychanged[$this->yind[$yoff++]] = 1;
- while ($xoff < $xlim)
- $this->xchanged[$this->xind[$xoff++]] = 1;
- }
- else {
- // Use the partitions to split this problem into subproblems.
- reset($seps);
- $pt1 = $seps[0];
- while ($pt2 = next($seps)) {
- $this->_compareseq ($pt1[0], $pt2[0], $pt1[1], $pt2[1]);
- $pt1 = $pt2;
+ if ($xoff == $xlim || $yoff == $ylim)
+ $lcs = 0;
+ else {
+ // This is ad hoc but seems to work well.
+ //$nchunks = sqrt(min($xlim - $xoff, $ylim - $yoff) / 2.5);
+ //$nchunks = max(2,min(8,(int)$nchunks));
+ $nchunks = min(7, $xlim - $xoff, $ylim - $yoff) + 1;
+ list ($lcs, $seps)
+ = $this->_diag($xoff,$xlim,$yoff, $ylim,$nchunks);
+ }
+
+ if ($lcs == 0) {
+ // X and Y sequences have no common subsequence:
+ // mark all changed.
+ while ($yoff < $ylim)
+ $this->ychanged[$this->yind[$yoff++]] = 1;
+ while ($xoff < $xlim)
+ $this->xchanged[$this->xind[$xoff++]] = 1;
+ } else {
+ // Use the partitions to split this problem into subproblems.
+ reset($seps);
+ $pt1 = $seps[0];
+ while ($pt2 = next($seps)) {
+ $this->_compareseq ($pt1[0], $pt2[0], $pt1[1], $pt2[1]);
+ $pt1 = $pt2;
}
}
+ wfProfileOut( $fname );
}
/* Adjust inserts/deletes of identical lines to join changes
* This is extracted verbatim from analyze.c (GNU diffutils-2.7).
*/
function _shift_boundaries ($lines, &$changed, $other_changed) {
- $i = 0;
- $j = 0;
-
- USE_ASSERTS && assert('sizeof($lines) == sizeof($changed)');
- $len = sizeof($lines);
- $other_len = sizeof($other_changed);
-
- while (1) {
- /*
- * Scan forwards to find beginning of another run of changes.
- * Also keep track of the corresponding point in the other file.
- *
- * Throughout this code, $i and $j are adjusted together so that
- * the first $i elements of $changed and the first $j elements
- * of $other_changed both contain the same number of zeros
- * (unchanged lines).
- * Furthermore, $j is always kept so that $j == $other_len or
- * $other_changed[$j] == false.
- */
- while ($j < $other_len && $other_changed[$j])
- $j++;
-
- while ($i < $len && ! $changed[$i]) {
- USE_ASSERTS && assert('$j < $other_len && ! $other_changed[$j]');
- $i++; $j++;
- while ($j < $other_len && $other_changed[$j])
- $j++;
+ $fname = '_DiffEngine::_shift_boundaries';
+ wfProfileIn( $fname );
+ $i = 0;
+ $j = 0;
+
+ USE_ASSERTS && assert('sizeof($lines) == sizeof($changed)');
+ $len = sizeof($lines);
+ $other_len = sizeof($other_changed);
+
+ while (1) {
+ /*
+ * Scan forwards to find beginning of another run of changes.
+ * Also keep track of the corresponding point in the other file.
+ *
+ * Throughout this code, $i and $j are adjusted together so that
+ * the first $i elements of $changed and the first $j elements
+ * of $other_changed both contain the same number of zeros
+ * (unchanged lines).
+ * Furthermore, $j is always kept so that $j == $other_len or
+ * $other_changed[$j] == false.
+ */
+ while ($j < $other_len && $other_changed[$j])
+ $j++;
+
+ while ($i < $len && ! $changed[$i]) {
+ USE_ASSERTS && assert('$j < $other_len && ! $other_changed[$j]');
+ $i++; $j++;
+ while ($j < $other_len && $other_changed[$j])
+ $j++;
}
-
- if ($i == $len)
- break;
-
- $start = $i;
-
- // Find the end of this run of changes.
- while (++$i < $len && $changed[$i])
- continue;
-
- do {
- /*
- * Record the length of this run of changes, so that
- * we can later determine whether the run has grown.
- */
- $runlength = $i - $start;
-
- /*
- * Move the changed region back, so long as the
- * previous unchanged line matches the last changed one.
- * This merges with previous changed regions.
- */
- while ($start > 0 && $lines[$start - 1] == $lines[$i - 1]) {
- $changed[--$start] = 1;
- $changed[--$i] = false;
- while ($start > 0 && $changed[$start - 1])
- $start--;
- USE_ASSERTS && assert('$j > 0');
- while ($other_changed[--$j])
- continue;
- USE_ASSERTS && assert('$j >= 0 && !$other_changed[$j]');
+
+ if ($i == $len)
+ break;
+
+ $start = $i;
+
+ // Find the end of this run of changes.
+ while (++$i < $len && $changed[$i])
+ continue;
+
+ do {
+ /*
+ * Record the length of this run of changes, so that
+ * we can later determine whether the run has grown.
+ */
+ $runlength = $i - $start;
+
+ /*
+ * Move the changed region back, so long as the
+ * previous unchanged line matches the last changed one.
+ * This merges with previous changed regions.
+ */
+ while ($start > 0 && $lines[$start - 1] == $lines[$i - 1]) {
+ $changed[--$start] = 1;
+ $changed[--$i] = false;
+ while ($start > 0 && $changed[$start - 1])
+ $start--;
+ USE_ASSERTS && assert('$j > 0');
+ while ($other_changed[--$j])
+ continue;
+ USE_ASSERTS && assert('$j >= 0 && !$other_changed[$j]');
}
- /*
- * Set CORRESPONDING to the end of the changed run, at the last
- * point where it corresponds to a changed run in the other file.
- * CORRESPONDING == LEN means no such point has been found.
- */
- $corresponding = $j < $other_len ? $i : $len;
-
- /*
- * Move the changed region forward, so long as the
- * first changed line matches the following unchanged one.
- * This merges with following changed regions.
- * Do this second, so that if there are no merges,
- * the changed region is moved forward as far as possible.
- */
- while ($i < $len && $lines[$start] == $lines[$i]) {
- $changed[$start++] = false;
- $changed[$i++] = 1;
- while ($i < $len && $changed[$i])
- $i++;
-
- USE_ASSERTS && assert('$j < $other_len && ! $other_changed[$j]');
- $j++;
- if ($j < $other_len && $other_changed[$j]) {
- $corresponding = $i;
- while ($j < $other_len && $other_changed[$j])
- $j++;
+ /*
+ * Set CORRESPONDING to the end of the changed run, at the last
+ * point where it corresponds to a changed run in the other file.
+ * CORRESPONDING == LEN means no such point has been found.
+ */
+ $corresponding = $j < $other_len ? $i : $len;
+
+ /*
+ * Move the changed region forward, so long as the
+ * first changed line matches the following unchanged one.
+ * This merges with following changed regions.
+ * Do this second, so that if there are no merges,
+ * the changed region is moved forward as far as possible.
+ */
+ while ($i < $len && $lines[$start] == $lines[$i]) {
+ $changed[$start++] = false;
+ $changed[$i++] = 1;
+ while ($i < $len && $changed[$i])
+ $i++;
+
+ USE_ASSERTS && assert('$j < $other_len && ! $other_changed[$j]');
+ $j++;
+ if ($j < $other_len && $other_changed[$j]) {
+ $corresponding = $i;
+ while ($j < $other_len && $other_changed[$j])
+ $j++;
}
}
} while ($runlength != $i - $start);
- /*
- * If possible, move the fully-merged run of changes
- * back to a corresponding run in the other file.
- */
- while ($corresponding < $i) {
- $changed[--$start] = 1;
- $changed[--$i] = 0;
- USE_ASSERTS && assert('$j > 0');
- while ($other_changed[--$j])
- continue;
- USE_ASSERTS && assert('$j >= 0 && !$other_changed[$j]');
+ /*
+ * If possible, move the fully-merged run of changes
+ * back to a corresponding run in the other file.
+ */
+ while ($corresponding < $i) {
+ $changed[--$start] = 1;
+ $changed[--$i] = 0;
+ USE_ASSERTS && assert('$j > 0');
+ while ($other_changed[--$j])
+ continue;
+ USE_ASSERTS && assert('$j >= 0 && !$other_changed[$j]');
}
}
+ wfProfileOut( $fname );
}
}
/**
* Class representing a 'diff' between two sequences of strings.
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
*/
-class Diff
+class Diff
{
var $edits;
* original diff.
*/
function reverse () {
- $rev = $this;
+ $rev = $this;
$rev->edits = array();
foreach ($this->edits as $edit) {
$rev->edits[] = $edit->reverse();
}
- return $rev;
+ return $rev;
}
/**
}
return true;
}
-
+
/**
* Compute the length of the Longest Common Subsequence (LCS).
*
* @return int The length of the LCS.
*/
function lcs () {
- $lcs = 0;
+ $lcs = 0;
foreach ($this->edits as $edit) {
if ($edit->type == 'copy')
$lcs += sizeof($edit->orig);
}
- return $lcs;
+ return $lcs;
}
/**
*/
function orig() {
$lines = array();
-
+
foreach ($this->edits as $edit) {
if ($edit->orig)
array_splice($lines, sizeof($lines), 0, $edit->orig);
*/
function closing() {
$lines = array();
-
+
foreach ($this->edits as $edit) {
if ($edit->closing)
array_splice($lines, sizeof($lines), 0, $edit->closing);
}
/**
- * Check a Diff for validity.
+ * Check a Diff for validity.
*
* This is here only for debugging purposes.
*/
function _check ($from_lines, $to_lines) {
+ $fname = 'Diff::_check';
+ wfProfileIn( $fname );
if (serialize($from_lines) != serialize($this->orig()))
trigger_error("Reconstructed original doesn't match", E_USER_ERROR);
if (serialize($to_lines) != serialize($this->closing()))
}
$lcs = $this->lcs();
- trigger_error("Diff okay: LCS = $lcs", E_USER_NOTICE);
+ trigger_error('Diff okay: LCS = '.$lcs, E_USER_NOTICE);
+ wfProfileOut( $fname );
}
}
-
+
/**
* FIXME: bad name.
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
*/
-class MappedDiff
-extends Diff
+class MappedDiff extends Diff
{
/**
* Constructor.
*/
function MappedDiff($from_lines, $to_lines,
$mapped_from_lines, $mapped_to_lines) {
+ $fname = 'MappedDiff::MappedDiff';
+ wfProfileIn( $fname );
assert(sizeof($from_lines) == sizeof($mapped_from_lines));
assert(sizeof($to_lines) == sizeof($mapped_to_lines));
-
+
$this->Diff($mapped_from_lines, $mapped_to_lines);
$xi = $yi = 0;
$orig = array_slice($from_lines, $xi, sizeof($orig));
$xi += sizeof($orig);
}
-
+
$closing = &$this->edits[$i]->closing;
if (is_array($closing)) {
$closing = array_slice($to_lines, $yi, sizeof($closing));
$yi += sizeof($closing);
}
}
+ wfProfileOut( $fname );
}
}
* This class formats the diff in classic diff format.
* It is intended that this class be customized via inheritance,
* to obtain fancier outputs.
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
*/
class DiffFormatter
{
* @return string The formatted output.
*/
function format($diff) {
+ $fname = 'DiffFormatter::format';
+ wfProfileIn( $fname );
$xi = $yi = 1;
$block = false;
$y0, $yi - $y0,
$block);
- return $this->_end_diff();
+ $end = $this->_end_diff();
+ wfProfileOut( $fname );
+ return $end;
}
function _block($xbeg, $xlen, $ybeg, $ylen, &$edits) {
+ $fname = 'DiffFormatter::_block';
+ wfProfileIn( $fname );
$this->_start_block($this->_block_header($xbeg, $xlen, $ybeg, $ylen));
foreach ($edits as $edit) {
if ($edit->type == 'copy')
elseif ($edit->type == 'change')
$this->_changed($edit->orig, $edit->closing);
else
- trigger_error("Unknown edit type", E_USER_ERROR);
+ trigger_error('Unknown edit type', E_USER_ERROR);
}
$this->_end_block();
+ wfProfileOut( $fname );
}
function _start_diff() {
return $xbeg . ($xlen ? ($ylen ? 'c' : 'd') : 'a') . $ybeg;
}
-
+
function _start_block($header) {
echo $header;
}
-
+
function _end_block() {
}
foreach ($lines as $line)
echo "$prefix $line\n";
}
-
+
function _context($lines) {
$this->_lines($lines);
}
function _added($lines) {
- $this->_lines($lines, ">");
+ $this->_lines($lines, '>');
}
function _deleted($lines) {
- $this->_lines($lines, "<");
+ $this->_lines($lines, '<');
}
function _changed($orig, $closing) {
/**
* Additions by Axel Boldt follow, partly taken from diff.php, phpwiki-1.3.3
- *
+ *
*/
-define('NBSP', "\xA0"); // iso-8859-x non-breaking space.
+define('NBSP', ' '); // iso-8859-x non-breaking space.
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class _HWLDF_WordAccumulator {
function _HWLDF_WordAccumulator () {
$this->_lines = array();
function _flushGroup ($new_tag) {
if ($this->_group !== '') {
- if ($this->_tag == 'mark')
- $this->_line .= "<font color=\"red\">$this->_group</font>";
- else
- $this->_line .= $this->_group;
- }
+ if ($this->_tag == 'mark')
+ $this->_line .= '<span class="diffchange">' .
+ htmlspecialchars ( $this->_group ) . '</span>';
+ else
+ $this->_line .= htmlspecialchars ( $this->_group );
+ }
$this->_group = '';
$this->_tag = $new_tag;
}
-
+
function _flushLine ($new_tag) {
$this->_flushGroup($new_tag);
if ($this->_line != '')
- $this->_lines[] = $this->_line;
+ array_push ( $this->_lines, $this->_line );
+ else
+ # make empty lines visible by inserting an NBSP
+ array_push ( $this->_lines, NBSP );
$this->_line = '';
}
-
+
function addWords ($words, $tag = '') {
if ($tag != $this->_tag)
$this->_flushGroup($tag);
if ($word == '')
continue;
if ($word[0] == "\n") {
- $this->_group .= NBSP;
$this->_flushLine($tag);
$word = substr($word, 1);
}
}
}
+/**
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
+ */
class WordLevelDiff extends MappedDiff
{
function WordLevelDiff ($orig_lines, $closing_lines) {
+ $fname = 'WordLevelDiff::WordLevelDiff';
+ wfProfileIn( $fname );
+
list ($orig_words, $orig_stripped) = $this->_split($orig_lines);
list ($closing_words, $closing_stripped) = $this->_split($closing_lines);
-
$this->MappedDiff($orig_words, $closing_words,
$orig_stripped, $closing_stripped);
+ wfProfileOut( $fname );
}
function _split($lines) {
- // FIXME: fix POSIX char class.
-# if (!preg_match_all('/ ( [^\S\n]+ | [[:alnum:]]+ | . ) (?: (?!< \n) [^\S\n])? /xs',
- if (!preg_match_all('/ ( [^\S\n]+ | [0-9_A-Za-z\x80-\xff]+ | . ) (?: (?!< \n) [^\S\n])? /xs',
- implode("\n", $lines),
- $m)) {
- return array(array(''), array(''));
+ $fname = 'WordLevelDiff::_split';
+ wfProfileIn( $fname );
+
+ $words = array();
+ $stripped = array();
+ $first = true;
+ foreach ( $lines as $line ) {
+ # If the line is too long, just pretend the entire line is one big word
+ # This prevents resource exhaustion problems
+ if ( $first ) {
+ $first = false;
+ } else {
+ $words[] = "\n";
+ $stripped[] = "\n";
+ }
+ if ( strlen( $line ) > MAX_DIFF_LINE ) {
+ $words[] = $line;
+ $stripped[] = $line;
+ } else {
+ if (preg_match_all('/ ( [^\S\n]+ | [0-9_A-Za-z\x80-\xff]+ | . ) (?: (?!< \n) [^\S\n])? /xs',
+ $line, $m))
+ {
+ $words = array_merge( $words, $m[0] );
+ $stripped = array_merge( $stripped, $m[1] );
+ }
+ }
}
- return array($m[0], $m[1]);
+ wfProfileOut( $fname );
+ return array($words, $stripped);
}
function orig () {
+ $fname = 'WordLevelDiff::orig';
+ wfProfileIn( $fname );
$orig = new _HWLDF_WordAccumulator;
-
+
foreach ($this->edits as $edit) {
if ($edit->type == 'copy')
$orig->addWords($edit->orig);
elseif ($edit->orig)
$orig->addWords($edit->orig, 'mark');
}
- return $orig->getLines();
+ $lines = $orig->getLines();
+ wfProfileOut( $fname );
+ return $lines;
}
function closing () {
+ $fname = 'WordLevelDiff::closing';
+ wfProfileIn( $fname );
$closing = new _HWLDF_WordAccumulator;
-
+
foreach ($this->edits as $edit) {
if ($edit->type == 'copy')
$closing->addWords($edit->closing);
elseif ($edit->closing)
$closing->addWords($edit->closing, 'mark');
}
- return $closing->getLines();
+ $lines = $closing->getLines();
+ wfProfileOut( $fname );
+ return $lines;
}
}
/**
* Wikipedia Table style diff formatter.
- *
+ * @todo document
+ * @private
+ * @package MediaWiki
+ * @subpackage DifferenceEngine
*/
class TableDiffFormatter extends DiffFormatter
{
$this->leading_context_lines = 2;
$this->trailing_context_lines = 2;
}
-
- function _block_header( $xbeg, $xlen, $ybeg, $ylen ) {
- $l1 = wfMsg( "lineno", $xbeg );
- $l2 = wfMsg( "lineno", $ybeg );
- $r = "<tr><td colspan='2' align='left'><strong>{$l1}</strong></td>\n" .
- "<td colspan='2' align='left'><strong>{$l2}</strong></td></tr>\n";
+ function _block_header( $xbeg, $xlen, $ybeg, $ylen ) {
+ $r = '<tr><td colspan="2" align="left"><strong><!--LINE '.$xbeg."--></strong></td>\n" .
+ '<td colspan="2" align="left"><strong><!--LINE '.$ybeg."--></strong></td></tr>\n";
return $r;
}
function _start_block( $header ) {
- global $wgOut;
- $wgOut->addHTML( $header );
+ echo $header;
}
function _end_block() {
}
- function _lines( $lines, $prefix=' ', $color="white" ) {
+ function _lines( $lines, $prefix=' ', $color='white' ) {
}
+ # HTML-escape parameter before calling this
function addedLine( $line ) {
- return "<td>+</td><td class='diff-addedline'>" .
- "<small>{$line}</small></td>";
+ return "<td>+</td><td class='diff-addedline'>{$line}</td>";
}
+ # HTML-escape parameter before calling this
function deletedLine( $line ) {
- return "<td>-</td><td class='diff-deletedline'>" .
- "<small>{$line}</small></td>";
+ return "<td>-</td><td class='diff-deletedline'>{$line}</td>";
}
- function emptyLine() {
- return "<td colspan='2'> </td>";
+ # HTML-escape parameter before calling this
+ function contextLine( $line ) {
+ return "<td> </td><td class='diff-context'>{$line}</td>";
}
- function contextLine( $line ) {
- return "<td> </td><td class='diff-context'><small>{$line}</small></td>";
+ function emptyLine() {
+ return '<td colspan="2"> </td>';
}
-
- function _added($lines) {
- global $wgOut;
+
+ function _added( $lines ) {
foreach ($lines as $line) {
- $wgOut->addHTML( "<tr>" . $this->emptyLine() .
- $this->addedLine( $line ) . "</tr>\n" );
+ echo '<tr>' . $this->emptyLine() .
+ $this->addedLine( htmlspecialchars ( $line ) ) . "</tr>\n";
}
}
function _deleted($lines) {
- global $wgOut;
foreach ($lines as $line) {
- $wgOut->addHTML( "<tr>" . $this->deletedLine( $line ) .
- $this->emptyLine() . "</tr>\n" );
+ echo '<tr>' . $this->deletedLine( htmlspecialchars ( $line ) ) .
+ $this->emptyLine() . "</tr>\n";
}
}
function _context( $lines ) {
- global $wgOut;
foreach ($lines as $line) {
- $wgOut->addHTML( "<tr>" . $this->contextLine( $line ) .
- $this->contextLine( $line ) . "</tr>\n" );
+ echo '<tr>' .
+ $this->contextLine( htmlspecialchars ( $line ) ) .
+ $this->contextLine( htmlspecialchars ( $line ) ) . "</tr>\n";
}
}
function _changed( $orig, $closing ) {
- global $wgOut;
+ $fname = 'TableDiffFormatter::_changed';
+ wfProfileIn( $fname );
+
$diff = new WordLevelDiff( $orig, $closing );
$del = $diff->orig();
$add = $diff->closing();
+ # Notice that WordLevelDiff returns HTML-escaped output.
+ # Hence, we will be calling addedLine/deletedLine without HTML-escaping.
+
while ( $line = array_shift( $del ) ) {
$aline = array_shift( $add );
- $wgOut->addHTML( "<tr>" . $this->deletedLine( $line ) .
- $this->addedLine( $aline ) . "</tr>\n" );
+ echo '<tr>' . $this->deletedLine( $line ) .
+ $this->addedLine( $aline ) . "</tr>\n";
}
- $this->_added( $add ); # If any leftovers
+ foreach ($add as $line) { # If any leftovers
+ echo '<tr>' . $this->emptyLine() .
+ $this->addedLine( $line ) . "</tr>\n";
+ }
+ wfProfileOut( $fname );
}
}