# with this program; if not, write to the Free Software Foundation, Inc.,
# 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
# http://www.gnu.org/copyleft/gpl.html
+/**
+ *
+ * @package MediaWiki
+ * @subpackage SpecialPage
+ */
-function wfSpecialExport( $page = "" ) {
+/** */
+require_once( 'Revision.php' );
+
+/**
+ *
+ */
+function wfSpecialExport( $page = '' ) {
global $wgOut, $wgLang, $wgRequest;
if( $wgRequest->getVal( 'action' ) == 'submit') {
$curonly = true;
}
- if( $page != "" ) {
+ if( $page != '' ) {
+ $wgOut->disable();
header( "Content-type: application/xml; charset=utf-8" );
$pages = explode( "\n", $page );
$xml = pages2xml( $pages, $curonly );
echo $xml;
- wfAbruptExit();
+ return;
}
$wgOut->addWikiText( wfMsg( "exporttext" ) );
<form method='post' action=\"$action\">
<input type='hidden' name='action' value='submit' />
<textarea name='pages' cols='40' rows='10'></textarea><br />
-<label><input type='checkbox' name='curonly' value='true' checked />
+<label><input type='checkbox' name='curonly' value='true' checked='checked' />
" . wfMsg( "exportcuronly" ) . "</label><br />
<input type='submit' />
</form>
}
function pages2xml( $pages, $curonly = false ) {
- global $wgLanguageCode, $wgInputEncoding, $wgLang;
+ $fname = 'pages2xml';
+ wfProfileIn( $fname );
+
+ global $wgContLanguageCode, $wgInputEncoding, $wgContLang;
$xml = "<" . "?xml version=\"1.0\" encoding=\"UTF-8\" ?" . ">\n" .
- "<mediawiki version=\"0.1\" xml:lang=\"$wgLanguageCode\">\n";
+ "<mediawiki version=\"0.1\" xml:lang=\"$wgContLanguageCode\">\n";
foreach( $pages as $page ) {
$xml .= page2xml( $page, $curonly );
}
$xml .= "</mediawiki>\n";
if($wgInputEncoding != "utf-8")
- $xml = $wgLang->iconv( $wgInputEncoding, "utf-8", $xml );
+ $xml = $wgContLang->iconv( $wgInputEncoding, "utf-8", $xml );
+
+ wfProfileOut( $fname );
return $xml;
}
function page2xml( $page, $curonly, $full = false ) {
- global $wgInputCharset, $wgLang;
+ global $wgLang;
+ $fname = 'page2xml';
+ wfProfileIn( $fname );
+
$title = Title::NewFromText( $page );
- if( !$title ) return "";
- $t = wfStrencode( $title->getDBKey() );
- $ns = $title->getNamespace();
- $sql = "SELECT cur_id as id,cur_timestamp as timestamp,cur_user as user,cur_user_text as user_text," .
- "cur_restrictions as restrictions,cur_comment as comment,cur_text as text FROM cur " .
- "WHERE cur_namespace=$ns AND cur_title='$t'";
- $res = wfQuery( $sql, DB_READ );
- if( $s = wfFetchObject( $res ) ) {
- $tl = htmlspecialchars( $title->getPrefixedText() );
+ if( !$title ) {
+ wfProfileOut( $fname );
+ return "";
+ }
+
+ $dbr =& wfGetDB( DB_SLAVE );
+ $s = $dbr->selectRow( 'page',
+ array( 'page_id', 'page_restrictions' ),
+ array( 'page_namespace' => $title->getNamespace(),
+ 'page_title' => $title->getDbkey() ) );
+ if( $s ) {
+ $tl = xmlsafe( $title->getPrefixedText() );
$xml = " <page>\n";
$xml .= " <title>$tl</title>\n";
+
if( $full ) {
- $xml .= " <id>$s->id</id>\n";
+ $xml .= " <id>$s->page_id</id>\n";
}
- if( $s->restrictions ) {
- $xml .= " <restrictions>$s->restrictions</restrictions>\n";
+ if( $s->page_restrictions ) {
+ $xml .= " <restrictions>" . xmlsafe( $s->page_restrictions ) . "</restrictions>\n";
}
- if( !$curonly ) {
- $sql = "SELECT old_id as id,old_timestamp as timestamp, old_user as user, old_user_text as user_text," .
- "old_comment as comment, old_text as text, old_flags as flags FROM old " .
- "WHERE old_namespace=$ns AND old_title='$t' ORDER BY old_timestamp";
- $res = wfQuery( $sql, DB_READ );
- while( $s2 = wfFetchObject( $res ) ) {
- $xml .= revision2xml( $s2, $full, false );
+ if( $curonly ) {
+ $res = Revision::fetchRevision( $title );
+ } else {
+ $res = Revision::fetchAllRevisions( $title );
+ }
+ if( $res ) {
+ while( $s = $res->fetchObject() ) {
+ $rev = new Revision( $s );
+ $xml .= revision2xml( $rev, $full, false );
}
+ $res->free();
}
- $xml .= revision2xml( $s, $full, true );
+
$xml .= " </page>\n";
+ wfProfileOut( $fname );
return $xml;
} else {
+ wfProfileOut( $fname );
return "";
}
}
-function revision2xml( $s, $full, $cur ) {
- $ts = wfTimestamp2ISO8601( $s->timestamp );
+/**
+ * @return string
+ * @param Revision $rev
+ * @param bool $full
+ * @access private
+ */
+function revision2xml( $rev, $full ) {
+ $fname = 'revision2xml';
+ wfProfileIn( $fname );
+
+ global $wgParser , $wgUseXMLparser ; # So we can return *real* XML
+
$xml = " <revision>\n";
- if($full && !$cur)
- $xml .= " <id>$s->id</id>\n";
+ if( $full )
+ $xml .= " <id>" . $rev->getId() . "</id>\n";
+
+ $ts = wfTimestamp2ISO8601( $rev->getTimestamp() );
$xml .= " <timestamp>$ts</timestamp>\n";
- if($s->user) {
- $u = "<username>" . htmlspecialchars( $s->user_text ) . "</username>";
- if($full)
- $u .= "<id>$s->user</id>";
+
+ if( $rev->getUser() ) {
+ $u = "<username>" . xmlsafe( $rev->getUserText() ) . "</username>";
+ if( $full )
+ $u .= "<id>" . $rev->getUser() . "</id>";
} else {
- $u = "<ip>" . htmlspecialchars( $s->user_text ) . "</ip>";
+ $u = "<ip>" . xmlsafe( $rev->getUserText() ) . "</ip>";
}
$xml .= " <contributor>$u</contributor>\n";
- if( !empty( $s->minor ) ) {
+
+ if( $rev->isMinor() ) {
$xml .= " <minor/>\n";
}
- if($s->comment != "") {
- $c = htmlspecialchars( $s->comment );
+ if($rev->getComment() != "") {
+ $c = xmlsafe( $rev->getComment() );
$xml .= " <comment>$c</comment>\n";
}
- $t = htmlspecialchars( Article::getRevisionText( $s, "" ) );
+
+ $t = $rev->getText() ;
+ if ( $wgUseXMLparser )
+ {
+ $wgParser->runXMLparser ( $t ) ;
+ }
+ $t = xmlsafe( $t ) ;
+
$xml .= " <text>$t</text>\n";
$xml .= " </revision>\n";
+ wfProfileOut( $fname );
return $xml;
}
return preg_replace( '/^(....)(..)(..)(..)(..)(..)$/', '$1-$2-$3T$4:$5:$6Z', $ts );
}
-?>
+function xmlsafe( $string ) {
+ $fname = 'xmlsafe';
+ wfProfileIn( $fname );
+
+ /**
+ * The page may contain old data which has not been properly normalized.
+ * Invalid UTF-8 sequences or forbidden control characters will make our
+ * XML output invalid, so be sure to strip them out.
+ */
+ global $wgUseLatin1;
+ if( $wgUseLatin1 ) {
+ /**
+ * We know the UTF-8 is valid since we converted outselves.
+ * Just check for forbidden controls...
+ */
+ $string = preg_replace( '/[\x00-\x08\x0b-\x1f]/', '', $string );
+ } else {
+ $string = UtfNormal::cleanUp( $string );
+ }
+
+ $string = htmlspecialchars( $string );
+ wfProfileOut( $fname );
+ return $string;
+}
+?>
\ No newline at end of file