/**
* Secure and split - main initialisation function for this object
*
* Assumes that mDbkeyform has been set, and is urldecoded
* and uses underscores, but not otherwise munged. This function
* removes illegal characters, splits off the interwiki and
* namespace prefixes, sets the other forms, and canonicalizes
* everything.
* @return bool true on success
*/
private function secureAndSplit()
{
global $wgContLang, $wgLocalInterwiki, $wgCapitalLinks;
# Initialisation
static $rxTc = false;
if (!$rxTc) {
# % is needed as well
$rxTc = '/[^' . Title::legalChars() . ']|%[0-9A-Fa-f]{2}/S';
}
$this->mInterwiki = $this->mFragment = '';
$this->mNamespace = $this->mDefaultNamespace;
# Usually NS_MAIN
$dbkey = $this->mDbkeyform;
# Strip Unicode bidi override characters.
# Sometimes they slip into cut-n-pasted page titles, where the
# override chars get included in list displays.
$dbkey = str_replace("", '', $dbkey);
// 200E LEFT-TO-RIGHT MARK
$dbkey = str_replace("", '', $dbkey);
// 200F RIGHT-TO-LEFT MARK
# Clean up whitespace
#
$dbkey = preg_replace('/[ _]+/', '_', $dbkey);
$dbkey = trim($dbkey, '_');
if ('' == $dbkey) {
return false;
}
if (false !== strpos($dbkey, UTF8_REPLACEMENT)) {
# Contained illegal UTF-8 sequences or forbidden Unicode chars.
return false;
}
$this->mDbkeyform = $dbkey;
# Initial colon indicates main namespace rather than specified default
# but should not create invalid {ns,title} pairs such as {0,Project:Foo}
if (':' == $dbkey[0]) {
$this->mNamespace = NS_MAIN;
$dbkey = substr($dbkey, 1);
# remove the colon but continue processing
$dbkey = trim($dbkey, '_');
# remove any subsequent whitespace
}
# Namespace or interwiki prefix
$firstPass = true;
do {
$m = array();
if (preg_match("/^(.+?)_*:_*(.*)\$/S", $dbkey, $m)) {
$p = $m[1];
if ($ns = $wgContLang->getNsIndex($p)) {
# Ordinary namespace
$dbkey = $m[2];
$this->mNamespace = $ns;
} elseif ($this->getInterwikiLink($p)) {
if (!$firstPass) {
# Can't make a local interwiki link to an interwiki link.
# That's just crazy!
return false;
}
# Interwiki link
$dbkey = $m[2];
$this->mInterwiki = $wgContLang->lc($p);
# Redundant interwiki prefix to the local wiki
if (0 == strcasecmp($this->mInterwiki, $wgLocalInterwiki)) {
if ($dbkey == '') {
# Can't have an empty self-link
return false;
}
$this->mInterwiki = '';
$firstPass = false;
# Do another namespace split...
continue;
}
# If there's an initial colon after the interwiki, that also
# resets the default namespace
if ($dbkey !== '' && $dbkey[0] == ':') {
$this->mNamespace = NS_MAIN;
$dbkey = substr($dbkey, 1);
}
}
# If there's no recognized interwiki or namespace,
# then let the colon expression be part of the title.
}
break;
} while (true);
# We already know that some pages won't be in the database!
#
if ('' != $this->mInterwiki || NS_SPECIAL == $this->mNamespace) {
$this->mArticleID = 0;
}
$fragment = strstr($dbkey, '#');
if (false !== $fragment) {
$this->setFragment($fragment);
$dbkey = substr($dbkey, 0, strlen($dbkey) - strlen($fragment));
# remove whitespace again: prevents "Foo_bar_#"
# becoming "Foo_bar_"
$dbkey = preg_replace('/_*$/', '', $dbkey);
}
# Reject illegal characters.
#
if (preg_match($rxTc, $dbkey)) {
return false;
}
/**
* Pages with "/./" or "/../" appearing in the URLs will
* often be unreachable due to the way web browsers deal
* with 'relative' URLs. Forbid them explicitly.
*/
if (strpos($dbkey, '.') !== false && ($dbkey === '.' || $dbkey === '..' || strpos($dbkey, './') === 0 || strpos($dbkey, '../') === 0 || strpos($dbkey, '/./') !== false || strpos($dbkey, '/../') !== false)) {
return false;
}
/**
* Magic tilde sequences? Nu-uh!
*/
if (strpos($dbkey, '~~~') !== false) {
return false;
}
/**
* Limit the size of titles to 255 bytes.
* This is typically the size of the underlying database field.
* We make an exception for special pages, which don't need to be stored
* in the database, and may edge over 255 bytes due to subpage syntax
* for long titles, e.g. [[Special:Block/Long name]]
*/
if ($this->mNamespace != NS_SPECIAL && strlen($dbkey) > 255 || strlen($dbkey) > 512) {
return false;
}
/**
* Normally, all wiki links are forced to have
* an initial capital letter so [[foo]] and [[Foo]]
* point to the same place.
*
* Don't force it for interwikis, since the other
* site might be case-sensitive.
*/
if ($wgCapitalLinks && $this->mInterwiki == '') {
$dbkey = $wgContLang->ucfirst($dbkey);
}
/**
* Can't make a link to a namespace alone...
* "empty" local links can only be self-links
* with a fragment identifier.
*/
if ($dbkey == '' && $this->mInterwiki == '' && $this->mNamespace != NS_MAIN) {
return false;
}
// Any remaining initial :s are illegal.
if ($dbkey !== '' && ':' == $dbkey[0]) {
return false;
}
# Fill fields
$this->mDbkeyform = $dbkey;
$this->mUrlform = ilWikiUtil::wfUrlencode($dbkey);
$this->mTextform = str_replace('_', ' ', $dbkey);
return true;
}
/**
* Set page parameter for Url Embedding
*/
static function makeUrlTitle($a_par)
{
$a_par = ilWikiUtil::removeUnsafeCharacters($a_par);
$a_par = str_replace(" ", "_", $a_par);
return ilWikiUtil::wfUrlencode($a_par);
}
