From f8f6417965b269b56e8098df3182c74812a422b4 Mon Sep 17 00:00:00 2001 From: Alexandre Emsenhuber Date: Sun, 7 Nov 2010 16:28:11 +0000 Subject: [PATCH] * (bug 24833) Files name in includes/diff/ are now less confusing Diff.php -> WikiDiff3.php, more descriptive DifferenceEngine.php -> WikiDiff.php, for consistency with the above and to make way for the file below DifferenceInterface.php -> DifferenceEngine.php, since it contains the DifferenceEngine class --- RELEASE-NOTES | 1 + includes/AutoLoader.php | 34 +- includes/diff/DifferenceEngine.php | 1985 ++++++++++----------- includes/diff/DifferenceInterface.php | 1072 ----------- includes/diff/WikiDiff.php | 1241 +++++++++++++ includes/diff/{Diff.php => WikiDiff3.php} | 0 6 files changed, 2167 insertions(+), 2166 deletions(-) delete mode 100644 includes/diff/DifferenceInterface.php create mode 100644 includes/diff/WikiDiff.php rename includes/diff/{Diff.php => WikiDiff3.php} (100%) diff --git a/RELEASE-NOTES b/RELEASE-NOTES index 53069e5bd8..5733e9af7a 100644 --- a/RELEASE-NOTES +++ b/RELEASE-NOTES @@ -401,6 +401,7 @@ LocalSettings.php. The specific bugs are listed below in the general notes. * Cache multiple sizes of InstantCommons thumbnails * (bug 25488) Disallowing anonymous users to read pages no longer throws error on discussion pages with vector as default skin +* (bug 24833) Files name in includes/diff/ are now less confusing === API changes in 1.17 === * (bug 22738) Allow filtering by action type on query=logevent. diff --git a/includes/AutoLoader.php b/includes/AutoLoader.php index ac49392665..6a6ed8d5a2 100644 --- a/includes/AutoLoader.php +++ b/includes/AutoLoader.php @@ -398,23 +398,23 @@ $wgAutoloadLocalClasses = array( 'IBM_DB2Field' => 'includes/db/DatabaseIbm_db2.php', # includes/diff - 'ArrayDiffFormatter' => 'includes/diff/DifferenceEngine.php', - '_DiffEngine' => 'includes/diff/DifferenceEngine.php', - 'DifferenceEngine' => 'includes/diff/DifferenceInterface.php', - 'DiffFormatter' => 'includes/diff/DifferenceEngine.php', - 'Diff' => 'includes/diff/DifferenceEngine.php', - '_DiffOp_Add' => 'includes/diff/DifferenceEngine.php', - '_DiffOp_Change' => 'includes/diff/DifferenceEngine.php', - '_DiffOp_Copy' => 'includes/diff/DifferenceEngine.php', - '_DiffOp_Delete' => 'includes/diff/DifferenceEngine.php', - '_DiffOp' => 'includes/diff/DifferenceEngine.php', - '_HWLDF_WordAccumulator' => 'includes/diff/DifferenceEngine.php', - 'MappedDiff' => 'includes/diff/DifferenceEngine.php', - 'RangeDifference' => 'includes/diff/Diff.php', - 'TableDiffFormatter' => 'includes/diff/DifferenceEngine.php', - 'UnifiedDiffFormatter' => 'includes/diff/DifferenceEngine.php', - 'WikiDiff3' => 'includes/diff/Diff.php', - 'WordLevelDiff' => 'includes/diff/DifferenceEngine.php', + 'ArrayDiffFormatter' => 'includes/diff/WikiDiff.php', + '_DiffEngine' => 'includes/diff/WikiDiff.php', + 'DifferenceEngine' => 'includes/diff/DifferenceEngine.php', + 'DiffFormatter' => 'includes/diff/WikiDiff.php', + 'Diff' => 'includes/diff/WikiDiff.php', + '_DiffOp_Add' => 'includes/diff/WikiDiff.php', + '_DiffOp_Change' => 'includes/diff/WikiDiff.php', + '_DiffOp_Copy' => 'includes/diff/WikiDiff.php', + '_DiffOp_Delete' => 'includes/diff/WikiDiff.php', + '_DiffOp' => 'includes/diff/WikiDiff.php', + '_HWLDF_WordAccumulator' => 'includes/diff/WikiDiff.php', + 'MappedDiff' => 'includes/diff/WikiDiff.php', + 'RangeDifference' => 'includes/diff/WikiDiff3.php', + 'TableDiffFormatter' => 'includes/diff/WikiDiff.php', + 'UnifiedDiffFormatter' => 'includes/diff/WikiDiff.php', + 'WikiDiff3' => 'includes/diff/WikiDiff3.php', + 'WordLevelDiff' => 'includes/diff/WikiDiff.php', # includes/filerepo 'ArchivedFile' => 'includes/filerepo/ArchivedFile.php', diff --git a/includes/diff/DifferenceEngine.php b/includes/diff/DifferenceEngine.php index c148d23853..d85dcfcf44 100644 --- a/includes/diff/DifferenceEngine.php +++ b/includes/diff/DifferenceEngine.php @@ -1,1241 +1,1072 @@ - * You may copy this code freely under the conditions of the GPL. + * User interface for the difference engine * * @file * @ingroup DifferenceEngine - * @defgroup DifferenceEngine DifferenceEngine */ - + /** - * @todo document - * @private - * @ingroup DifferenceEngine + * Constant to indicate diff cache compatibility. + * Bump this when changing the diff formatting in a way that + * fixes important bugs or such to force cached diff views to + * clear. */ -class _DiffOp { - var $type; - var $orig; - var $closing; - - function reverse() { - trigger_error('pure virtual', E_USER_ERROR); - } - - function norig() { - return $this->orig ? sizeof($this->orig) : 0; - } - - function nclosing() { - return $this->closing ? sizeof($this->closing) : 0; - } -} +define( 'MW_DIFF_VERSION', '1.11a' ); /** * @todo document - * @private * @ingroup DifferenceEngine */ -class _DiffOp_Copy extends _DiffOp { - var $type = 'copy'; - - function __construct ($orig, $closing = false) { - if (!is_array($closing)) - $closing = $orig; - $this->orig = $orig; - $this->closing = $closing; - } +class DifferenceEngine { + /**#@+ + * @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? + var $mCacheHit = false; // Was the diff fetched from cache? - function reverse() { - return new _DiffOp_Copy($this->closing, $this->orig); - } -} + /** + * Set this to true to add debug info to the HTML output. + * Warning: this may cause RSS readers to spuriously mark articles as "new" + * (bug 20601) + */ + var $enableDebugComment = false; -/** - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class _DiffOp_Delete extends _DiffOp { - var $type = 'delete'; + // If true, line X is not displayed when X is 1, for example to increase + // readability and conserve space with many small diffs. + protected $mReducedLineNumbers = false; - function __construct ($lines) { - $this->orig = $lines; - $this->closing = false; - } + protected $unhide = false; # show rev_deleted content if allowed + /**#@-*/ - function reverse() { - return new _DiffOp_Add($this->orig); + /** + * 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) + * @param $refreshCache boolean If set, refreshes the diff cache + * @param $unhide boolean If set, allow viewing deleted revs + */ + function __construct( $titleObj = null, $old = 0, $new = 0, $rcid = 0, + $refreshCache = false, $unhide = false ) + { + if ( $titleObj ) { + $this->mTitle = $titleObj; + } else { + global $wgTitle; + $this->mTitle = $wgTitle; + } + wfDebug("DifferenceEngine old '$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 next one. + # Get next 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); + wfRunHooks( 'NewDifferenceEngine', array(&$titleObj, &$this->mOldid, &$this->mNewid, $old, $new) ); + } + $this->mRcidMarkPatrolled = intval($rcid); # force it to be an integer + $this->mRefreshCache = $refreshCache; + $this->unhide = $unhide; } -} - -/** - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class _DiffOp_Add extends _DiffOp { - var $type = 'add'; - function __construct ($lines) { - $this->closing = $lines; - $this->orig = false; + function setReducedLineNumbers( $value = true ) { + $this->mReducedLineNumbers = $value; } - function reverse() { - return new _DiffOp_Delete($this->closing); + function getTitle() { + return $this->mTitle; } -} - -/** - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class _DiffOp_Change extends _DiffOp { - var $type = 'change'; - - function __construct ($orig, $closing) { - $this->orig = $orig; - $this->closing = $closing; + + function wasCacheHit() { + return $this->mCacheHit; } - - function reverse() { - return new _DiffOp_Change($this->closing, $this->orig); + + function getOldid() { + return $this->mOldid; + } + + function getNewid() { + return $this->mNewid; } -} - -/** - * Class used internally by Diff to actually compute the diffs. - * - * The algorithm used here is mostly lifted from the perl module - * Algorithm::Diff (version 1.06) by Ned Konz, which is available at: - * http://www.perl.com/CPAN/authors/id/N/NE/NEDKONZ/Algorithm-Diff-1.06.zip - * - * More ideas are taken from: - * http://www.ics.uci.edu/~eppstein/161/960229.html - * - * Some ideas are (and a bit of code) are from from analyze.c, from GNU - * diffutils-2.7, which can be found at: - * ftp://gnudist.gnu.org/pub/gnu/diffutils/diffutils-2.7.tar.gz - * - * closingly, some ideas (subdivision by NCHUNKS > 2, and some optimizations) - * are my own. - * - * Line length limits for robustness added by Tim Starling, 2005-08-31 - * Alternative implementation added by Guy Van den Broeck, 2008-07-30 - * - * @author Geoffrey T. Dairiki, Tim Starling, Guy Van den Broeck - * @private - * @ingroup DifferenceEngine - */ -class _DiffEngine { - - const MAX_XREF_LENGTH = 10000; - function diff ($from_lines, $to_lines){ + function showDiffPage( $diffOnly = false ) { + global $wgUser, $wgOut, $wgUseExternalEditor, $wgUseRCPatrol; wfProfileIn( __METHOD__ ); - // Diff and store locally - $this->diff_local($from_lines, $to_lines); - - // Merge edits when possible - $this->_shift_boundaries($from_lines, $this->xchanged, $this->ychanged); - $this->_shift_boundaries($to_lines, $this->ychanged, $this->xchanged); - - // Compute the edit operations. - $n_from = sizeof($from_lines); - $n_to = sizeof($to_lines); - - $edits = array(); - $xi = $yi = 0; - while ($xi < $n_from || $yi < $n_to) { - assert($yi < $n_to || $this->xchanged[$xi]); - assert($xi < $n_from || $this->ychanged[$yi]); - - // Skip matching "snake". - $copy = array(); - while ( $xi < $n_from && $yi < $n_to - && !$this->xchanged[$xi] && !$this->ychanged[$yi]) { - $copy[] = $from_lines[$xi++]; - ++$yi; - } - if ($copy) - $edits[] = new _DiffOp_Copy($copy); - - // Find deletes & adds. - $delete = array(); - while ($xi < $n_from && $this->xchanged[$xi]) - $delete[] = $from_lines[$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) - $edits[] = new _DiffOp_Delete($delete); - elseif ($add) - $edits[] = new _DiffOp_Add($add); + + # 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( array( + 'action' => 'raw', + 'oldid' => $this->mOldid + ) ); + $url2=$this->mTitle->getFullURL( array( + 'action' => 'raw', + 'oldid' => $this->mNewid + ) ); + $special=$wgLang->getNsText(NS_SPECIAL); + $control=<<diff($from_lines, $to_lines); - $this->xchanged = $wikidiff3->removed; - $this->ychanged = $wikidiff3->added; - unset($wikidiff3); - }else{ - // old diff - $n_from = sizeof($from_lines); - $n_to = sizeof($to_lines); - $this->xchanged = $this->ychanged = array(); - $this->xv = $this->yv = array(); - $this->xind = $this->yind = array(); - unset($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]) - 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]) - break; - $this->xchanged[$xi] = $this->ychanged[$yi] = false; - } + $wgOut->setArticleFlag( false ); + if ( !$this->loadRevisionData() ) { + $t = $this->mTitle->getPrefixedText(); + $d = wfMsgExt( 'missingarticle-diff', array( 'escape' ), $this->mOldid, $this->mNewid ); + $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) ); + $wgOut->addWikiMsg( 'missing-article', "$t", $d ); + wfProfileOut( __METHOD__ ); + return; + } - // Ignore lines which do not exist in both files. - for ($xi = $skip; $xi < $n_from - $endskip; $xi++) { - $xhash[$this->_line_hash($from_lines[$xi])] = 1; - } + wfRunHooks( 'DiffViewHeader', array( $this, $this->mOldRev, $this->mNewRev ) ); - for ($yi = $skip; $yi < $n_to - $endskip; $yi++) { - $line = $to_lines[$yi]; - if ( ($this->ychanged[$yi] = empty($xhash[$this->_line_hash($line)])) ) - continue; - $yhash[$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($yhash[$this->_line_hash($line)])) ) - continue; - $this->xv[] = $line; - $this->xind[] = $xi; - } + if ( $this->mNewRev->isCurrent() ) { + $wgOut->setArticleFlag( true ); + } - // Find the LCS. - $this->_compareseq(0, sizeof($this->xv), 0, sizeof($this->yv)); + # 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(); + $this->renderNewRevision(); // should we respect $diffOnly here or not? + wfProfileOut( __METHOD__ ); + return; } - wfProfileOut( __METHOD__ ); - } - /** - * Returns the whole line if it's small enough, or the MD5 hash otherwise - */ - function _line_hash( $line ) { - if ( strlen( $line ) > self::MAX_XREF_LENGTH ) { - return md5( $line ); + $wgOut->suppressQuickbar(); + + $oldTitle = $this->mOldPage->getPrefixedText(); + $newTitle = $this->mNewPage->getPrefixedText(); + if( $oldTitle == $newTitle ) { + $wgOut->setPageTitle( $newTitle ); } else { - return $line; + $wgOut->setPageTitle( $oldTitle . ', ' . $newTitle ); + } + if ( $this->mNewPage->equals( $this->mOldPage ) ) { + $wgOut->setSubtitle( wfMsgExt( 'difference', array( 'parseinline' ) ) ); + } else { + $wgOut->setSubtitle( wfMsgExt( 'difference-multipage', array( 'parseinline' ) ) ); + } + $wgOut->setRobotPolicy( 'noindex,nofollow' ); + + if ( !$this->mOldPage->userCanRead() || !$this->mNewPage->userCanRead() ) { + $wgOut->loginToUse(); + $wgOut->output(); + $wgOut->disable(); + wfProfileOut( __METHOD__ ); + return; } - } - /* Divide the Largest Common Subsequence (LCS) of the sequences - * [XOFF, XLIM) and [YOFF, YLIM) into NCHUNKS approximately equally - * sized segments. - * - * Returns (LCS, PTS). LCS is the length of the LCS. PTS is an - * array of NCHUNKS+1 (X, Y) indexes giving the diving points between - * sub sequences. The first sub-sequence is contained in [X0, X1), - * [Y0, Y1), the second in [X1, X2), [Y1, Y2) and so on. Note - * that (X0, Y0) == (XOFF, YOFF) and - * (X[NCHUNKS], Y[NCHUNKS]) == (XLIM, YLIM). - * - * This function assumes that the first lines of the specified portions - * of the two files do not match, and likewise that the last lines do not - * match. The caller must trim matching lines from the beginning and end - * of the portions it is going to specify. - */ - function _diag ($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); + $sk = $wgUser->getSkin(); + + // Check if page is editable + $editable = $this->mNewRev->getTitle()->userCan( 'edit' ); + if ( $editable && $this->mNewRev->isCurrent() && $wgUser->isAllowed( 'rollback' ) ) { + $rollback = '   ' . $sk->generateRollback( $this->mNewRev ); + } else { + $rollback = ''; } - if ($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++) { - $line = $flip ? $this->yv[$x] : $this->xv[$x]; - 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 ); - assert( $k > 0 ); - $ymids[$k] = $ymids[$k-1]; - break; - } - } - while (list ( /* $junk */, $y) = each($matches)) { - if ($y > $this->seq[$k-1]) { - 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); - assert($k > 0); - $ymids[$k] = $ymids[$k-1]; - } + // Prepare a change patrol link, if applicable + if( $wgUseRCPatrol && $this->mTitle->userCan('patrol') ) { + // If we've been given an explicit change identifier, use it; saves time + if( $this->mRcidMarkPatrolled ) { + $rcid = $this->mRcidMarkPatrolled; + $rc = RecentChange::newFromId( $rcid ); + // Already patrolled? + $rcid = is_object($rc) && !$rc->getAttribute('rc_patrolled') ? $rcid : 0; + } else { + // Look for an unpatrolled change corresponding to this diff + $db = wfGetDB( DB_SLAVE ); + $change = RecentChange::newFromConds( + array( + // Redundant user,timestamp condition so we can use the existing index + 'rc_user_text' => $this->mNewRev->getRawUserText(), + 'rc_timestamp' => $db->timestamp( $this->mNewRev->getTimestamp() ), + 'rc_this_oldid' => $this->mNewid, + 'rc_last_oldid' => $this->mOldid, + 'rc_patrolled' => 0 + ), + __METHOD__ + ); + if( $change instanceof RecentChange ) { + $rcid = $change->mAttribs['rc_id']; + $this->mRcidMarkPatrolled = $rcid; + } else { + // None found + $rcid = 0; } } + // Build the link + if( $rcid ) { + $token = $wgUser->editToken( $rcid ); + $patrol = ' [' . $sk->link( + $this->mTitle, + wfMsgHtml( 'markaspatrolleddiff' ), + array(), + array( + 'action' => 'markpatrolled', + 'rcid' => $rcid, + 'token' => $token, + ), + array( + 'known', + 'noclasses' + ) + ) . ']'; + } else { + $patrol = ''; + } + } else { + $patrol = ''; } - $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); + # Carry over 'diffonly' param via navigation links + if( $diffOnly != $wgUser->getBoolOption('diffonly') ) { + $query['diffonly'] = $diffOnly; } - $seps[] = $flip ? array($ylim, $xlim) : array($xlim, $ylim); - return array($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; + # Make "previous revision link" + $query['diff'] = 'prev'; + $query['oldid'] = $this->mOldid; + # Cascade unhide param in links for easy deletion browsing + if( $this->unhide ) { + $query['unhide'] = 1; } - - $beg = 1; - while ($beg < $end) { - $mid = (int)(($beg + $end) / 2); - if ( $ypos > $this->seq[$mid] ) - $beg = $mid + 1; - else - $end = $mid; + if( !$this->mOldRev->getPrevious() ) { + $prevlink = ' '; + } else { + $prevlink = $sk->link( + $this->mTitle, + wfMsgHtml( 'previousdiff' ), + array( + 'id' => 'differences-prevlink' + ), + $query, + array( + 'known', + 'noclasses' + ) + ); } - assert($ypos != $this->seq[$end]); + # Make "next revision link" + $query['diff'] = 'next'; + $query['oldid'] = $this->mNewid; + # Skip next link on the top revision + if( $this->mNewRev->isCurrent() ) { + $nextlink = ' '; + } else { + $nextlink = $sk->link( + $this->mTitle, + wfMsgHtml( 'nextdiff' ), + array( + 'id' => 'differences-nextlink' + ), + $query, + array( + 'known', + 'noclasses' + ) + ); + } - $this->in_seq[$this->seq[$end]] = false; - $this->seq[$end] = $ypos; - $this->in_seq[$ypos] = 1; - return $end; - } + $oldminor = ''; + $newminor = ''; - /* Find LCS of two sequences. - * - * The results are recorded in the vectors $this->{x,y}changed[], by - * storing a 1 in the element for each line that is an insertion - * or deletion (ie. is not in the LCS). - * - * The subsequence of file 0 is [XOFF, XLIM) and likewise for file 1. - * - * Note that XLIM, YLIM are exclusive bounds. - * All line numbers are origin-0 and discarded lines are not counted. - */ - function _compareseq ($xoff, $xlim, $yoff, $ylim) { - // Slide down the bottom initial diagonal. - while ($xoff < $xlim && $yoff < $ylim - && $this->xv[$xoff] == $this->yv[$yoff]) { - ++$xoff; - ++$yoff; + if( $this->mOldRev->isMinor() ) { + $oldminor = ChangesList::flag( 'minor' ); } - - // Slide up the top initial diagonal. - while ($xlim > $xoff && $ylim > $yoff - && $this->xv[$xlim - 1] == $this->yv[$ylim - 1]) { - --$xlim; - --$ylim; + if( $this->mNewRev->isMinor() ) { + $newminor = ChangesList::flag( 'minor' ); } - if ($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); + # Handle RevisionDelete links... + $ldel = $this->revisionDeleteLink( $this->mOldRev ); + $rdel = $this->revisionDeleteLink( $this->mNewRev ); + + $oldHeader = '
'.$this->mOldtitle.'
' . + '
' . + $sk->revUserTools( $this->mOldRev, !$this->unhide ).'
' . + '
' . $oldminor . + $sk->revComment( $this->mOldRev, !$diffOnly, !$this->unhide ).$ldel.'
' . + '
' . $prevlink .'
'; + $newHeader = '
'.$this->mNewtitle.'
' . + '
' . $sk->revUserTools( $this->mNewRev, !$this->unhide ) . + " $rollback
" . + '
' . $newminor . + $sk->revComment( $this->mNewRev, !$diffOnly, !$this->unhide ).$rdel.'
' . + '
' . $nextlink . $patrol . '
'; + + # Check if this user can see the revisions + $allowed = $this->mOldRev->userCan(Revision::DELETED_TEXT) + && $this->mNewRev->userCan(Revision::DELETED_TEXT); + # Check if one of the revisions is deleted/suppressed + $deleted = $suppressed = false; + if( $this->mOldRev->isDeleted(Revision::DELETED_TEXT) ) { + $deleted = true; // old revisions text is hidden + if( $this->mOldRev->isDeleted(Revision::DELETED_RESTRICTED) ) + $suppressed = true; // also suppressed } - - if ($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; + if( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { + $deleted = true; // new revisions text is hidden + if( $this->mNewRev->isDeleted(Revision::DELETED_RESTRICTED) ) + $suppressed = true; // also suppressed + } + # If the diff cannot be shown due to a deleted revision, then output + # the diff header and links to unhide (if available)... + if( $deleted && (!$this->unhide || !$allowed) ) { + $this->showDiffStyle(); + $multi = $this->getMultiNotice(); + $wgOut->addHTML( $this->addHeader( '', $oldHeader, $newHeader, $multi ) ); + if( !$allowed ) { + $msg = $suppressed ? 'rev-suppressed-no-diff' : 'rev-deleted-no-diff'; + # Give explanation for why revision is not visible + $wgOut->wrapWikiMsg( "\n", + array( $msg ) ); + } else { + # Give explanation and add a link to view the diff... + $link = $this->mTitle->getFullUrl( array( + 'diff' => $this->mNewid, + 'oldid' => $this->mOldid, + 'unhide' => 1 + ) ); + $msg = $suppressed ? 'rev-suppressed-unhide-diff' : 'rev-deleted-unhide-diff'; + $wgOut->wrapWikiMsg( "\n", array( $msg, $link ) ); + } + # Otherwise, output a regular diff... } else { - // Use the partitions to split this problem into subproblems. - reset($seps); - $pt1 = $seps[0]; - while ($pt2 = next($seps)) { - $this->_compareseq ($pt1[0], $pt2[0], $pt1[1], $pt2[1]); - $pt1 = $pt2; + # Add deletion notice if the user is viewing deleted content + $notice = ''; + if( $deleted ) { + $msg = $suppressed ? 'rev-suppressed-diff-view' : 'rev-deleted-diff-view'; + $notice = "\n"; + } + $this->showDiff( $oldHeader, $newHeader, $notice ); + if( !$diffOnly ) { + $this->renderNewRevision(); } } + wfProfileOut( __METHOD__ ); } - - /* Adjust inserts/deletes of identical lines to join changes - * as much as possible. - * - * We do something when a run of changed lines include a - * line at one end and has an excluded, identical line at the other. - * We are free to choose which identical line is included. - * `compareseq' usually chooses the one at the beginning, - * but usually it is cleaner to consider the following identical line - * to be the "change". - * - * This is extracted verbatim from analyze.c (GNU diffutils-2.7). - */ - function _shift_boundaries ($lines, &$changed, $other_changed) { - wfProfileIn( __METHOD__ ); - $i = 0; - $j = 0; - - assert('sizeof($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]) { - 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]) - $start--; - assert('$j > 0'); - while ($other_changed[--$j]) - continue; - 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]) - $i++; - - assert('$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; - assert('$j > 0'); - while ($other_changed[--$j]) - continue; - assert('$j >= 0 && !$other_changed[$j]'); + + protected function revisionDeleteLink( $rev ) { + global $wgUser; + $link = ''; + $canHide = $wgUser->isAllowed( 'deleterevision' ); + // Show del/undel link if: + // (a) the user can delete revisions, or + // (b) the user can view deleted revision *and* this one is deleted + if( $canHide || ($rev->getVisibility() && $wgUser->isAllowed( 'deletedhistory' )) ) { + $sk = $wgUser->getSkin(); + if( !$rev->userCan( Revision::DELETED_RESTRICTED ) ) { + $link = $sk->revDeleteLinkDisabled( $canHide ); // revision was hidden from sysops + } else { + $query = array( + 'type' => 'revision', + 'target' => $rev->mTitle->getPrefixedDbkey(), + 'ids' => $rev->getId() + ); + $link = $sk->revDeleteLink( $query, + $rev->isDeleted( Revision::DELETED_RESTRICTED ), $canHide ); } + $link = '   ' . $link . ' '; } - wfProfileOut( __METHOD__ ); + return $link; } -} - -/** - * Class representing a 'diff' between two sequences of strings. - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class Diff -{ - var $edits; /** - * Constructor. - * Computes diff between sequences of strings. - * - * @param $from_lines array An array of strings. - * (Typically these are lines from a file.) - * @param $to_lines array An array of strings. + * Show the new revision of the page. */ - function __construct($from_lines, $to_lines) { - $eng = new _DiffEngine; - $this->edits = $eng->diff($from_lines, $to_lines); - //$this->_check($from_lines, $to_lines); - } + function renderNewRevision() { + global $wgOut, $wgUser; + wfProfileIn( __METHOD__ ); - /** - * Compute reversed Diff. - * - * SYNOPSIS: - * - * $diff = new Diff($lines1, $lines2); - * $rev = $diff->reverse(); - * @return object A Diff object representing the inverse of the - * original diff. - */ - function reverse () { - $rev = $this; - $rev->edits = array(); - foreach ($this->edits as $edit) { - $rev->edits[] = $edit->reverse(); + $wgOut->addHTML( "

{$this->mPagetitle}

\n" ); + # Add deleted rev tag if needed + if( !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { + $wgOut->wrapWikiMsg( "\n", 'rev-deleted-text-permission' ); + } else if( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { + $wgOut->wrapWikiMsg( "\n", 'rev-deleted-text-view' ); } - return $rev; - } - /** - * Check for empty diff. - * - * @return bool True iff two sequences were identical. - */ - function isEmpty () { - foreach ($this->edits as $edit) { - if ($edit->type != 'copy') - return false; + $pCache = true; + if( !$this->mNewRev->isCurrent() ) { + $oldEditSectionSetting = $wgOut->parserOptions()->setEditSection( false ); + $pCache = false; } - return true; - } - /** - * Compute the length of the Longest Common Subsequence (LCS). - * - * This is mostly for diagnostic purposed. - * - * @return int The length of the LCS. - */ - function lcs () { - $lcs = 0; - foreach ($this->edits as $edit) { - if ($edit->type == 'copy') - $lcs += sizeof($edit->orig); + $this->loadNewText(); + if( is_object( $this->mNewRev ) ) { + $wgOut->setRevisionId( $this->mNewRev->getId() ); } - return $lcs; - } - /** - * Get the original set of lines. - * - * This reconstructs the $from_lines parameter passed to the - * constructor. - * - * @return array The original sequence of strings. - */ - function orig() { - $lines = array(); - - foreach ($this->edits as $edit) { - if ($edit->orig) - array_splice($lines, sizeof($lines), 0, $edit->orig); + if( $this->mTitle->isCssJsSubpage() || $this->mTitle->isCssOrJsPage() ) { + // Stolen from Article::view --AG 2007-10-11 + // Give hooks a chance to customise the output + if( wfRunHooks( 'ShowRawCssJs', array( $this->mNewtext, $this->mTitle, $wgOut ) ) ) { + // Wrap the whole lot in a
 and don't parse
+				$m = array();
+				preg_match( '!\.(css|js)$!u', $this->mTitle->getText(), $m );
+				$wgOut->addHTML( "
\n" );
+				$wgOut->addHTML( htmlspecialchars( $this->mNewtext ) );
+				$wgOut->addHTML( "\n
\n" ); + } + } elseif( wfRunHooks( 'ArticleContentOnDiff', array( $this, $wgOut ) ) ) { + if ( $pCache ) { + $article = new Article( $this->mTitle, 0 ); + $pOutput = ParserCache::singleton()->get( $article, $wgOut->parserOptions() ); + if( $pOutput ) { + $wgOut->addParserOutput( $pOutput ); + } else { + $article->doViewParse(); + } + } else { + $wgOut->addWikiTextTidy( $this->mNewtext ); + } } - return $lines; - } - - /** - * Get the closing set of lines. - * - * This reconstructs the $to_lines parameter passed to the - * constructor. - * - * @return array The sequence of strings. - */ - function closing() { - $lines = array(); - - foreach ($this->edits as $edit) { - if ($edit->closing) - array_splice($lines, sizeof($lines), 0, $edit->closing); + + if( is_object( $this->mNewRev ) && !$this->mNewRev->isCurrent() ) { + $wgOut->parserOptions()->setEditSection( $oldEditSectionSetting ); } - return $lines; - } - /** - * Check a Diff for validity. - * - * This is here only for debugging purposes. - */ - function _check ($from_lines, $to_lines) { - wfProfileIn( __METHOD__ ); - if (serialize($from_lines) != serialize($this->orig())) - trigger_error("Reconstructed original doesn't match", E_USER_ERROR); - if (serialize($to_lines) != serialize($this->closing())) - trigger_error("Reconstructed closing doesn't match", E_USER_ERROR); - - $rev = $this->reverse(); - if (serialize($to_lines) != serialize($rev->orig())) - trigger_error("Reversed original doesn't match", E_USER_ERROR); - if (serialize($from_lines) != serialize($rev->closing())) - trigger_error("Reversed closing doesn't match", E_USER_ERROR); - - - $prevtype = 'none'; - foreach ($this->edits as $edit) { - if ( $prevtype == $edit->type ) - trigger_error("Edit sequence is non-optimal", E_USER_ERROR); - $prevtype = $edit->type; + # Add redundant patrol link on bottom... + if( $this->mRcidMarkPatrolled && $this->mTitle->quickUserCan('patrol') ) { + $sk = $wgUser->getSkin(); + $token = $wgUser->editToken( $this->mRcidMarkPatrolled ); + $wgOut->addHTML( + "' + ); } - $lcs = $this->lcs(); - trigger_error('Diff okay: LCS = '.$lcs, E_USER_NOTICE); wfProfileOut( __METHOD__ ); } -} -/** - * @todo document, bad name. - * @private - * @ingroup DifferenceEngine - */ -class MappedDiff extends Diff -{ /** - * Constructor. - * - * Computes diff between sequences of strings. - * - * This can be used to compute things like - * case-insensitve diffs, or diffs which ignore - * changes in white-space. - * - * @param $from_lines array An array of strings. - * (Typically these are lines from a file.) - * - * @param $to_lines array An array of strings. - * - * @param $mapped_from_lines array This array should - * have the same size number of elements as $from_lines. - * The elements in $mapped_from_lines and - * $mapped_to_lines are what is actually compared - * when computing the diff. - * - * @param $mapped_to_lines array This array should - * have the same number of elements as $to_lines. + * Show the first revision of an article. Uses normal diff headers in + * contrast to normal "old revision" display style. */ - function __construct($from_lines, $to_lines, - $mapped_from_lines, $mapped_to_lines) { + function showFirstRevision() { + global $wgOut, $wgUser; wfProfileIn( __METHOD__ ); - assert(sizeof($from_lines) == sizeof($mapped_from_lines)); - assert(sizeof($to_lines) == sizeof($mapped_to_lines)); + # Get article text from the DB + # + if ( ! $this->loadNewText() ) { + $t = $this->mTitle->getPrefixedText(); + $d = wfMsgExt( 'missingarticle-diff', array( 'escape' ), $this->mOldid, $this->mNewid ); + $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) ); + $wgOut->addWikiMsg( 'missing-article', "$t", $d ); + wfProfileOut( __METHOD__ ); + return; + } + if ( $this->mNewRev->isCurrent() ) { + $wgOut->setArticleFlag( true ); + } - parent::__construct($mapped_from_lines, $mapped_to_lines); + # Check if user is allowed to look at this page. If not, bail out. + # + if ( !$this->mTitle->userCanRead() ) { + $wgOut->loginToUse(); + $wgOut->output(); + wfProfileOut( __METHOD__ ); + throw new MWException("Permission Error: you do not have access to view this page"); + } - $xi = $yi = 0; - for ($i = 0; $i < sizeof($this->edits); $i++) { - $orig = &$this->edits[$i]->orig; - if (is_array($orig)) { - $orig = array_slice($from_lines, $xi, sizeof($orig)); - $xi += sizeof($orig); - } + # Prepare the header box + # + $sk = $wgUser->getSkin(); - $closing = &$this->edits[$i]->closing; - if (is_array($closing)) { - $closing = array_slice($to_lines, $yi, sizeof($closing)); - $yi += sizeof($closing); - } + $next = $this->mTitle->getNextRevisionID( $this->mNewid ); + if( !$next ) { + $nextlink = ''; + } else { + $nextlink = '
' . $sk->link( + $this->mTitle, + wfMsgHtml( 'nextdiff' ), + array( + 'id' => 'differences-nextlink' + ), + array( + 'diff' => 'next', + 'oldid' => $this->mNewid, + ), + array( + 'known', + 'noclasses' + ) + ); } + $header = "
" . + $sk->revUserTools( $this->mNewRev ) . "
" . $sk->revComment( $this->mNewRev ) . $nextlink . "
\n"; + + $wgOut->addHTML( $header ); + + $wgOut->setSubtitle( wfMsgExt( 'difference', array( 'parseinline' ) ) ); + $wgOut->setRobotPolicy( 'noindex,nofollow' ); + wfProfileOut( __METHOD__ ); } -} -/** - * A class to format Diffs - * - * 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 - * @ingroup DifferenceEngine - */ -class DiffFormatter { /** - * Number of leading context "lines" to preserve. - * - * This should be left at zero for this class, but subclasses - * may want to set this to other values. + * Get the diff text, send it to $wgOut + * Returns false if the diff could not be generated, otherwise returns true */ - var $leading_context_lines = 0; + function showDiff( $otitle, $ntitle, $notice = '' ) { + global $wgOut; + $diff = $this->getDiff( $otitle, $ntitle, $notice ); + if ( $diff === false ) { + $wgOut->addWikiMsg( 'missing-article', "(fixme, bug)", '' ); + return false; + } else { + $this->showDiffStyle(); + $wgOut->addHTML( $diff ); + return true; + } + } /** - * Number of trailing context "lines" to preserve. + * Add style sheets and supporting JS for diff display. + */ + function showDiffStyle() { + global $wgOut; + $wgOut->addModules( 'mediawiki.legacy.diff' ); + } + + /** + * Get complete diff table, including header * - * This should be left at zero for this class, but subclasses - * may want to set this to other values. + * @param $otitle Title: old title + * @param $ntitle Title: new title + * @param $notice String: HTML between diff header and body + * @return mixed */ - var $trailing_context_lines = 0; + function getDiff( $otitle, $ntitle, $notice = '' ) { + $body = $this->getDiffBody(); + if ( $body === false ) { + return false; + } else { + $multi = $this->getMultiNotice(); + return $this->addHeader( $body, $otitle, $ntitle, $multi, $notice ); + } + } /** - * Format a diff. + * Get the diff table body, without header * - * @param $diff object A Diff object. - * @return string The formatted output. + * @return mixed */ - function format($diff) { + function getDiffBody() { + global $wgMemc; wfProfileIn( __METHOD__ ); - - $xi = $yi = 1; - $block = false; - $context = array(); - - $nlead = $this->leading_context_lines; - $ntrail = $this->trailing_context_lines; - - $this->_start_diff(); - - foreach ($diff->edits as $edit) { - if ($edit->type == 'copy') { - if (is_array($block)) { - if (sizeof($edit->orig) <= $nlead + $ntrail) { - $block[] = $edit; - } - else{ - if ($ntrail) { - $context = array_slice($edit->orig, 0, $ntrail); - $block[] = new _DiffOp_Copy($context); - } - $this->_block($x0, $ntrail + $xi - $x0, - $y0, $ntrail + $yi - $y0, - $block); - $block = false; - } - } - $context = $edit->orig; - } - else { - if (! is_array($block)) { - $context = array_slice($context, sizeof($context) - $nlead); - $x0 = $xi - sizeof($context); - $y0 = $yi - sizeof($context); - $block = array(); - if ($context) - $block[] = new _DiffOp_Copy($context); + $this->mCacheHit = true; + // Check if the diff should be hidden from this user + if ( !$this->loadRevisionData() ) + return ''; + if ( $this->mOldRev && !$this->mOldRev->userCan(Revision::DELETED_TEXT) ) { + return ''; + } else if ( $this->mNewRev && !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { + return ''; + } else if ( $this->mOldRev && $this->mNewRev && $this->mOldRev->getID() == $this->mNewRev->getID() ) { + return ''; + } + // Cacheable? + $key = false; + if ( $this->mOldid && $this->mNewid ) { + $key = wfMemcKey( 'diff', 'version', MW_DIFF_VERSION, 'oldid', $this->mOldid, 'newid', $this->mNewid ); + // Try cache + if ( !$this->mRefreshCache ) { + $difftext = $wgMemc->get( $key ); + if ( $difftext ) { + wfIncrStats( 'diff_cache_hit' ); + $difftext = $this->localiseLineNumbers( $difftext ); + $difftext .= "\n\n"; + wfProfileOut( __METHOD__ ); + return $difftext; } - $block[] = $edit; - } - - if ($edit->orig) - $xi += sizeof($edit->orig); - if ($edit->closing) - $yi += sizeof($edit->closing); + } // don't try to load but save the result } + $this->mCacheHit = false; - if (is_array($block)) - $this->_block($x0, $xi - $x0, - $y0, $yi - $y0, - $block); + // Loadtext is permission safe, this just clears out the diff + if ( !$this->loadText() ) { + wfProfileOut( __METHOD__ ); + return false; + } - $end = $this->_end_diff(); - wfProfileOut( __METHOD__ ); - return $end; - } + $difftext = $this->generateDiffBody( $this->mOldtext, $this->mNewtext ); - function _block($xbeg, $xlen, $ybeg, $ylen, &$edits) { - wfProfileIn( __METHOD__ ); - $this->_start_block($this->_block_header($xbeg, $xlen, $ybeg, $ylen)); - foreach ($edits as $edit) { - if ($edit->type == 'copy') - $this->_context($edit->orig); - elseif ($edit->type == 'add') - $this->_added($edit->closing); - elseif ($edit->type == 'delete') - $this->_deleted($edit->orig); - elseif ($edit->type == 'change') - $this->_changed($edit->orig, $edit->closing); - else - trigger_error('Unknown edit type', E_USER_ERROR); + // Save to cache for 7 days + if ( !wfRunHooks( 'AbortDiffCache', array( &$this ) ) ) { + wfIncrStats( 'diff_uncacheable' ); + } else 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 ); } - $this->_end_block(); wfProfileOut( __METHOD__ ); + return $difftext; } - function _start_diff() { - ob_start(); - } - - function _end_diff() { - $val = ob_get_contents(); - ob_end_clean(); - return $val; - } - - function _block_header($xbeg, $xlen, $ybeg, $ylen) { - if ($xlen > 1) - $xbeg .= "," . ($xbeg + $xlen - 1); - if ($ylen > 1) - $ybeg .= "," . ($ybeg + $ylen - 1); - - return $xbeg . ($xlen ? ($ylen ? 'c' : 'd') : 'a') . $ybeg; - } - - function _start_block($header) { - echo $header . "\n"; - } - - function _end_block() { - } - - function _lines($lines, $prefix = ' ') { - foreach ($lines as $line) - echo "$prefix $line\n"; - } - - function _context($lines) { - $this->_lines($lines); - } - - function _added($lines) { - $this->_lines($lines, '>'); - } - function _deleted($lines) { - $this->_lines($lines, '<'); - } - - function _changed($orig, $closing) { - $this->_deleted($orig); - echo "---\n"; - $this->_added($closing); + /** + * Make sure the proper modules are loaded before we try to + * make the diff + */ + private function initDiffEngines() { + global $wgExternalDiffEngine; + if ( $wgExternalDiffEngine == 'wikidiff' && !function_exists( 'wikidiff_do_diff' ) ) { + wfProfileIn( __METHOD__ . '-php_wikidiff.so' ); + wfSuppressWarnings(); + dl( 'php_wikidiff.so' ); + wfRestoreWarnings(); + wfProfileOut( __METHOD__ . '-php_wikidiff.so' ); + } + else if ( $wgExternalDiffEngine == 'wikidiff2' && !function_exists( 'wikidiff2_do_diff' ) ) { + wfProfileIn( __METHOD__ . '-php_wikidiff2.so' ); + wfSuppressWarnings(); + wfDl( 'wikidiff2' ); + wfRestoreWarnings(); + wfProfileOut( __METHOD__ . '-php_wikidiff2.so' ); + } } -} -/** - * A formatter that outputs unified diffs - * @ingroup DifferenceEngine - */ + /** + * Generate a diff, no caching + * + * @param $otext String: old text, must be already segmented + * @param $ntext String: new text, must be already segmented + */ + function generateDiffBody( $otext, $ntext ) { + global $wgExternalDiffEngine, $wgContLang; -class UnifiedDiffFormatter extends DiffFormatter { - var $leading_context_lines = 2; - var $trailing_context_lines = 2; + $otext = str_replace( "\r\n", "\n", $otext ); + $ntext = str_replace( "\r\n", "\n", $ntext ); - function _added($lines) { - $this->_lines($lines, '+'); - } - function _deleted($lines) { - $this->_lines($lines, '-'); - } - function _changed($orig, $closing) { - $this->_deleted($orig); - $this->_added($closing); - } - function _block_header($xbeg, $xlen, $ybeg, $ylen) { - return "@@ -$xbeg,$xlen +$ybeg,$ylen @@"; - } -} + $this->initDiffEngines(); -/** - * A pseudo-formatter that just passes along the Diff::$edits array - * @ingroup DifferenceEngine - */ -class ArrayDiffFormatter extends DiffFormatter { - function format($diff) { - $oldline = 1; - $newline = 1; - $retval = array(); - foreach($diff->edits as $edit) - switch($edit->type) { - case 'add': - foreach($edit->closing as $l) { - $retval[] = array( - 'action' => 'add', - 'new'=> $l, - 'newline' => $newline++ - ); - } - break; - case 'delete': - foreach($edit->orig as $l) { - $retval[] = array( - 'action' => 'delete', - 'old' => $l, - 'oldline' => $oldline++, - ); - } - break; - case 'change': - foreach($edit->orig as $i => $l) { - $retval[] = array( - 'action' => 'change', - 'old' => $l, - 'new' => @$edit->closing[$i], - 'oldline' => $oldline++, - 'newline' => $newline++, - ); - } - break; - case 'copy': - $oldline += count($edit->orig); - $newline += count($edit->orig); + if ( $wgExternalDiffEngine == 'wikidiff' && function_exists( 'wikidiff_do_diff' ) ) { + # For historical reasons, external diff engine expects + # input text to be HTML-escaped already + $otext = htmlspecialchars ( $wgContLang->segmentForDiff( $otext ) ); + $ntext = htmlspecialchars ( $wgContLang->segmentForDiff( $ntext ) ); + return $wgContLang->unsegementForDiff( wikidiff_do_diff( $otext, $ntext, 2 ) ) . + $this->debug( 'wikidiff1' ); } - return $retval; - } -} - -/** - * Additions by Axel Boldt follow, partly taken from diff.php, phpwiki-1.3.3 - * - */ -define('NBSP', ' '); // iso-8859-x non-breaking space. - -/** - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class _HWLDF_WordAccumulator { - function __construct () { - $this->_lines = array(); - $this->_line = ''; - $this->_group = ''; - $this->_tag = ''; - } - - function _flushGroup ($new_tag) { - if ($this->_group !== '') { - if ($this->_tag == 'ins') - $this->_line .= '' . - htmlspecialchars ( $this->_group ) . ''; - elseif ($this->_tag == 'del') - $this->_line .= '' . - htmlspecialchars ( $this->_group ) . ''; - else - $this->_line .= htmlspecialchars ( $this->_group ); + if ( $wgExternalDiffEngine == 'wikidiff2' && function_exists( 'wikidiff2_do_diff' ) ) { + # Better external diff engine, the 2 may some day be dropped + # This one does the escaping and segmenting itself + wfProfileIn( 'wikidiff2_do_diff' ); + $text = wikidiff2_do_diff( $otext, $ntext, 2 ); + $text .= $this->debug( 'wikidiff2' ); + wfProfileOut( 'wikidiff2_do_diff' ); + return $text; + } + if ( $wgExternalDiffEngine != 'wikidiff3' && $wgExternalDiffEngine !== false ) { + # Diff via the shell + global $wgTmpDirectory; + $tempName1 = tempnam( $wgTmpDirectory, 'diff_' ); + $tempName2 = tempnam( $wgTmpDirectory, 'diff_' ); + + $tempFile1 = fopen( $tempName1, "w" ); + if ( !$tempFile1 ) { + wfProfileOut( __METHOD__ ); + return false; + } + $tempFile2 = fopen( $tempName2, "w" ); + if ( !$tempFile2 ) { + wfProfileOut( __METHOD__ ); + return false; + } + fwrite( $tempFile1, $otext ); + fwrite( $tempFile2, $ntext ); + fclose( $tempFile1 ); + fclose( $tempFile2 ); + $cmd = wfEscapeShellArg( $wgExternalDiffEngine, $tempName1, $tempName2 ); + wfProfileIn( __METHOD__ . "-shellexec" ); + $difftext = wfShellExec( $cmd ); + $difftext .= $this->debug( "external $wgExternalDiffEngine" ); + wfProfileOut( __METHOD__ . "-shellexec" ); + unlink( $tempName1 ); + unlink( $tempName2 ); + return $difftext; } - $this->_group = ''; - $this->_tag = $new_tag; - } - function _flushLine ($new_tag) { - $this->_flushGroup($new_tag); - if ($this->_line != '') - array_push ( $this->_lines, $this->_line ); - else - # make empty lines visible by inserting an NBSP - array_push ( $this->_lines, NBSP ); - $this->_line = ''; + # 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 ) ) . + $this->debug(); } - function addWords ($words, $tag = '') { - if ($tag != $this->_tag) - $this->_flushGroup($tag); - - foreach ($words as $word) { - // new-line should only come as first char of word. - if ($word == '') - continue; - if ($word[0] == "\n") { - $this->_flushLine($tag); - $word = substr($word, 1); - } - assert(!strstr($word, "\n")); - $this->_group .= $word; + /** + * Generate a debug comment indicating diff generating time, + * server node, and generator backend. + */ + protected function debug( $generator="internal" ) { + global $wgShowHostnames; + if ( !$this->enableDebugComment ) { + return ''; } + $data = array( $generator ); + if( $wgShowHostnames ) { + $data[] = wfHostname(); + } + $data[] = wfTimestamp( TS_DB ); + return "\n"; } - function getLines() { - $this->_flushLine('~done'); - return preg_replace( '/\n/m', '', $this->_lines); + /** + * Replace line numbers with the text in the user's language + */ + function localiseLineNumbers( $text ) { + return preg_replace_callback( '//', + array( &$this, 'localiseLineNumbersCb' ), $text ); } -} -/** - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class WordLevelDiff extends MappedDiff { - const MAX_LINE_LENGTH = 10000; + function localiseLineNumbersCb( $matches ) { + global $wgLang; + if ( $matches[1] === '1' && $this->mReducedLineNumbers ) return ''; + return wfMsgExt( 'lineno', 'escape', $wgLang->formatNum( $matches[1] ) ); + } - function __construct ($orig_lines, $closing_lines) { - wfProfileIn( __METHOD__ ); - list ($orig_words, $orig_stripped) = $this->_split($orig_lines); - list ($closing_words, $closing_stripped) = $this->_split($closing_lines); + /** + * If there are revisions between the ones being compared, return a note saying so. + */ + function getMultiNotice() { + if ( !is_object($this->mOldRev) || !is_object($this->mNewRev) ) + return ''; - parent::__construct($orig_words, $closing_words, - $orig_stripped, $closing_stripped); - wfProfileOut( __METHOD__ ); - } + if( !$this->mOldPage->equals( $this->mNewPage ) ) { + // Comparing two different pages? Count would be meaningless. + return ''; + } - function _split($lines) { - wfProfileIn( __METHOD__ ); + $oldid = $this->mOldRev->getId(); + $newid = $this->mNewRev->getId(); + if ( $oldid > $newid ) { + $tmp = $oldid; $oldid = $newid; $newid = $tmp; + } - $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 ) > self::MAX_LINE_LENGTH ) { - $words[] = $line; - $stripped[] = $line; + $n = $this->mTitle->countRevisionsBetween( $oldid, $newid ); + if ( !$n ) { + return ''; + } else { + global $wgLang; + $dbr = wfGetDB( DB_SLAVE ); + + // Actually, the limit is $limit + 1. We do this so we can detect + // if there are > 100 authors in a given revision range. If they + // are, $limit will be passed to diff-multi-manyusers for l10n. + $limit = 100; + $res = $dbr->select( 'revision', 'DISTINCT rev_user_text', + array( + 'rev_page = ' . $this->mOldRev->getPage(), + 'rev_id > ' . $this->mOldRev->getId(), + 'rev_id < ' . $this->mNewRev->getId() + ), __METHOD__, + array( 'LIMIT' => $limit + 1 ) + ); + $numUsers = $dbr->numRows( $res ); + if( $numUsers > $limit ) { + $msg = 'diff-multi-manyusers'; + $numUsers = $limit; } else { - $m = array(); - 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] ); - } + $msg = 'diff-multi'; } + return wfMsgExt( $msg, array( 'parseinline' ), $wgLang->formatnum( $n ), + $wgLang->formatnum( $numUsers ) ); } - wfProfileOut( __METHOD__ ); - return array($words, $stripped); } - function orig () { - wfProfileIn( __METHOD__ ); - $orig = new _HWLDF_WordAccumulator; - foreach ($this->edits as $edit) { - if ($edit->type == 'copy') - $orig->addWords($edit->orig); - elseif ($edit->orig) - $orig->addWords($edit->orig, 'del'); + /** + * Add the header to a diff body + */ + static function addHeader( $diff, $otitle, $ntitle, $multi = '', $notice = '' ) { + $header = ""; + if( $diff ) { // Safari/Chrome show broken output if cols not used + $header .= " + + + + "; + $colspan = 2; + $multiColspan = 4; + } else { + $colspan = 1; + $multiColspan = 2; } - $lines = $orig->getLines(); - wfProfileOut( __METHOD__ ); - return $lines; - } - - function closing () { - wfProfileIn( __METHOD__ ); - $closing = new _HWLDF_WordAccumulator; - - foreach ($this->edits as $edit) { - if ($edit->type == 'copy') - $closing->addWords($edit->closing); - elseif ($edit->closing) - $closing->addWords($edit->closing, 'ins'); + $header .= " + + + + "; + + if ( $multi != '' ) { + $header .= ""; + } + if ( $notice != '' ) { + $header .= ""; } - $lines = $closing->getLines(); - wfProfileOut( __METHOD__ ); - return $lines; - } -} - -/** - * Wikipedia Table style diff formatter. - * @todo document - * @private - * @ingroup DifferenceEngine - */ -class TableDiffFormatter extends DiffFormatter { - function __construct() { - $this->leading_context_lines = 2; - $this->trailing_context_lines = 2; - } - public static function escapeWhiteSpace( $msg ) { - $msg = preg_replace( '/^ /m', '  ', $msg ); - $msg = preg_replace( '/ $/m', '  ', $msg ); - $msg = preg_replace( '/ /', '  ', $msg ); - return $msg; + return $header . $diff . "
{$otitle}{$ntitle}
{$multi}
{$notice}
"; } - function _block_header( $xbeg, $xlen, $ybeg, $ylen ) { - $r = '\n" . - '\n"; - return $r; + /** + * Use specified text instead of loading from the database + */ + function setText( $oldText, $newText ) { + $this->mOldtext = $oldText; + $this->mNewtext = $newText; + $this->mTextLoaded = 2; + $this->mRevisionsLoaded = true; } - function _start_block( $header ) { - echo $header; - } + /** + * 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, $wgUser; + if ( $this->mRevisionsLoaded ) { + return true; + } else { + // Whether it succeeds or fails, we don't want to try again + $this->mRevisionsLoaded = true; + } - function _end_block() { - } + // Load the new revision object + $this->mNewRev = $this->mNewid + ? Revision::newFromId( $this->mNewid ) + : Revision::newFromTitle( $this->mTitle ); + if( !$this->mNewRev instanceof Revision ) + return false; - function _lines( $lines, $prefix=' ', $color='white' ) { - } + // Update the new revision ID in case it was 0 (makes life easier doing UI stuff) + $this->mNewid = $this->mNewRev->getId(); + + // Check if page is editable + $editable = $this->mNewRev->getTitle()->userCan( 'edit' ); + + // Set assorted variables + $timestamp = $wgLang->timeanddate( $this->mNewRev->getTimestamp(), true ); + $dateofrev = $wgLang->date( $this->mNewRev->getTimestamp(), true ); + $timeofrev = $wgLang->time( $this->mNewRev->getTimestamp(), true ); + $this->mNewPage = $this->mNewRev->getTitle(); + if( $this->mNewRev->isCurrent() ) { + $newLink = $this->mNewPage->escapeLocalUrl( array( + 'oldid' => $this->mNewid + ) ); + $this->mPagetitle = htmlspecialchars( wfMsg( + 'currentrev-asof', + $timestamp, + $dateofrev, + $timeofrev + ) ); + $newEdit = $this->mNewPage->escapeLocalUrl( array( + 'action' => 'edit' + ) ); + + $this->mNewtitle = "{$this->mPagetitle}"; + $this->mNewtitle .= " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; + } else { + $newLink = $this->mNewPage->escapeLocalUrl( array( + 'oldid' => $this->mNewid + ) ); + $newEdit = $this->mNewPage->escapeLocalUrl( array( + 'action' => 'edit', + 'oldid' => $this->mNewid + ) ); + $this->mPagetitle = htmlspecialchars( wfMsg( + 'revisionasof', + $timestamp, + $dateofrev, + $timeofrev + ) ); + + $this->mNewtitle = "{$this->mPagetitle}"; + $this->mNewtitle .= " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; + } + if( !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { + $this->mNewtitle = "{$this->mPagetitle}"; + } else if ( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { + $this->mNewtitle = "{$this->mNewtitle}"; + } - # HTML-escape parameter before calling this - function addedLine( $line ) { - return $this->wrapLine( '+', 'diff-addedline', $line ); - } + // 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; */ - # HTML-escape parameter before calling this - function deletedLine( $line ) { - return $this->wrapLine( '−', 'diff-deletedline', $line ); - } + if( is_null( $this->mOldRev ) ) { + return false; + } - # HTML-escape parameter before calling this - function contextLine( $line ) { - return $this->wrapLine( ' ', 'diff-context', $line ); - } + if ( $this->mOldRev ) { + $this->mOldPage = $this->mOldRev->getTitle(); + + $t = $wgLang->timeanddate( $this->mOldRev->getTimestamp(), true ); + $dateofrev = $wgLang->date( $this->mOldRev->getTimestamp(), true ); + $timeofrev = $wgLang->time( $this->mOldRev->getTimestamp(), true ); + $oldLink = $this->mOldPage->escapeLocalUrl( array( + 'oldid' => $this->mOldid + ) ); + $oldEdit = $this->mOldPage->escapeLocalUrl( array( + 'action' => 'edit', + 'oldid' => $this->mOldid + ) ); + $this->mOldPagetitle = htmlspecialchars( wfMsg( 'revisionasof', $t, $dateofrev, $timeofrev ) ); + + $this->mOldtitle = "{$this->mOldPagetitle}" + . " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; + // Add an "undo" link + $newUndo = $this->mNewPage->escapeLocalUrl( array( + 'action' => 'edit', + 'undoafter' => $this->mOldid, + 'undo' => $this->mNewid + ) ); + $htmlLink = htmlspecialchars( wfMsg( 'editundo' ) ); + $htmlTitle = $wgUser->getSkin()->titleAttrib( 'undo' ); + if( $editable && !$this->mOldRev->isDeleted( Revision::DELETED_TEXT ) && !$this->mNewRev->isDeleted( Revision::DELETED_TEXT ) ) { + $this->mNewtitle .= " (" . $htmlLink . ")"; + } - private function wrapLine( $marker, $class, $line ) { - if( $line !== '' ) { - // The
wrapper is needed for 'overflow: auto' style to scroll properly - $line = Xml::tags( 'div', null, $this->escapeWhiteSpace( $line ) ); + if( !$this->mOldRev->userCan( Revision::DELETED_TEXT ) ) { + $this->mOldtitle = '' . $this->mOldPagetitle . ''; + } else if( $this->mOldRev->isDeleted( Revision::DELETED_TEXT ) ) { + $this->mOldtitle = '' . $this->mOldtitle . ''; + } } - return "$marker$line"; - } - function emptyLine() { - return ' '; + return true; } - function _added( $lines ) { - foreach ($lines as $line) { - echo '' . $this->emptyLine() . - $this->addedLine( '' . - htmlspecialchars ( $line ) . '' ) . "\n"; + /** + * 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 + $this->mTextLoaded = 2; } - } - function _deleted($lines) { - foreach ($lines as $line) { - echo '' . $this->deletedLine( '' . - htmlspecialchars ( $line ) . '' ) . - $this->emptyLine() . "\n"; + if ( !$this->loadRevisionData() ) { + return false; } - } - - function _context( $lines ) { - foreach ($lines as $line) { - echo '' . - $this->contextLine( htmlspecialchars ( $line ) ) . - $this->contextLine( htmlspecialchars ( $line ) ) . "\n"; + if ( $this->mOldRev ) { + $this->mOldtext = $this->mOldRev->getText( Revision::FOR_THIS_USER ); + if ( $this->mOldtext === false ) { + return false; + } + } + if ( $this->mNewRev ) { + $this->mNewtext = $this->mNewRev->getText( Revision::FOR_THIS_USER ); + if ( $this->mNewtext === false ) { + return false; + } } + return true; } - function _changed( $orig, $closing ) { - wfProfileIn( __METHOD__ ); - - $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 ); - echo '' . $this->deletedLine( $line ) . - $this->addedLine( $aline ) . "\n"; + /** + * Load the text of the new revision, not the old one + */ + function loadNewText() { + if ( $this->mTextLoaded >= 1 ) { + return true; + } else { + $this->mTextLoaded = 1; } - foreach ($add as $line) { # If any leftovers - echo '' . $this->emptyLine() . - $this->addedLine( $line ) . "\n"; + if ( !$this->loadRevisionData() ) { + return false; } - wfProfileOut( __METHOD__ ); + $this->mNewtext = $this->mNewRev->getText( Revision::FOR_THIS_USER ); + return true; } } diff --git a/includes/diff/DifferenceInterface.php b/includes/diff/DifferenceInterface.php deleted file mode 100644 index d85dcfcf44..0000000000 --- a/includes/diff/DifferenceInterface.php +++ /dev/null @@ -1,1072 +0,0 @@ -mTitle = $titleObj; - } else { - global $wgTitle; - $this->mTitle = $wgTitle; - } - wfDebug("DifferenceEngine old '$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 next one. - # Get next 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); - wfRunHooks( 'NewDifferenceEngine', array(&$titleObj, &$this->mOldid, &$this->mNewid, $old, $new) ); - } - $this->mRcidMarkPatrolled = intval($rcid); # force it to be an integer - $this->mRefreshCache = $refreshCache; - $this->unhide = $unhide; - } - - function setReducedLineNumbers( $value = true ) { - $this->mReducedLineNumbers = $value; - } - - function getTitle() { - return $this->mTitle; - } - - function wasCacheHit() { - return $this->mCacheHit; - } - - function getOldid() { - return $this->mOldid; - } - - function getNewid() { - return $this->mNewid; - } - - function showDiffPage( $diffOnly = false ) { - global $wgUser, $wgOut, $wgUseExternalEditor, $wgUseRCPatrol; - wfProfileIn( __METHOD__ ); - - - # If external diffs are enabled both globally and for the user, - # we'll use the application/x-external-editor interface to call - # an external diff tool like kompare, kdiff3, etc. - if($wgUseExternalEditor && $wgUser->getOption('externaldiff')) { - global $wgInputEncoding,$wgServer,$wgScript,$wgLang; - $wgOut->disable(); - header ( "Content-type: application/x-external-editor; charset=".$wgInputEncoding ); - $url1=$this->mTitle->getFullURL( array( - 'action' => 'raw', - 'oldid' => $this->mOldid - ) ); - $url2=$this->mTitle->getFullURL( array( - 'action' => 'raw', - 'oldid' => $this->mNewid - ) ); - $special=$wgLang->getNsText(NS_SPECIAL); - $control=<<setArticleFlag( false ); - if ( !$this->loadRevisionData() ) { - $t = $this->mTitle->getPrefixedText(); - $d = wfMsgExt( 'missingarticle-diff', array( 'escape' ), $this->mOldid, $this->mNewid ); - $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) ); - $wgOut->addWikiMsg( 'missing-article', "$t", $d ); - wfProfileOut( __METHOD__ ); - return; - } - - 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(); - $this->renderNewRevision(); // should we respect $diffOnly here or not? - wfProfileOut( __METHOD__ ); - return; - } - - $wgOut->suppressQuickbar(); - - $oldTitle = $this->mOldPage->getPrefixedText(); - $newTitle = $this->mNewPage->getPrefixedText(); - if( $oldTitle == $newTitle ) { - $wgOut->setPageTitle( $newTitle ); - } else { - $wgOut->setPageTitle( $oldTitle . ', ' . $newTitle ); - } - if ( $this->mNewPage->equals( $this->mOldPage ) ) { - $wgOut->setSubtitle( wfMsgExt( 'difference', array( 'parseinline' ) ) ); - } else { - $wgOut->setSubtitle( wfMsgExt( 'difference-multipage', array( 'parseinline' ) ) ); - } - $wgOut->setRobotPolicy( 'noindex,nofollow' ); - - if ( !$this->mOldPage->userCanRead() || !$this->mNewPage->userCanRead() ) { - $wgOut->loginToUse(); - $wgOut->output(); - $wgOut->disable(); - wfProfileOut( __METHOD__ ); - return; - } - - $sk = $wgUser->getSkin(); - - // Check if page is editable - $editable = $this->mNewRev->getTitle()->userCan( 'edit' ); - if ( $editable && $this->mNewRev->isCurrent() && $wgUser->isAllowed( 'rollback' ) ) { - $rollback = '   ' . $sk->generateRollback( $this->mNewRev ); - } else { - $rollback = ''; - } - - // Prepare a change patrol link, if applicable - if( $wgUseRCPatrol && $this->mTitle->userCan('patrol') ) { - // If we've been given an explicit change identifier, use it; saves time - if( $this->mRcidMarkPatrolled ) { - $rcid = $this->mRcidMarkPatrolled; - $rc = RecentChange::newFromId( $rcid ); - // Already patrolled? - $rcid = is_object($rc) && !$rc->getAttribute('rc_patrolled') ? $rcid : 0; - } else { - // Look for an unpatrolled change corresponding to this diff - $db = wfGetDB( DB_SLAVE ); - $change = RecentChange::newFromConds( - array( - // Redundant user,timestamp condition so we can use the existing index - 'rc_user_text' => $this->mNewRev->getRawUserText(), - 'rc_timestamp' => $db->timestamp( $this->mNewRev->getTimestamp() ), - 'rc_this_oldid' => $this->mNewid, - 'rc_last_oldid' => $this->mOldid, - 'rc_patrolled' => 0 - ), - __METHOD__ - ); - if( $change instanceof RecentChange ) { - $rcid = $change->mAttribs['rc_id']; - $this->mRcidMarkPatrolled = $rcid; - } else { - // None found - $rcid = 0; - } - } - // Build the link - if( $rcid ) { - $token = $wgUser->editToken( $rcid ); - $patrol = ' [' . $sk->link( - $this->mTitle, - wfMsgHtml( 'markaspatrolleddiff' ), - array(), - array( - 'action' => 'markpatrolled', - 'rcid' => $rcid, - 'token' => $token, - ), - array( - 'known', - 'noclasses' - ) - ) . ']'; - } else { - $patrol = ''; - } - } else { - $patrol = ''; - } - - # Carry over 'diffonly' param via navigation links - if( $diffOnly != $wgUser->getBoolOption('diffonly') ) { - $query['diffonly'] = $diffOnly; - } - - # Make "previous revision link" - $query['diff'] = 'prev'; - $query['oldid'] = $this->mOldid; - # Cascade unhide param in links for easy deletion browsing - if( $this->unhide ) { - $query['unhide'] = 1; - } - if( !$this->mOldRev->getPrevious() ) { - $prevlink = ' '; - } else { - $prevlink = $sk->link( - $this->mTitle, - wfMsgHtml( 'previousdiff' ), - array( - 'id' => 'differences-prevlink' - ), - $query, - array( - 'known', - 'noclasses' - ) - ); - } - - # Make "next revision link" - $query['diff'] = 'next'; - $query['oldid'] = $this->mNewid; - # Skip next link on the top revision - if( $this->mNewRev->isCurrent() ) { - $nextlink = ' '; - } else { - $nextlink = $sk->link( - $this->mTitle, - wfMsgHtml( 'nextdiff' ), - array( - 'id' => 'differences-nextlink' - ), - $query, - array( - 'known', - 'noclasses' - ) - ); - } - - $oldminor = ''; - $newminor = ''; - - if( $this->mOldRev->isMinor() ) { - $oldminor = ChangesList::flag( 'minor' ); - } - if( $this->mNewRev->isMinor() ) { - $newminor = ChangesList::flag( 'minor' ); - } - - # Handle RevisionDelete links... - $ldel = $this->revisionDeleteLink( $this->mOldRev ); - $rdel = $this->revisionDeleteLink( $this->mNewRev ); - - $oldHeader = '
'.$this->mOldtitle.'
' . - '
' . - $sk->revUserTools( $this->mOldRev, !$this->unhide ).'
' . - '
' . $oldminor . - $sk->revComment( $this->mOldRev, !$diffOnly, !$this->unhide ).$ldel.'
' . - '
' . $prevlink .'
'; - $newHeader = '
'.$this->mNewtitle.'
' . - '
' . $sk->revUserTools( $this->mNewRev, !$this->unhide ) . - " $rollback
" . - '
' . $newminor . - $sk->revComment( $this->mNewRev, !$diffOnly, !$this->unhide ).$rdel.'
' . - '
' . $nextlink . $patrol . '
'; - - # Check if this user can see the revisions - $allowed = $this->mOldRev->userCan(Revision::DELETED_TEXT) - && $this->mNewRev->userCan(Revision::DELETED_TEXT); - # Check if one of the revisions is deleted/suppressed - $deleted = $suppressed = false; - if( $this->mOldRev->isDeleted(Revision::DELETED_TEXT) ) { - $deleted = true; // old revisions text is hidden - if( $this->mOldRev->isDeleted(Revision::DELETED_RESTRICTED) ) - $suppressed = true; // also suppressed - } - if( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { - $deleted = true; // new revisions text is hidden - if( $this->mNewRev->isDeleted(Revision::DELETED_RESTRICTED) ) - $suppressed = true; // also suppressed - } - # If the diff cannot be shown due to a deleted revision, then output - # the diff header and links to unhide (if available)... - if( $deleted && (!$this->unhide || !$allowed) ) { - $this->showDiffStyle(); - $multi = $this->getMultiNotice(); - $wgOut->addHTML( $this->addHeader( '', $oldHeader, $newHeader, $multi ) ); - if( !$allowed ) { - $msg = $suppressed ? 'rev-suppressed-no-diff' : 'rev-deleted-no-diff'; - # Give explanation for why revision is not visible - $wgOut->wrapWikiMsg( "\n", - array( $msg ) ); - } else { - # Give explanation and add a link to view the diff... - $link = $this->mTitle->getFullUrl( array( - 'diff' => $this->mNewid, - 'oldid' => $this->mOldid, - 'unhide' => 1 - ) ); - $msg = $suppressed ? 'rev-suppressed-unhide-diff' : 'rev-deleted-unhide-diff'; - $wgOut->wrapWikiMsg( "\n", array( $msg, $link ) ); - } - # Otherwise, output a regular diff... - } else { - # Add deletion notice if the user is viewing deleted content - $notice = ''; - if( $deleted ) { - $msg = $suppressed ? 'rev-suppressed-diff-view' : 'rev-deleted-diff-view'; - $notice = "\n"; - } - $this->showDiff( $oldHeader, $newHeader, $notice ); - if( !$diffOnly ) { - $this->renderNewRevision(); - } - } - wfProfileOut( __METHOD__ ); - } - - protected function revisionDeleteLink( $rev ) { - global $wgUser; - $link = ''; - $canHide = $wgUser->isAllowed( 'deleterevision' ); - // Show del/undel link if: - // (a) the user can delete revisions, or - // (b) the user can view deleted revision *and* this one is deleted - if( $canHide || ($rev->getVisibility() && $wgUser->isAllowed( 'deletedhistory' )) ) { - $sk = $wgUser->getSkin(); - if( !$rev->userCan( Revision::DELETED_RESTRICTED ) ) { - $link = $sk->revDeleteLinkDisabled( $canHide ); // revision was hidden from sysops - } else { - $query = array( - 'type' => 'revision', - 'target' => $rev->mTitle->getPrefixedDbkey(), - 'ids' => $rev->getId() - ); - $link = $sk->revDeleteLink( $query, - $rev->isDeleted( Revision::DELETED_RESTRICTED ), $canHide ); - } - $link = '   ' . $link . ' '; - } - return $link; - } - - /** - * Show the new revision of the page. - */ - function renderNewRevision() { - global $wgOut, $wgUser; - wfProfileIn( __METHOD__ ); - - $wgOut->addHTML( "

{$this->mPagetitle}

\n" ); - # Add deleted rev tag if needed - if( !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { - $wgOut->wrapWikiMsg( "\n", 'rev-deleted-text-permission' ); - } else if( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { - $wgOut->wrapWikiMsg( "\n", 'rev-deleted-text-view' ); - } - - $pCache = true; - if( !$this->mNewRev->isCurrent() ) { - $oldEditSectionSetting = $wgOut->parserOptions()->setEditSection( false ); - $pCache = false; - } - - $this->loadNewText(); - if( is_object( $this->mNewRev ) ) { - $wgOut->setRevisionId( $this->mNewRev->getId() ); - } - - if( $this->mTitle->isCssJsSubpage() || $this->mTitle->isCssOrJsPage() ) { - // Stolen from Article::view --AG 2007-10-11 - // Give hooks a chance to customise the output - if( wfRunHooks( 'ShowRawCssJs', array( $this->mNewtext, $this->mTitle, $wgOut ) ) ) { - // Wrap the whole lot in a
 and don't parse
-				$m = array();
-				preg_match( '!\.(css|js)$!u', $this->mTitle->getText(), $m );
-				$wgOut->addHTML( "
\n" );
-				$wgOut->addHTML( htmlspecialchars( $this->mNewtext ) );
-				$wgOut->addHTML( "\n
\n" ); - } - } elseif( wfRunHooks( 'ArticleContentOnDiff', array( $this, $wgOut ) ) ) { - if ( $pCache ) { - $article = new Article( $this->mTitle, 0 ); - $pOutput = ParserCache::singleton()->get( $article, $wgOut->parserOptions() ); - if( $pOutput ) { - $wgOut->addParserOutput( $pOutput ); - } else { - $article->doViewParse(); - } - } else { - $wgOut->addWikiTextTidy( $this->mNewtext ); - } - } - - if( is_object( $this->mNewRev ) && !$this->mNewRev->isCurrent() ) { - $wgOut->parserOptions()->setEditSection( $oldEditSectionSetting ); - } - - # Add redundant patrol link on bottom... - if( $this->mRcidMarkPatrolled && $this->mTitle->quickUserCan('patrol') ) { - $sk = $wgUser->getSkin(); - $token = $wgUser->editToken( $this->mRcidMarkPatrolled ); - $wgOut->addHTML( - "' - ); - } - - wfProfileOut( __METHOD__ ); - } - - /** - * Show the first revision of an article. Uses normal diff headers in - * contrast to normal "old revision" display style. - */ - function showFirstRevision() { - global $wgOut, $wgUser; - wfProfileIn( __METHOD__ ); - - # Get article text from the DB - # - if ( ! $this->loadNewText() ) { - $t = $this->mTitle->getPrefixedText(); - $d = wfMsgExt( 'missingarticle-diff', array( 'escape' ), $this->mOldid, $this->mNewid ); - $wgOut->setPagetitle( wfMsg( 'errorpagetitle' ) ); - $wgOut->addWikiMsg( 'missing-article', "$t", $d ); - wfProfileOut( __METHOD__ ); - 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( __METHOD__ ); - throw new MWException("Permission Error: you do not have access to view this page"); - } - - # Prepare the header box - # - $sk = $wgUser->getSkin(); - - $next = $this->mTitle->getNextRevisionID( $this->mNewid ); - if( !$next ) { - $nextlink = ''; - } else { - $nextlink = '
' . $sk->link( - $this->mTitle, - wfMsgHtml( 'nextdiff' ), - array( - 'id' => 'differences-nextlink' - ), - array( - 'diff' => 'next', - 'oldid' => $this->mNewid, - ), - array( - 'known', - 'noclasses' - ) - ); - } - $header = "
" . - $sk->revUserTools( $this->mNewRev ) . "
" . $sk->revComment( $this->mNewRev ) . $nextlink . "
\n"; - - $wgOut->addHTML( $header ); - - $wgOut->setSubtitle( wfMsgExt( 'difference', array( 'parseinline' ) ) ); - $wgOut->setRobotPolicy( 'noindex,nofollow' ); - - wfProfileOut( __METHOD__ ); - } - - /** - * Get the diff text, send it to $wgOut - * Returns false if the diff could not be generated, otherwise returns true - */ - function showDiff( $otitle, $ntitle, $notice = '' ) { - global $wgOut; - $diff = $this->getDiff( $otitle, $ntitle, $notice ); - if ( $diff === false ) { - $wgOut->addWikiMsg( 'missing-article', "(fixme, bug)", '' ); - return false; - } else { - $this->showDiffStyle(); - $wgOut->addHTML( $diff ); - return true; - } - } - - /** - * Add style sheets and supporting JS for diff display. - */ - function showDiffStyle() { - global $wgOut; - $wgOut->addModules( 'mediawiki.legacy.diff' ); - } - - /** - * Get complete diff table, including header - * - * @param $otitle Title: old title - * @param $ntitle Title: new title - * @param $notice String: HTML between diff header and body - * @return mixed - */ - function getDiff( $otitle, $ntitle, $notice = '' ) { - $body = $this->getDiffBody(); - if ( $body === false ) { - return false; - } else { - $multi = $this->getMultiNotice(); - return $this->addHeader( $body, $otitle, $ntitle, $multi, $notice ); - } - } - - /** - * Get the diff table body, without header - * - * @return mixed - */ - function getDiffBody() { - global $wgMemc; - wfProfileIn( __METHOD__ ); - $this->mCacheHit = true; - // Check if the diff should be hidden from this user - if ( !$this->loadRevisionData() ) - return ''; - if ( $this->mOldRev && !$this->mOldRev->userCan(Revision::DELETED_TEXT) ) { - return ''; - } else if ( $this->mNewRev && !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { - return ''; - } else if ( $this->mOldRev && $this->mNewRev && $this->mOldRev->getID() == $this->mNewRev->getID() ) { - return ''; - } - // Cacheable? - $key = false; - if ( $this->mOldid && $this->mNewid ) { - $key = wfMemcKey( 'diff', 'version', MW_DIFF_VERSION, 'oldid', $this->mOldid, 'newid', $this->mNewid ); - // Try cache - if ( !$this->mRefreshCache ) { - $difftext = $wgMemc->get( $key ); - if ( $difftext ) { - wfIncrStats( 'diff_cache_hit' ); - $difftext = $this->localiseLineNumbers( $difftext ); - $difftext .= "\n\n"; - wfProfileOut( __METHOD__ ); - return $difftext; - } - } // don't try to load but save the result - } - $this->mCacheHit = false; - - // Loadtext is permission safe, this just clears out the diff - if ( !$this->loadText() ) { - wfProfileOut( __METHOD__ ); - return false; - } - - $difftext = $this->generateDiffBody( $this->mOldtext, $this->mNewtext ); - - // Save to cache for 7 days - if ( !wfRunHooks( 'AbortDiffCache', array( &$this ) ) ) { - wfIncrStats( 'diff_uncacheable' ); - } else 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( __METHOD__ ); - return $difftext; - } - - /** - * Make sure the proper modules are loaded before we try to - * make the diff - */ - private function initDiffEngines() { - global $wgExternalDiffEngine; - if ( $wgExternalDiffEngine == 'wikidiff' && !function_exists( 'wikidiff_do_diff' ) ) { - wfProfileIn( __METHOD__ . '-php_wikidiff.so' ); - wfSuppressWarnings(); - dl( 'php_wikidiff.so' ); - wfRestoreWarnings(); - wfProfileOut( __METHOD__ . '-php_wikidiff.so' ); - } - else if ( $wgExternalDiffEngine == 'wikidiff2' && !function_exists( 'wikidiff2_do_diff' ) ) { - wfProfileIn( __METHOD__ . '-php_wikidiff2.so' ); - wfSuppressWarnings(); - wfDl( 'wikidiff2' ); - wfRestoreWarnings(); - wfProfileOut( __METHOD__ . '-php_wikidiff2.so' ); - } - } - - /** - * Generate a diff, no caching - * - * @param $otext String: old text, must be already segmented - * @param $ntext String: new text, must be already segmented - */ - function generateDiffBody( $otext, $ntext ) { - global $wgExternalDiffEngine, $wgContLang; - - $otext = str_replace( "\r\n", "\n", $otext ); - $ntext = str_replace( "\r\n", "\n", $ntext ); - - $this->initDiffEngines(); - - if ( $wgExternalDiffEngine == 'wikidiff' && function_exists( 'wikidiff_do_diff' ) ) { - # For historical reasons, external diff engine expects - # input text to be HTML-escaped already - $otext = htmlspecialchars ( $wgContLang->segmentForDiff( $otext ) ); - $ntext = htmlspecialchars ( $wgContLang->segmentForDiff( $ntext ) ); - return $wgContLang->unsegementForDiff( wikidiff_do_diff( $otext, $ntext, 2 ) ) . - $this->debug( 'wikidiff1' ); - } - - if ( $wgExternalDiffEngine == 'wikidiff2' && function_exists( 'wikidiff2_do_diff' ) ) { - # Better external diff engine, the 2 may some day be dropped - # This one does the escaping and segmenting itself - wfProfileIn( 'wikidiff2_do_diff' ); - $text = wikidiff2_do_diff( $otext, $ntext, 2 ); - $text .= $this->debug( 'wikidiff2' ); - wfProfileOut( 'wikidiff2_do_diff' ); - return $text; - } - if ( $wgExternalDiffEngine != 'wikidiff3' && $wgExternalDiffEngine !== false ) { - # Diff via the shell - global $wgTmpDirectory; - $tempName1 = tempnam( $wgTmpDirectory, 'diff_' ); - $tempName2 = tempnam( $wgTmpDirectory, 'diff_' ); - - $tempFile1 = fopen( $tempName1, "w" ); - if ( !$tempFile1 ) { - wfProfileOut( __METHOD__ ); - return false; - } - $tempFile2 = fopen( $tempName2, "w" ); - if ( !$tempFile2 ) { - wfProfileOut( __METHOD__ ); - return false; - } - fwrite( $tempFile1, $otext ); - fwrite( $tempFile2, $ntext ); - fclose( $tempFile1 ); - fclose( $tempFile2 ); - $cmd = wfEscapeShellArg( $wgExternalDiffEngine, $tempName1, $tempName2 ); - wfProfileIn( __METHOD__ . "-shellexec" ); - $difftext = wfShellExec( $cmd ); - $difftext .= $this->debug( "external $wgExternalDiffEngine" ); - wfProfileOut( __METHOD__ . "-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 ) ) . - $this->debug(); - } - - /** - * Generate a debug comment indicating diff generating time, - * server node, and generator backend. - */ - protected function debug( $generator="internal" ) { - global $wgShowHostnames; - if ( !$this->enableDebugComment ) { - return ''; - } - $data = array( $generator ); - if( $wgShowHostnames ) { - $data[] = wfHostname(); - } - $data[] = wfTimestamp( TS_DB ); - return "\n"; - } - - /** - * Replace line numbers with the text in the user's language - */ - function localiseLineNumbers( $text ) { - return preg_replace_callback( '//', - array( &$this, 'localiseLineNumbersCb' ), $text ); - } - - function localiseLineNumbersCb( $matches ) { - global $wgLang; - if ( $matches[1] === '1' && $this->mReducedLineNumbers ) return ''; - return wfMsgExt( 'lineno', 'escape', $wgLang->formatNum( $matches[1] ) ); - } - - - /** - * If there are revisions between the ones being compared, return a note saying so. - */ - function getMultiNotice() { - if ( !is_object($this->mOldRev) || !is_object($this->mNewRev) ) - return ''; - - if( !$this->mOldPage->equals( $this->mNewPage ) ) { - // Comparing two different pages? Count would be meaningless. - return ''; - } - - $oldid = $this->mOldRev->getId(); - $newid = $this->mNewRev->getId(); - if ( $oldid > $newid ) { - $tmp = $oldid; $oldid = $newid; $newid = $tmp; - } - - $n = $this->mTitle->countRevisionsBetween( $oldid, $newid ); - if ( !$n ) { - return ''; - } else { - global $wgLang; - $dbr = wfGetDB( DB_SLAVE ); - - // Actually, the limit is $limit + 1. We do this so we can detect - // if there are > 100 authors in a given revision range. If they - // are, $limit will be passed to diff-multi-manyusers for l10n. - $limit = 100; - $res = $dbr->select( 'revision', 'DISTINCT rev_user_text', - array( - 'rev_page = ' . $this->mOldRev->getPage(), - 'rev_id > ' . $this->mOldRev->getId(), - 'rev_id < ' . $this->mNewRev->getId() - ), __METHOD__, - array( 'LIMIT' => $limit + 1 ) - ); - $numUsers = $dbr->numRows( $res ); - if( $numUsers > $limit ) { - $msg = 'diff-multi-manyusers'; - $numUsers = $limit; - } else { - $msg = 'diff-multi'; - } - return wfMsgExt( $msg, array( 'parseinline' ), $wgLang->formatnum( $n ), - $wgLang->formatnum( $numUsers ) ); - } - } - - - /** - * Add the header to a diff body - */ - static function addHeader( $diff, $otitle, $ntitle, $multi = '', $notice = '' ) { - $header = ""; - if( $diff ) { // Safari/Chrome show broken output if cols not used - $header .= " - - - - "; - $colspan = 2; - $multiColspan = 4; - } else { - $colspan = 1; - $multiColspan = 2; - } - $header .= " - - - - "; - - if ( $multi != '' ) { - $header .= ""; - } - if ( $notice != '' ) { - $header .= ""; - } - - return $header . $diff . "
{$otitle}{$ntitle}
{$multi}
{$notice}
"; - } - - /** - * Use specified text instead of loading from the database - */ - function setText( $oldText, $newText ) { - $this->mOldtext = $oldText; - $this->mNewtext = $newText; - $this->mTextLoaded = 2; - $this->mRevisionsLoaded = true; - } - - /** - * 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, $wgUser; - if ( $this->mRevisionsLoaded ) { - return true; - } else { - // Whether it succeeds or fails, we don't want to try again - $this->mRevisionsLoaded = true; - } - - // Load the new revision object - $this->mNewRev = $this->mNewid - ? Revision::newFromId( $this->mNewid ) - : Revision::newFromTitle( $this->mTitle ); - if( !$this->mNewRev instanceof Revision ) - return false; - - // Update the new revision ID in case it was 0 (makes life easier doing UI stuff) - $this->mNewid = $this->mNewRev->getId(); - - // Check if page is editable - $editable = $this->mNewRev->getTitle()->userCan( 'edit' ); - - // Set assorted variables - $timestamp = $wgLang->timeanddate( $this->mNewRev->getTimestamp(), true ); - $dateofrev = $wgLang->date( $this->mNewRev->getTimestamp(), true ); - $timeofrev = $wgLang->time( $this->mNewRev->getTimestamp(), true ); - $this->mNewPage = $this->mNewRev->getTitle(); - if( $this->mNewRev->isCurrent() ) { - $newLink = $this->mNewPage->escapeLocalUrl( array( - 'oldid' => $this->mNewid - ) ); - $this->mPagetitle = htmlspecialchars( wfMsg( - 'currentrev-asof', - $timestamp, - $dateofrev, - $timeofrev - ) ); - $newEdit = $this->mNewPage->escapeLocalUrl( array( - 'action' => 'edit' - ) ); - - $this->mNewtitle = "{$this->mPagetitle}"; - $this->mNewtitle .= " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; - } else { - $newLink = $this->mNewPage->escapeLocalUrl( array( - 'oldid' => $this->mNewid - ) ); - $newEdit = $this->mNewPage->escapeLocalUrl( array( - 'action' => 'edit', - 'oldid' => $this->mNewid - ) ); - $this->mPagetitle = htmlspecialchars( wfMsg( - 'revisionasof', - $timestamp, - $dateofrev, - $timeofrev - ) ); - - $this->mNewtitle = "{$this->mPagetitle}"; - $this->mNewtitle .= " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; - } - if( !$this->mNewRev->userCan(Revision::DELETED_TEXT) ) { - $this->mNewtitle = "{$this->mPagetitle}"; - } else if ( $this->mNewRev->isDeleted(Revision::DELETED_TEXT) ) { - $this->mNewtitle = "{$this->mNewtitle}"; - } - - // 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 ); - $dateofrev = $wgLang->date( $this->mOldRev->getTimestamp(), true ); - $timeofrev = $wgLang->time( $this->mOldRev->getTimestamp(), true ); - $oldLink = $this->mOldPage->escapeLocalUrl( array( - 'oldid' => $this->mOldid - ) ); - $oldEdit = $this->mOldPage->escapeLocalUrl( array( - 'action' => 'edit', - 'oldid' => $this->mOldid - ) ); - $this->mOldPagetitle = htmlspecialchars( wfMsg( 'revisionasof', $t, $dateofrev, $timeofrev ) ); - - $this->mOldtitle = "{$this->mOldPagetitle}" - . " (" . wfMsgHtml( $editable ? 'editold' : 'viewsourceold' ) . ")"; - // Add an "undo" link - $newUndo = $this->mNewPage->escapeLocalUrl( array( - 'action' => 'edit', - 'undoafter' => $this->mOldid, - 'undo' => $this->mNewid - ) ); - $htmlLink = htmlspecialchars( wfMsg( 'editundo' ) ); - $htmlTitle = $wgUser->getSkin()->titleAttrib( 'undo' ); - if( $editable && !$this->mOldRev->isDeleted( Revision::DELETED_TEXT ) && !$this->mNewRev->isDeleted( Revision::DELETED_TEXT ) ) { - $this->mNewtitle .= " (" . $htmlLink . ")"; - } - - if( !$this->mOldRev->userCan( Revision::DELETED_TEXT ) ) { - $this->mOldtitle = '' . $this->mOldPagetitle . ''; - } else if( $this->mOldRev->isDeleted( Revision::DELETED_TEXT ) ) { - $this->mOldtitle = '' . $this->mOldtitle . ''; - } - } - - 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 ) { - $this->mOldtext = $this->mOldRev->getText( Revision::FOR_THIS_USER ); - if ( $this->mOldtext === false ) { - return false; - } - } - if ( $this->mNewRev ) { - $this->mNewtext = $this->mNewRev->getText( Revision::FOR_THIS_USER ); - 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( Revision::FOR_THIS_USER ); - return true; - } -} diff --git a/includes/diff/WikiDiff.php b/includes/diff/WikiDiff.php new file mode 100644 index 0000000000..c148d23853 --- /dev/null +++ b/includes/diff/WikiDiff.php @@ -0,0 +1,1241 @@ + + * You may copy this code freely under the conditions of the GPL. + * + * @file + * @ingroup DifferenceEngine + * @defgroup DifferenceEngine DifferenceEngine + */ + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _DiffOp { + var $type; + var $orig; + var $closing; + + function reverse() { + trigger_error('pure virtual', E_USER_ERROR); + } + + function norig() { + return $this->orig ? sizeof($this->orig) : 0; + } + + function nclosing() { + return $this->closing ? sizeof($this->closing) : 0; + } +} + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _DiffOp_Copy extends _DiffOp { + var $type = 'copy'; + + function __construct ($orig, $closing = false) { + if (!is_array($closing)) + $closing = $orig; + $this->orig = $orig; + $this->closing = $closing; + } + + function reverse() { + return new _DiffOp_Copy($this->closing, $this->orig); + } +} + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _DiffOp_Delete extends _DiffOp { + var $type = 'delete'; + + function __construct ($lines) { + $this->orig = $lines; + $this->closing = false; + } + + function reverse() { + return new _DiffOp_Add($this->orig); + } +} + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _DiffOp_Add extends _DiffOp { + var $type = 'add'; + + function __construct ($lines) { + $this->closing = $lines; + $this->orig = false; + } + + function reverse() { + return new _DiffOp_Delete($this->closing); + } +} + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _DiffOp_Change extends _DiffOp { + var $type = 'change'; + + function __construct ($orig, $closing) { + $this->orig = $orig; + $this->closing = $closing; + } + + function reverse() { + return new _DiffOp_Change($this->closing, $this->orig); + } +} + +/** + * Class used internally by Diff to actually compute the diffs. + * + * The algorithm used here is mostly lifted from the perl module + * Algorithm::Diff (version 1.06) by Ned Konz, which is available at: + * http://www.perl.com/CPAN/authors/id/N/NE/NEDKONZ/Algorithm-Diff-1.06.zip + * + * More ideas are taken from: + * http://www.ics.uci.edu/~eppstein/161/960229.html + * + * Some ideas are (and a bit of code) are from from analyze.c, from GNU + * diffutils-2.7, which can be found at: + * ftp://gnudist.gnu.org/pub/gnu/diffutils/diffutils-2.7.tar.gz + * + * closingly, some ideas (subdivision by NCHUNKS > 2, and some optimizations) + * are my own. + * + * Line length limits for robustness added by Tim Starling, 2005-08-31 + * Alternative implementation added by Guy Van den Broeck, 2008-07-30 + * + * @author Geoffrey T. Dairiki, Tim Starling, Guy Van den Broeck + * @private + * @ingroup DifferenceEngine + */ +class _DiffEngine { + + const MAX_XREF_LENGTH = 10000; + + function diff ($from_lines, $to_lines){ + wfProfileIn( __METHOD__ ); + + // Diff and store locally + $this->diff_local($from_lines, $to_lines); + + // Merge edits when possible + $this->_shift_boundaries($from_lines, $this->xchanged, $this->ychanged); + $this->_shift_boundaries($to_lines, $this->ychanged, $this->xchanged); + + // Compute the edit operations. + $n_from = sizeof($from_lines); + $n_to = sizeof($to_lines); + + $edits = array(); + $xi = $yi = 0; + while ($xi < $n_from || $yi < $n_to) { + assert($yi < $n_to || $this->xchanged[$xi]); + assert($xi < $n_from || $this->ychanged[$yi]); + + // Skip matching "snake". + $copy = array(); + while ( $xi < $n_from && $yi < $n_to + && !$this->xchanged[$xi] && !$this->ychanged[$yi]) { + $copy[] = $from_lines[$xi++]; + ++$yi; + } + if ($copy) + $edits[] = new _DiffOp_Copy($copy); + + // Find deletes & adds. + $delete = array(); + while ($xi < $n_from && $this->xchanged[$xi]) + $delete[] = $from_lines[$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) + $edits[] = new _DiffOp_Delete($delete); + elseif ($add) + $edits[] = new _DiffOp_Add($add); + } + wfProfileOut( __METHOD__ ); + return $edits; + } + + function diff_local ($from_lines, $to_lines) { + global $wgExternalDiffEngine; + wfProfileIn( __METHOD__); + + if($wgExternalDiffEngine == 'wikidiff3'){ + // wikidiff3 + $wikidiff3 = new WikiDiff3(); + $wikidiff3->diff($from_lines, $to_lines); + $this->xchanged = $wikidiff3->removed; + $this->ychanged = $wikidiff3->added; + unset($wikidiff3); + }else{ + // old diff + $n_from = sizeof($from_lines); + $n_to = sizeof($to_lines); + $this->xchanged = $this->ychanged = array(); + $this->xv = $this->yv = array(); + $this->xind = $this->yind = array(); + unset($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]) + 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]) + 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++) { + $xhash[$this->_line_hash($from_lines[$xi])] = 1; + } + + for ($yi = $skip; $yi < $n_to - $endskip; $yi++) { + $line = $to_lines[$yi]; + if ( ($this->ychanged[$yi] = empty($xhash[$this->_line_hash($line)])) ) + continue; + $yhash[$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($yhash[$this->_line_hash($line)])) ) + continue; + $this->xv[] = $line; + $this->xind[] = $xi; + } + + // Find the LCS. + $this->_compareseq(0, sizeof($this->xv), 0, sizeof($this->yv)); + } + wfProfileOut( __METHOD__ ); + } + + /** + * Returns the whole line if it's small enough, or the MD5 hash otherwise + */ + function _line_hash( $line ) { + if ( strlen( $line ) > self::MAX_XREF_LENGTH ) { + return md5( $line ); + } else { + return $line; + } + } + + /* Divide the Largest Common Subsequence (LCS) of the sequences + * [XOFF, XLIM) and [YOFF, YLIM) into NCHUNKS approximately equally + * sized segments. + * + * Returns (LCS, PTS). LCS is the length of the LCS. PTS is an + * array of NCHUNKS+1 (X, Y) indexes giving the diving points between + * sub sequences. The first sub-sequence is contained in [X0, X1), + * [Y0, Y1), the second in [X1, X2), [Y1, Y2) and so on. Note + * that (X0, Y0) == (XOFF, YOFF) and + * (X[NCHUNKS], Y[NCHUNKS]) == (XLIM, YLIM). + * + * This function assumes that the first lines of the specified portions + * of the two files do not match, and likewise that the last lines do not + * match. The caller must trim matching lines from the beginning and end + * of the portions it is going to specify. + */ + function _diag ($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++) { + $line = $flip ? $this->yv[$x] : $this->xv[$x]; + 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 ); + assert( $k > 0 ); + $ymids[$k] = $ymids[$k-1]; + break; + } + } + while (list ( /* $junk */, $y) = each($matches)) { + if ($y > $this->seq[$k-1]) { + 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); + assert($k > 0); + $ymids[$k] = $ymids[$k-1]; + } + } + } + } + + $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, $ylim); + + return array($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; + } + + $beg = 1; + while ($beg < $end) { + $mid = (int)(($beg + $end) / 2); + if ( $ypos > $this->seq[$mid] ) + $beg = $mid + 1; + else + $end = $mid; + } + + assert($ypos != $this->seq[$end]); + + $this->in_seq[$this->seq[$end]] = false; + $this->seq[$end] = $ypos; + $this->in_seq[$ypos] = 1; + return $end; + } + + /* Find LCS of two sequences. + * + * The results are recorded in the vectors $this->{x,y}changed[], by + * storing a 1 in the element for each line that is an insertion + * or deletion (ie. is not in the LCS). + * + * The subsequence of file 0 is [XOFF, XLIM) and likewise for file 1. + * + * Note that XLIM, YLIM are exclusive bounds. + * All line numbers are origin-0 and discarded lines are not counted. + */ + function _compareseq ($xoff, $xlim, $yoff, $ylim) { + // Slide down the bottom initial diagonal. + while ($xoff < $xlim && $yoff < $ylim + && $this->xv[$xoff] == $this->yv[$yoff]) { + ++$xoff; + ++$yoff; + } + + // Slide up the top initial diagonal. + while ($xlim > $xoff && $ylim > $yoff + && $this->xv[$xlim - 1] == $this->yv[$ylim - 1]) { + --$xlim; + --$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 = $pt2; + } + } + } + + /* Adjust inserts/deletes of identical lines to join changes + * as much as possible. + * + * We do something when a run of changed lines include a + * line at one end and has an excluded, identical line at the other. + * We are free to choose which identical line is included. + * `compareseq' usually chooses the one at the beginning, + * but usually it is cleaner to consider the following identical line + * to be the "change". + * + * This is extracted verbatim from analyze.c (GNU diffutils-2.7). + */ + function _shift_boundaries ($lines, &$changed, $other_changed) { + wfProfileIn( __METHOD__ ); + $i = 0; + $j = 0; + + assert('sizeof($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]) { + 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]) + $start--; + assert('$j > 0'); + while ($other_changed[--$j]) + continue; + 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]) + $i++; + + assert('$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; + assert('$j > 0'); + while ($other_changed[--$j]) + continue; + assert('$j >= 0 && !$other_changed[$j]'); + } + } + wfProfileOut( __METHOD__ ); + } +} + +/** + * Class representing a 'diff' between two sequences of strings. + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class Diff +{ + var $edits; + + /** + * Constructor. + * Computes diff between sequences of strings. + * + * @param $from_lines array An array of strings. + * (Typically these are lines from a file.) + * @param $to_lines array An array of strings. + */ + function __construct($from_lines, $to_lines) { + $eng = new _DiffEngine; + $this->edits = $eng->diff($from_lines, $to_lines); + //$this->_check($from_lines, $to_lines); + } + + /** + * Compute reversed Diff. + * + * SYNOPSIS: + * + * $diff = new Diff($lines1, $lines2); + * $rev = $diff->reverse(); + * @return object A Diff object representing the inverse of the + * original diff. + */ + function reverse () { + $rev = $this; + $rev->edits = array(); + foreach ($this->edits as $edit) { + $rev->edits[] = $edit->reverse(); + } + return $rev; + } + + /** + * Check for empty diff. + * + * @return bool True iff two sequences were identical. + */ + function isEmpty () { + foreach ($this->edits as $edit) { + if ($edit->type != 'copy') + return false; + } + return true; + } + + /** + * Compute the length of the Longest Common Subsequence (LCS). + * + * This is mostly for diagnostic purposed. + * + * @return int The length of the LCS. + */ + function lcs () { + $lcs = 0; + foreach ($this->edits as $edit) { + if ($edit->type == 'copy') + $lcs += sizeof($edit->orig); + } + return $lcs; + } + + /** + * Get the original set of lines. + * + * This reconstructs the $from_lines parameter passed to the + * constructor. + * + * @return array The original sequence of strings. + */ + function orig() { + $lines = array(); + + foreach ($this->edits as $edit) { + if ($edit->orig) + array_splice($lines, sizeof($lines), 0, $edit->orig); + } + return $lines; + } + + /** + * Get the closing set of lines. + * + * This reconstructs the $to_lines parameter passed to the + * constructor. + * + * @return array The sequence of strings. + */ + function closing() { + $lines = array(); + + foreach ($this->edits as $edit) { + if ($edit->closing) + array_splice($lines, sizeof($lines), 0, $edit->closing); + } + return $lines; + } + + /** + * Check a Diff for validity. + * + * This is here only for debugging purposes. + */ + function _check ($from_lines, $to_lines) { + wfProfileIn( __METHOD__ ); + if (serialize($from_lines) != serialize($this->orig())) + trigger_error("Reconstructed original doesn't match", E_USER_ERROR); + if (serialize($to_lines) != serialize($this->closing())) + trigger_error("Reconstructed closing doesn't match", E_USER_ERROR); + + $rev = $this->reverse(); + if (serialize($to_lines) != serialize($rev->orig())) + trigger_error("Reversed original doesn't match", E_USER_ERROR); + if (serialize($from_lines) != serialize($rev->closing())) + trigger_error("Reversed closing doesn't match", E_USER_ERROR); + + + $prevtype = 'none'; + foreach ($this->edits as $edit) { + if ( $prevtype == $edit->type ) + trigger_error("Edit sequence is non-optimal", E_USER_ERROR); + $prevtype = $edit->type; + } + + $lcs = $this->lcs(); + trigger_error('Diff okay: LCS = '.$lcs, E_USER_NOTICE); + wfProfileOut( __METHOD__ ); + } +} + +/** + * @todo document, bad name. + * @private + * @ingroup DifferenceEngine + */ +class MappedDiff extends Diff +{ + /** + * Constructor. + * + * Computes diff between sequences of strings. + * + * This can be used to compute things like + * case-insensitve diffs, or diffs which ignore + * changes in white-space. + * + * @param $from_lines array An array of strings. + * (Typically these are lines from a file.) + * + * @param $to_lines array An array of strings. + * + * @param $mapped_from_lines array This array should + * have the same size number of elements as $from_lines. + * The elements in $mapped_from_lines and + * $mapped_to_lines are what is actually compared + * when computing the diff. + * + * @param $mapped_to_lines array This array should + * have the same number of elements as $to_lines. + */ + function __construct($from_lines, $to_lines, + $mapped_from_lines, $mapped_to_lines) { + wfProfileIn( __METHOD__ ); + + assert(sizeof($from_lines) == sizeof($mapped_from_lines)); + assert(sizeof($to_lines) == sizeof($mapped_to_lines)); + + parent::__construct($mapped_from_lines, $mapped_to_lines); + + $xi = $yi = 0; + for ($i = 0; $i < sizeof($this->edits); $i++) { + $orig = &$this->edits[$i]->orig; + if (is_array($orig)) { + $orig = array_slice($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( __METHOD__ ); + } +} + +/** + * A class to format Diffs + * + * 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 + * @ingroup DifferenceEngine + */ +class DiffFormatter { + /** + * Number of leading context "lines" to preserve. + * + * This should be left at zero for this class, but subclasses + * may want to set this to other values. + */ + var $leading_context_lines = 0; + + /** + * Number of trailing context "lines" to preserve. + * + * This should be left at zero for this class, but subclasses + * may want to set this to other values. + */ + var $trailing_context_lines = 0; + + /** + * Format a diff. + * + * @param $diff object A Diff object. + * @return string The formatted output. + */ + function format($diff) { + wfProfileIn( __METHOD__ ); + + $xi = $yi = 1; + $block = false; + $context = array(); + + $nlead = $this->leading_context_lines; + $ntrail = $this->trailing_context_lines; + + $this->_start_diff(); + + foreach ($diff->edits as $edit) { + if ($edit->type == 'copy') { + if (is_array($block)) { + if (sizeof($edit->orig) <= $nlead + $ntrail) { + $block[] = $edit; + } + else{ + if ($ntrail) { + $context = array_slice($edit->orig, 0, $ntrail); + $block[] = new _DiffOp_Copy($context); + } + $this->_block($x0, $ntrail + $xi - $x0, + $y0, $ntrail + $yi - $y0, + $block); + $block = false; + } + } + $context = $edit->orig; + } + else { + if (! is_array($block)) { + $context = array_slice($context, sizeof($context) - $nlead); + $x0 = $xi - sizeof($context); + $y0 = $yi - sizeof($context); + $block = array(); + if ($context) + $block[] = new _DiffOp_Copy($context); + } + $block[] = $edit; + } + + if ($edit->orig) + $xi += sizeof($edit->orig); + if ($edit->closing) + $yi += sizeof($edit->closing); + } + + if (is_array($block)) + $this->_block($x0, $xi - $x0, + $y0, $yi - $y0, + $block); + + $end = $this->_end_diff(); + wfProfileOut( __METHOD__ ); + return $end; + } + + function _block($xbeg, $xlen, $ybeg, $ylen, &$edits) { + wfProfileIn( __METHOD__ ); + $this->_start_block($this->_block_header($xbeg, $xlen, $ybeg, $ylen)); + foreach ($edits as $edit) { + if ($edit->type == 'copy') + $this->_context($edit->orig); + elseif ($edit->type == 'add') + $this->_added($edit->closing); + elseif ($edit->type == 'delete') + $this->_deleted($edit->orig); + elseif ($edit->type == 'change') + $this->_changed($edit->orig, $edit->closing); + else + trigger_error('Unknown edit type', E_USER_ERROR); + } + $this->_end_block(); + wfProfileOut( __METHOD__ ); + } + + function _start_diff() { + ob_start(); + } + + function _end_diff() { + $val = ob_get_contents(); + ob_end_clean(); + return $val; + } + + function _block_header($xbeg, $xlen, $ybeg, $ylen) { + if ($xlen > 1) + $xbeg .= "," . ($xbeg + $xlen - 1); + if ($ylen > 1) + $ybeg .= "," . ($ybeg + $ylen - 1); + + return $xbeg . ($xlen ? ($ylen ? 'c' : 'd') : 'a') . $ybeg; + } + + function _start_block($header) { + echo $header . "\n"; + } + + function _end_block() { + } + + function _lines($lines, $prefix = ' ') { + foreach ($lines as $line) + echo "$prefix $line\n"; + } + + function _context($lines) { + $this->_lines($lines); + } + + function _added($lines) { + $this->_lines($lines, '>'); + } + function _deleted($lines) { + $this->_lines($lines, '<'); + } + + function _changed($orig, $closing) { + $this->_deleted($orig); + echo "---\n"; + $this->_added($closing); + } +} + +/** + * A formatter that outputs unified diffs + * @ingroup DifferenceEngine + */ + +class UnifiedDiffFormatter extends DiffFormatter { + var $leading_context_lines = 2; + var $trailing_context_lines = 2; + + function _added($lines) { + $this->_lines($lines, '+'); + } + function _deleted($lines) { + $this->_lines($lines, '-'); + } + function _changed($orig, $closing) { + $this->_deleted($orig); + $this->_added($closing); + } + function _block_header($xbeg, $xlen, $ybeg, $ylen) { + return "@@ -$xbeg,$xlen +$ybeg,$ylen @@"; + } +} + +/** + * A pseudo-formatter that just passes along the Diff::$edits array + * @ingroup DifferenceEngine + */ +class ArrayDiffFormatter extends DiffFormatter { + function format($diff) { + $oldline = 1; + $newline = 1; + $retval = array(); + foreach($diff->edits as $edit) + switch($edit->type) { + case 'add': + foreach($edit->closing as $l) { + $retval[] = array( + 'action' => 'add', + 'new'=> $l, + 'newline' => $newline++ + ); + } + break; + case 'delete': + foreach($edit->orig as $l) { + $retval[] = array( + 'action' => 'delete', + 'old' => $l, + 'oldline' => $oldline++, + ); + } + break; + case 'change': + foreach($edit->orig as $i => $l) { + $retval[] = array( + 'action' => 'change', + 'old' => $l, + 'new' => @$edit->closing[$i], + 'oldline' => $oldline++, + 'newline' => $newline++, + ); + } + break; + case 'copy': + $oldline += count($edit->orig); + $newline += count($edit->orig); + } + return $retval; + } +} + +/** + * Additions by Axel Boldt follow, partly taken from diff.php, phpwiki-1.3.3 + * + */ + +define('NBSP', ' '); // iso-8859-x non-breaking space. + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class _HWLDF_WordAccumulator { + function __construct () { + $this->_lines = array(); + $this->_line = ''; + $this->_group = ''; + $this->_tag = ''; + } + + function _flushGroup ($new_tag) { + if ($this->_group !== '') { + if ($this->_tag == 'ins') + $this->_line .= '' . + htmlspecialchars ( $this->_group ) . ''; + elseif ($this->_tag == 'del') + $this->_line .= '' . + htmlspecialchars ( $this->_group ) . ''; + else + $this->_line .= htmlspecialchars ( $this->_group ); + } + $this->_group = ''; + $this->_tag = $new_tag; + } + + function _flushLine ($new_tag) { + $this->_flushGroup($new_tag); + if ($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); + + foreach ($words as $word) { + // new-line should only come as first char of word. + if ($word == '') + continue; + if ($word[0] == "\n") { + $this->_flushLine($tag); + $word = substr($word, 1); + } + assert(!strstr($word, "\n")); + $this->_group .= $word; + } + } + + function getLines() { + $this->_flushLine('~done'); + return preg_replace( '/\n/m', '', $this->_lines); + } +} + +/** + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class WordLevelDiff extends MappedDiff { + const MAX_LINE_LENGTH = 10000; + + function __construct ($orig_lines, $closing_lines) { + wfProfileIn( __METHOD__ ); + + list ($orig_words, $orig_stripped) = $this->_split($orig_lines); + list ($closing_words, $closing_stripped) = $this->_split($closing_lines); + + parent::__construct($orig_words, $closing_words, + $orig_stripped, $closing_stripped); + wfProfileOut( __METHOD__ ); + } + + function _split($lines) { + wfProfileIn( __METHOD__ ); + + $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 ) > self::MAX_LINE_LENGTH ) { + $words[] = $line; + $stripped[] = $line; + } else { + $m = array(); + 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, $m[1] ); + } + } + } + wfProfileOut( __METHOD__ ); + return array($words, $stripped); + } + + function orig () { + wfProfileIn( __METHOD__ ); + $orig = new _HWLDF_WordAccumulator; + + foreach ($this->edits as $edit) { + if ($edit->type == 'copy') + $orig->addWords($edit->orig); + elseif ($edit->orig) + $orig->addWords($edit->orig, 'del'); + } + $lines = $orig->getLines(); + wfProfileOut( __METHOD__ ); + return $lines; + } + + function closing () { + wfProfileIn( __METHOD__ ); + $closing = new _HWLDF_WordAccumulator; + + foreach ($this->edits as $edit) { + if ($edit->type == 'copy') + $closing->addWords($edit->closing); + elseif ($edit->closing) + $closing->addWords($edit->closing, 'ins'); + } + $lines = $closing->getLines(); + wfProfileOut( __METHOD__ ); + return $lines; + } +} + +/** + * Wikipedia Table style diff formatter. + * @todo document + * @private + * @ingroup DifferenceEngine + */ +class TableDiffFormatter extends DiffFormatter { + function __construct() { + $this->leading_context_lines = 2; + $this->trailing_context_lines = 2; + } + + public static function escapeWhiteSpace( $msg ) { + $msg = preg_replace( '/^ /m', '  ', $msg ); + $msg = preg_replace( '/ $/m', '  ', $msg ); + $msg = preg_replace( '/ /', '  ', $msg ); + return $msg; + } + + function _block_header( $xbeg, $xlen, $ybeg, $ylen ) { + $r = '\n" . + '\n"; + return $r; + } + + function _start_block( $header ) { + echo $header; + } + + function _end_block() { + } + + function _lines( $lines, $prefix=' ', $color='white' ) { + } + + # HTML-escape parameter before calling this + function addedLine( $line ) { + return $this->wrapLine( '+', 'diff-addedline', $line ); + } + + # HTML-escape parameter before calling this + function deletedLine( $line ) { + return $this->wrapLine( '−', 'diff-deletedline', $line ); + } + + # HTML-escape parameter before calling this + function contextLine( $line ) { + return $this->wrapLine( ' ', 'diff-context', $line ); + } + + private function wrapLine( $marker, $class, $line ) { + if( $line !== '' ) { + // The
wrapper is needed for 'overflow: auto' style to scroll properly + $line = Xml::tags( 'div', null, $this->escapeWhiteSpace( $line ) ); + } + return "$marker$line"; + } + + function emptyLine() { + return ' '; + } + + function _added( $lines ) { + foreach ($lines as $line) { + echo '' . $this->emptyLine() . + $this->addedLine( '' . + htmlspecialchars ( $line ) . '' ) . "\n"; + } + } + + function _deleted($lines) { + foreach ($lines as $line) { + echo '' . $this->deletedLine( '' . + htmlspecialchars ( $line ) . '' ) . + $this->emptyLine() . "\n"; + } + } + + function _context( $lines ) { + foreach ($lines as $line) { + echo '' . + $this->contextLine( htmlspecialchars ( $line ) ) . + $this->contextLine( htmlspecialchars ( $line ) ) . "\n"; + } + } + + function _changed( $orig, $closing ) { + wfProfileIn( __METHOD__ ); + + $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 ); + echo '' . $this->deletedLine( $line ) . + $this->addedLine( $aline ) . "\n"; + } + foreach ($add as $line) { # If any leftovers + echo '' . $this->emptyLine() . + $this->addedLine( $line ) . "\n"; + } + wfProfileOut( __METHOD__ ); + } +} diff --git a/includes/diff/Diff.php b/includes/diff/WikiDiff3.php similarity index 100% rename from includes/diff/Diff.php rename to includes/diff/WikiDiff3.php -- 2.20.1