* Internet Explorer data URI length limit. See encodeImageAsDataURI().
*/
const DATA_URI_SIZE_LIMIT = 32768;
- const URL_REGEX = 'url\(\s*[\'"]?(?P<file>[^\?\)\'"]*?)(?P<query>\?[^\)\'"]*?|)[\'"]?\s*\)';
+
const EMBED_REGEX = '\/\*\s*\@embed\s*\*\/';
const COMMENT_REGEX = '\/\*.*?\*\/';
$files = [];
$rFlags = PREG_OFFSET_CAPTURE | PREG_SET_ORDER;
- if ( preg_match_all( '/' . self::URL_REGEX . '/', $stripped, $matches, $rFlags ) ) {
+ if ( preg_match_all( '/' . self::getUrlRegex() . '/', $stripped, $matches, $rFlags ) ) {
foreach ( $matches as $match ) {
+ self::processUrlMatch( $match, $rFlags );
$url = $match['file'][0];
// Skip fully-qualified and protocol-relative URLs and data URIs
// quotation marks (e.g. "foo /* bar").
$comments = [];
- $pattern = '/(?!' . CSSMin::EMBED_REGEX . ')(' . CSSMin::COMMENT_REGEX . ')/s';
+ $pattern = '/(?!' . self::EMBED_REGEX . ')(' . self::COMMENT_REGEX . ')/s';
$source = preg_replace_callback(
$pattern,
// appears in the rule itself, e.g. in a quoted string. You are advised
// not to use such characters in file names. We also match start/end of
// the string to be consistent in edge-cases ('@import url(…)').
- $pattern = '/(?:^|[;{])\K[^;{}]*' . CSSMin::URL_REGEX . '[^;}]*(?=[;}]|$)/';
+ $pattern = '/(?:^|[;{])\K[^;{}]*' . self::getUrlRegex() . '[^;}]*(?=[;}]|$)/';
$source = preg_replace_callback(
$pattern,
// Build two versions of current rule: with remapped URLs
// and with embedded data: URIs (where possible).
- $pattern = '/(?P<embed>' . CSSMin::EMBED_REGEX . '\s*|)' . CSSMin::URL_REGEX . '/';
+ $pattern = '/(?P<embed>' . CSSMin::EMBED_REGEX . '\s*|)' . self::getUrlRegex() . '/';
$ruleWithRemapped = preg_replace_callback(
$pattern,
function ( $match ) use ( $local, $remote ) {
- $remapped = CSSMin::remapOne( $match['file'], $match['query'], $local, $remote, false );
+ self::processUrlMatch( $match );
+ $remapped = CSSMin::remapOne( $match['file'], $match['query'], $local, $remote, false );
return CSSMin::buildUrlValue( $remapped );
},
$rule
$ruleWithEmbedded = preg_replace_callback(
$pattern,
function ( $match ) use ( $embedAll, $local, $remote, &$mimeTypes ) {
+ self::processUrlMatch( $match );
+
$embed = $embedAll || $match['embed'];
$embedded = CSSMin::remapOne(
$match['file'],
}, $source );
// Re-insert comments
- $pattern = '/' . CSSMin::PLACEHOLDER . '(\d+)x/';
- $source = preg_replace_callback( $pattern, function( $match ) use ( &$comments ) {
+ $pattern = '/' . self::PLACEHOLDER . '(\d+)x/';
+ $source = preg_replace_callback( $pattern, function ( $match ) use ( &$comments ) {
return $comments[ $match[1] ];
}, $source );
return false;
}
+ private static function getUrlRegex() {
+ static $urlRegex;
+ if ( $urlRegex === null ) {
+ // Match these three variants separately to avoid broken urls when
+ // e.g. a double quoted url contains a parenthesis, or when a
+ // single quoted url contains a double quote, etc.
+ // Note: PCRE doesn't support multiple capture groups with the same name by default.
+ // - PCRE 6.7 introduced the "J" modifier (PCRE_INFO_JCHANGED for PCRE_DUPNAMES).
+ // https://secure.php.net/manual/en/reference.pcre.pattern.modifiers.php
+ // However this isn't useful since it just ignores all but the first one.
+ // Also, while the modifier was introduced in PCRE 6.7 (PHP 5.2+) it was
+ // not exposed to public preg_* functions until PHP 5.6.0.
+ // - PCRE 8.36 fixed this to work as expected (e.g. merge conceptually to
+ // only return the one matched in the part that actually matched).
+ // However MediaWiki supports 5.5.9, which has PCRE 8.32
+ // Per https://secure.php.net/manual/en/pcre.installation.php:
+ // - PCRE 8.32 (PHP 5.5.0)
+ // - PCRE 8.34 (PHP 5.5.10, PHP 5.6.0)
+ // - PCRE 8.37 (PHP 5.5.26, PHP 5.6.9, PHP 7.0.0)
+ // Workaround by using different groups and merge via processUrlMatch().
+ // - Using string concatenation for class constant or member assignments
+ // is only supported in PHP 5.6. Use a getter method for now.
+ $urlRegex = '(' .
+ // Unquoted url
+ 'url\(\s*(?P<file0>[^\'"][^\?\)]*?)(?P<query0>\?[^\)]*?|)\s*\)' .
+ // Single quoted url
+ '|url\(\s*\'(?P<file1>[^\?\']*?)(?P<query1>\?[^\']*?|)\'\s*\)' .
+ // Double quoted url
+ '|url\(\s*"(?P<file2>[^\?"]*?)(?P<query2>\?[^"]*?|)"\s*\)' .
+ ')';
+ }
+ return $urlRegex;
+ }
+
+ private static function processUrlMatch( array &$match, $flags = 0 ) {
+ if ( $flags & PREG_SET_ORDER ) {
+ // preg_match_all with PREG_SET_ORDER will return each group in each
+ // match array, and if it didn't match, instead of the sub array
+ // being an empty array it is `[ '', -1 ]`...
+ if ( isset( $match['file0'] ) && $match['file0'][1] !== -1 ) {
+ $match['file'] = $match['file0'];
+ $match['query'] = $match['query0'];
+ } elseif ( isset( $match['file1'] ) && $match['file1'][1] !== -1 ) {
+ $match['file'] = $match['file1'];
+ $match['query'] = $match['query1'];
+ } else {
+ $match['file'] = $match['file2'];
+ $match['query'] = $match['query2'];
+ }
+ } else {
+ if ( isset( $match['file0'] ) && $match['file0'] !== '' ) {
+ $match['file'] = $match['file0'];
+ $match['query'] = $match['query0'];
+ } elseif ( isset( $match['file1'] ) && $match['file1'] !== '' ) {
+ $match['file'] = $match['file1'];
+ $match['query'] = $match['query1'];
+ } else {
+ $match['file'] = $match['file2'];
+ $match['query'] = $match['query2'];
+ }
+ }
+ }
+
/**
* Remap or embed a CSS URL path.
*
dépôts / lhc / web / wiklou.git / blobdiff