/**
* Build a set of regular expressions from a MediaWiki message.
* Will be correctly empty if the message isn't present.
*
* @param $message string
* @return array of regular expressions, potentially empty
*/
static function regexesFromMessage( $message ) {
$source = wfMsgForContent( $message );
if( $source && !wfEmptyMsg( $message, $source ) ) {
return SpamRegexBatch::regexesFromText( $source );
} else {
return array();
}
}
function buildSharedBlacklists() {
$regexes = array();
$listType = $this->getBlacklistType();
# Load lists
wfDebugLog( 'SpamBlacklist', "Constructing $listType blacklist\n" );
foreach ( $this->files as $fileName ) {
$matches = array();
if ( preg_match( '/^DB: ([\w-]*) (.*)$/', $fileName, $matches ) ) {
$text = $this->getArticleText( $matches[1], $matches[2] );
} elseif ( preg_match( '/^http:\/\//', $fileName ) ) {
$text = $this->getHttpText( $fileName );
} else {
$text = file_get_contents( $fileName );
wfDebugLog( 'SpamBlacklist', "got from file $fileName\n" );
}
// Build a separate batch of regexes from each source.
// While in theory we could squeeze a little efficiency
// out of combining multiple sources in one regex, if
// there's a bad line in one of them we'll gain more
// from only having to break that set into smaller pieces.
$regexes = array_merge( $regexes,
SpamRegexBatch::regexesFromText( $text, $fileName ) );
}
return $regexes;
}
/**
* Build a set of regular expressions from a MediaWiki message.
* Will be correctly empty if the message isn't present.
*
* @param $message string
* @param $blacklist BaseBlacklist
* @return array of regular expressions, potentially empty
*/
static function regexesFromMessage($message, BaseBlacklist $blacklist)
{
$source = wfMessage($message)->inContentLanguage();
if (!$source->isDisabled()) {
return SpamRegexBatch::regexesFromText($source->plain(), $blacklist);
} else {
return array();
}
}