<?php
/**
* See diff.doc
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
/** */
require_once( 'Revision.php' );
+define( 'MAX_DIFF_LINE', 10000 );
+define( 'MAX_DIFF_XREF_LENGTH', 10000 );
+
/**
* @todo document
* @access public
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class DifferenceEngine {
- /* private */ var $mOldid, $mNewid;
- /* private */ var $mOldtitle, $mNewtitle, $mPagetitle;
- /* private */ var $mOldtext, $mNewtext;
- /* private */ var $mOldUser, $mNewUser;
- /* private */ var $mOldComment, $mNewComment;
- /* private */ var $mOldPage, $mNewPage;
- /* private */ var $mRcidMarkPatrolled;
-
- function DifferenceEngine( $old, $new, $rcid = 0 )
- {
- global $wgTitle;
+ /**#@+
+ * @access private
+ */
+ var $mOldid, $mNewid, $mTitle;
+ var $mOldtitle, $mNewtitle, $mPagetitle;
+ var $mOldtext, $mNewtext;
+ var $mOldUser, $mNewUser;
+ var $mOldComment, $mNewComment;
+ 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 Title $titleObj Title object that the diff is associated with
+ * @param integer $old Old ID we want to show and diff with.
+ * @param string $new Either 'prev' or 'next'.
+ * @param integer $rcid ??? (default 0)
+ */
+ function DifferenceEngine( $titleObj = null, $old = 0, $new = 0, $rcid = 0 ) {
+ $this->mTitle = $titleObj;
+
if ( 'prev' == $new ) {
# Show diff between revision $old and the previous one.
# Get previous one from DB.
#
$this->mNewid = intval($old);
- $this->mOldid = $wgTitle->getPreviousRevisionID( $this->mNewid );
+ $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 = $wgTitle->getNextRevisionID( $this->mOldid );
+ $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".
}
} else {
-
$this->mOldid = intval($old);
$this->mNewid = intval($new);
}
$this->mRcidMarkPatrolled = intval($rcid); # force it to be an integer
}
- function showDiffPage()
- {
- global $wgUser, $wgTitle, $wgOut, $wgContLang, $wgOnlySysopsCanPatrol, $wgUseRCPatrol;
+ function showDiffPage() {
+ global $wgUser, $wgOut, $wgContLang, $wgOnlySysopsCanPatrol,
+ $wgUseExternalEditor, $wgUseRCPatrol;
$fname = 'DifferenceEngine::showDiffPage';
wfProfileIn( $fname );
+ # If external diffs are enabled both globally and for the user,
+ # we'll use the application/x-external-editor interface to call
+ # an external diff tool like kompare, kdiff3, etc.
+ if($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;
+ }
+
# 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.
return;
}
- $t = $wgTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
+ $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() ) {
+ if ( ! $this->loadRevisionData() ) {
$wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) );
- $wgOut->addHTML( $mtext );
+ $wgOut->addWikitext( $mtext );
wfProfileOut( $fname );
return;
}
+ if ( $this->mNewRev->isCurrent() ) {
+ $wgOut->setArticleFlag( true );
+ }
+
$wgOut->suppressQuickbar();
$oldTitle = $this->mOldPage->getPrefixedText();
'target=' . urlencode($this->mOldUser) );
$newContribs = $sk->makeKnownLinkObj( Title::makeTitle( NS_SPECIAL, 'Contributions' ), $contribs,
'target=' . urlencode($this->mNewUser) );
- if ( !$this->mNewid && $wgUser->isAllowed('rollback') ) {
- $rollback = ' <strong>[' . $sk->makeKnownLinkObj( $wgTitle, wfMsg( 'rollbacklink' ),
- 'action=rollback&from=' . urlencode($this->mNewUser) ) . ']</strong>';
+ if ( $this->mNewRev->isCurrent() && $wgUser->isAllowed('rollback') ) {
+ $rollback = ' <strong>[' . $sk->makeKnownLinkObj( $this->mTitle, wfMsg( 'rollbacklink' ),
+ 'action=rollback&from=' . urlencode($this->mNewUser) .
+ '&token=' . urlencode( $wgUser->editToken( array( $this->mTitle->getPrefixedText(), $this->mNewUser ) ) ) ) .
+ ']</strong>';
} else {
$rollback = '';
}
- if ( $wgUseRCPatrol && $this->mRcidMarkPatrolled != 0 && $wgUser->getID() != 0 &&
+ if ( $wgUseRCPatrol && $this->mRcidMarkPatrolled != 0 && $wgUser->isLoggedIn() &&
( $wgUser->isAllowed('rollback') || !$wgOnlySysopsCanPatrol ) )
{
- $patrol = ' [' . $sk->makeKnownLinkObj( $wgTitle, wfMsg( 'markaspatrolleddiff' ),
+ $patrol = ' [' . $sk->makeKnownLinkObj( $this->mTitle, wfMsg( 'markaspatrolleddiff' ),
"action=markpatrolled&rcid={$this->mRcidMarkPatrolled}" ) . ']';
} else {
$patrol = '';
}
- $prevlink = $sk->makeKnownLinkObj( $wgTitle, wfMsg( 'previousdiff' ), 'diff=prev&oldid='.$this->mOldid );
- if ( $this->mNewid == 0 ) {
+ $prevlink = $sk->makeKnownLinkObj( $this->mTitle, wfMsgHtml( 'previousdiff' ),
+ 'diff=prev&oldid='.$this->mOldid, '', '', 'id="differences-prevlink"' );
+ if ( $this->mNewRev->isCurrent() ) {
$nextlink = '';
} else {
- $nextlink = $sk->makeKnownLinkObj( $wgTitle, wfMsg( 'nextdiff' ), 'diff=next&oldid='.$this->mNewid );
+ $nextlink = $sk->makeKnownLinkObj( $this->mTitle, wfMsgHtml( 'nextdiff' ),
+ 'diff=next&oldid='.$this->mNewid, '', '', 'id="differences-nextlink"' );
}
$oldHeader = "<strong>{$this->mOldtitle}</strong><br />$oldUserLink " .
"($newUTLink | $newContribs) $rollback<br />" . $this->mNewComment .
'<br />' . $nextlink . $patrol;
- DifferenceEngine::showDiff( $this->mOldtext, $this->mNewtext,
- $oldHeader, $newHeader );
+ $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->addWikiText( $this->mNewtext );
+ if( !$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, $wgTitle, $wgUser, $wgLang;
+ /**
+ * Show the first revision of an article. Uses normal diff headers in
+ * contrast to normal "old revision" display style.
+ */
+ function showFirstRevision() {
+ global $wgOut, $wgUser, $wgLang;
$fname = 'DifferenceEngine::showFirstRevision';
wfProfileIn( $fname );
-
- $this->mOldid = $this->mNewid; # hack to make loadText() work.
-
# Get article text from the DB
#
- if ( ! $this->loadText() ) {
- $t = $wgTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
+ if ( ! $this->loadNewText() ) {
+ $t = $this->mTitle->getPrefixedText() . " (Diff: {$this->mOldid}, " .
"{$this->mNewid})";
- $mtext = wfMsg( 'missingarticle', $t );
+ $mtext = wfMsg( 'missingarticle', "<nowiki>$t</nowiki>" );
$wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) );
- $wgOut->addHTML( $mtext );
+ $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->mOldPage->userCanRead() ) ) {
+ if ( !( $this->mTitle->userCanRead() ) ) {
$wgOut->loginToUse();
$wgOut->output();
wfProfileOut( $fname );
$userLink = $sk->makeLinkObj( Title::makeTitleSafe( NS_USER, $this->mOldUser ), $this->mOldUser );
$contribs = $sk->makeKnownLinkObj( Title::makeTitle( NS_SPECIAL, 'Contributions' ), wfMsg( 'contribslink' ),
'target=' . urlencode($this->mOldUser) );
- $nextlink = $sk->makeKnownLinkObj( $wgTitle, wfMsg( 'nextdiff' ), 'diff=next&oldid='.$this->mNewid );
+ $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 />$userLink " .
"($uTLink | $contribs)<br />" . $this->mOldComment .
'<br />' . $nextlink. "</div>\n";
# Show current revision
#
$wgOut->addHTML( "<hr /><h2>{$this->mPagetitle}</h2>\n" );
+ if( is_object( $this->mNewRev ) ) {
+ $wgOut->setRevisionId( $this->mNewRev->getId() );
+ }
$wgOut->addWikiText( $this->mNewtext );
wfProfileOut( $fname );
}
- function showDiff( $otext, $ntext, $otitle, $ntitle )
- {
+ /**
+ * 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;
- $wgOut->addHTML( DifferenceEngine::getDiff( $otext, $ntext, $otitle, $ntitle ) );
+ $diff = $this->getDiff( $otitle, $ntitle );
+ if ( $diff === false ) {
+ $wgOut->addWikitext( wfMsg( 'missingarticle', "<nowiki>$t</nowiki>" ) );
+ return false;
+ } else {
+ $wgOut->addHTML( $diff );
+ return true;
+ }
}
-
- function getDiff( $otext, $ntext, $otitle, $ntitle ) {
- global $wgUseExternalDiffEngine, $wgContLang;
- $out = "
- <table border='0' width='98%' cellpadding='0' cellspacing='4' class='diff'>
- <tr>
- <td colspan='2' width='50%' align='center' class='diff-otitle'>{$otitle}</td>
- <td colspan='2' width='50%' align='center' class='diff-ntitle'>{$ntitle}</td>
- </tr>
- ";
- $otext = $wgContLang->segmentForDiff($otext);
- $ntext = $wgContLang->segmentForDiff($ntext);
- $difftext='';
+
+ /**
+ * 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 {
+ return $this->addHeader( $body, $otitle, $ntitle );
+ }
+ }
+
+ /**
+ * Get the diff table body, without header
+ * Returns false on error
+ */
+ function getDiffBody() {
+ global $wgUseExternalDiffEngine, $wgContLang, $wgMemc, $wgDBname;
+
+ // 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' );
+ return $difftext;
+ }
+ }
+
+ if ( !$this->loadText() ) {
+ return false;
+ }
+
+ $otext = $wgContLang->segmentForDiff($this->mOldtext);
+ $ntext = $wgContLang->segmentForDiff($this->mNewtext);
if ( $wgUseExternalDiffEngine ) {
# For historical reasons, external diff engine expects
# input text to be HTML-escaped already
$difftext = $formatter->format( $diffs );
}
$difftext = $wgContLang->unsegmentForDiff($difftext);
- $out .= $difftext."</table>\n";
+
+ // Save to cache for 7 days
+ if ( $key !== false ) {
+ wfIncrStats( 'diff_cache_miss' );
+ $wgMemc->set( $key, $difftext, 7*86400 );
+ } else {
+ wfIncrStats( 'diff_uncacheable' );
+ }
+ return $difftext;
+ }
+
+ /**
+ * 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;
}
- # 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;
- $fname = 'DifferenceEngine::loadText';
+ /**
+ * 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 {
+ // Whether it succeeds or fails, we don't want to try again
+ $this->mRevisionsLoaded = true;
+ }
- $dbr =& wfGetDB( DB_SLAVE );
+ // Load the new revision object
if( $this->mNewid ) {
- $this->newRev =& Revision::newFromId( $this->mNewid );
+ $this->mNewRev = Revision::newFromId( $this->mNewid );
} else {
- $this->newRev =& Revision::newFromTitle( $wgTitle );
+ $this->mNewRev = Revision::newFromTitle( $this->mTitle );
}
-
- if( $this->newRev->isCurrent() ) {
+
+ if( is_null( $this->mNewRev ) ) {
+ return false;
+ }
+
+ // Set assorted variables
+ if( $this->mNewRev->isCurrent() ) {
$this->mPagetitle = htmlspecialchars( wfMsg( 'currentrev' ) );
- $this->mNewPage = $wgTitle;
+ $this->mNewPage = $this->mTitle;
$newLink = $this->mNewPage->escapeLocalUrl();
$this->mNewtitle = "<a href='$newLink'>{$this->mPagetitle}</a>";
} else {
- $this->mNewPage = $this->newRev->getTitle();
+ $this->mNewPage = $this->mNewRev->getTitle();
$newLink = $this->mNewPage->escapeLocalUrl ('oldid=' . $this->mNewid );
- $t = $wgLang->timeanddate( $this->newRev->getTimestamp(), true );
+ $t = $wgLang->timeanddate( $this->mNewRev->getTimestamp(), true );
$this->mPagetitle = htmlspecialchars( wfMsg( 'revisionasof', $t ) );
$this->mNewtitle = "<a href='$newLink'>{$this->mPagetitle}</a>";
}
-
+
+ $this->mNewUser = $this->mNewRev->getUserText();
+ $this->mNewComment = $this->mNewRev->getComment();
+
+ // Load the old revision object
+ $this->mOldRev = false;
if( $this->mOldid ) {
- $this->oldRev =& Revision::newFromId( $this->mOldid );
+ $this->mOldRev = Revision::newFromId( $this->mOldid );
+ } elseif ( $this->mOldid === 0 ) {
+ $this->mOldRev = $this->mNewRev->getPrevious();
+ $this->mOldid = $this->mOldRev->getId();
+ }/* 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 );
+ $this->mOldtitle = "<a href='$oldLink'>" . htmlspecialchars( wfMsg( 'revisionasof', $t ) ) . '</a>';
+
+
+ $this->mOldUser = $this->mOldRev->getUserText();
+ $this->mOldComment = $this->mOldRev->getComment();
+ }
+
+ return true;
+ }
+
+ /**
+ * Load the text of the revisions, as well as revision data.
+ */
+ function loadText() {
+ if ( $this->mTextLoaded == 2 ) {
+ return true;
} else {
- $this->oldRev =& $this->newRev->getPrevious();
- $this->mOldid = $this->oldRev->getId();
- }
-
- $this->mOldPage = $this->oldRev->getTitle();
-
- $t = $wgLang->timeanddate( $this->oldRev->getTimestamp(), true );
- $oldLink = $this->mOldPage->escapeLocalUrl( 'oldid=' . $this->mOldid );
- $this->mOldtitle = "<a href='$oldLink'>" . htmlspecialchars( wfMsg( 'revisionasof', $t ) ) . '</a>';
-
- $this->mNewUser = $this->newRev->getUserText();
- $this->mNewComment = $this->newRev->getComment();
- $this->mNewtext = $this->newRev->getText();
-
- $this->mOldUser = $this->oldRev->getUserText();
- $this->mOldComment = $this->oldRev->getComment();
- $this->mOldtext = $this->oldRev->getText();
+ // Whether it succeeds or fails, we don't want to try again
+ $this->mTextLoaded = 2;
+ }
+ if ( !$this->loadRevisionData() ) {
+ return false;
+ }
+ if ( $this->mOldRev ) {
+ $this->mOldtext = $this->mOldRev->getText();
+ }
+ if ( $this->mNewRev ) {
+ $this->mNewtext = $this->mNewRev->getText();
+ }
+ 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 = $this->mNewRev->getText();
return true;
}
+
+
}
// A PHP diff engine for phpwiki. (Taken from phpwiki-1.3.3)
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffOp {
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffOp_Copy extends _DiffOp {
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffOp_Delete extends _DiffOp {
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffOp_Add extends _DiffOp {
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffOp_Change extends _DiffOp {
* closingly, some ideas (subdivision by NCHUNKS > 2, and some optimizations)
* are my own.
*
- * @author Geoffrey T. Dairiki
+ * Line length limits for robustness added by Tim Starling, 2005-08-31
+ *
+ * @author Geoffrey T. Dairiki, Tim Starling
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _DiffEngine
function diff ($from_lines, $to_lines) {
$fname = '_DiffEngine::diff';
wfProfileIn( $fname );
-
+
$n_from = sizeof($from_lines);
$n_to = sizeof($to_lines);
// 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;
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 $line;
+ }
+ }
+
/* Divide the Largest Common Subsequence (LCS) of the sequences
* [XOFF, XLIM) and [YOFF, YLIM) into NCHUNKS approximately equally
else
for ($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++) {
$line = $flip ? $this->yv[$x] : $this->xv[$x];
}
wfProfileOut( "$fname-chunk" );
}
-
+
$seps[] = $flip ? array($yoff, $xoff) : array($xoff, $yoff);
$ymid = $ymids[$this->lcs];
for ($n = 0; $n < $nchunks - 1; $n++) {
$seps[] = $flip ? array($y1, $x1) : array($x1, $y1);
}
$seps[] = $flip ? array($ylim, $xlim) : array($xlim, $ylim);
-
+
wfProfileOut( $fname );
return array($this->lcs, $seps);
}
function _lcs_pos ($ypos) {
$fname = '_DiffEngine::_lcs_pos';
wfProfileIn( $fname );
-
+
$end = $this->lcs;
if ($end == 0 || $ypos > $this->seq[$end]) {
$this->seq[++$this->lcs] = $ypos;
else
$end = $mid;
}
-
+
USE_ASSERTS && assert($ypos != $this->seq[$end]);
-
+
$this->in_seq[$this->seq[$end]] = false;
$this->seq[$end] = $ypos;
$this->in_seq[$ypos] = 1;
function _compareseq ($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]) {
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.
*/
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 - $start;
-
+
/*
* Move the changed region back, so long as the
* previous unchanged line matches the last changed one.
continue;
USE_ASSERTS && assert('$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.
$changed[$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]) {
}
}
} while ($runlength != $i - $start);
-
+
/*
* If possible, move the fully-merged run of changes
* back to a corresponding run in the other file.
* Class representing a 'diff' between two sequences of strings.
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class Diff
* FIXME: bad name.
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class MappedDiff extends Diff
$mapped_from_lines, $mapped_to_lines) {
$fname = 'MappedDiff::MappedDiff';
wfProfileIn( $fname );
-
+
assert(sizeof($from_lines) == sizeof($mapped_from_lines));
assert(sizeof($to_lines) == sizeof($mapped_to_lines));
* to obtain fancier outputs.
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class DiffFormatter
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class _HWLDF_WordAccumulator {
/**
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class WordLevelDiff extends MappedDiff
function WordLevelDiff ($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) {
$fname = 'WordLevelDiff::_split';
wfProfileIn( $fname );
- if (!preg_match_all('/ ( [^\S\n]+ | [0-9_A-Za-z\x80-\xff]+ | . ) (?: (?!< \n) [^\S\n])? /xs',
- implode("\n", $lines),
- $m)) {
- wfProfileOut( $fname );
- return array(array(''), array(''));
+
+ $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] );
+ }
+ }
}
wfProfileOut( $fname );
- return array($m[0], $m[1]);
+ return array($words, $stripped);
}
function orig () {
* Wikipedia Table style diff formatter.
* @todo document
* @access private
- * @package MediaWiki
+ * @package MediaWiki
* @subpackage DifferenceEngine
*/
class TableDiffFormatter extends DiffFormatter
function _changed( $orig, $closing ) {
$fname = 'TableDiffFormatter::_changed';
wfProfileIn( $fname );
-
+
$diff = new WordLevelDiff( $orig, $closing );
$del = $diff->orig();
$add = $diff->closing();
dépôts / lhc / web / wiklou.git / blobdiff