<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head><meta http-equiv="content-type" content="text/html; charset=utf-8" />
<title>[32364] trunk: WPDB: When checking that a string can be sent to MySQL, we shouldn't use `mb_convert_encoding()`, as it behaves differently to MySQL's character encoding conversion.</title>
</head>
<body>
<style type="text/css"><!--
#msg dl.meta { border: 1px #006 solid; background: #369; padding: 6px; color: #fff; }
#msg dl.meta dt { float: left; width: 6em; font-weight: bold; }
#msg dt:after { content:':';}
#msg dl, #msg dt, #msg ul, #msg li, #header, #footer, #logmsg { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt; }
#msg dl a { font-weight: bold}
#msg dl a:link { color:#fc3; }
#msg dl a:active { color:#ff0; }
#msg dl a:visited { color:#cc6; }
h3 { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt; font-weight: bold; }
#msg pre { overflow: auto; background: #ffc; border: 1px #fa0 solid; padding: 6px; }
#logmsg { background: #ffc; border: 1px #fa0 solid; padding: 1em 1em 0 1em; }
#logmsg p, #logmsg pre, #logmsg blockquote { margin: 0 0 1em 0; }
#logmsg p, #logmsg li, #logmsg dt, #logmsg dd { line-height: 14pt; }
#logmsg h1, #logmsg h2, #logmsg h3, #logmsg h4, #logmsg h5, #logmsg h6 { margin: .5em 0; }
#logmsg h1:first-child, #logmsg h2:first-child, #logmsg h3:first-child, #logmsg h4:first-child, #logmsg h5:first-child, #logmsg h6:first-child { margin-top: 0; }
#logmsg ul, #logmsg ol { padding: 0; list-style-position: inside; margin: 0 0 0 1em; }
#logmsg ul { text-indent: -1em; padding-left: 1em; }#logmsg ol { text-indent: -1.5em; padding-left: 1.5em; }
#logmsg > ul, #logmsg > ol { margin: 0 0 1em 0; }
#logmsg pre { background: #eee; padding: 1em; }
#logmsg blockquote { border: 1px solid #fa0; border-left-width: 10px; padding: 1em 1em 0 1em; background: white;}
#logmsg dl { margin: 0; }
#logmsg dt { font-weight: bold; }
#logmsg dd { margin: 0; padding: 0 0 0.5em 0; }
#logmsg dd:before { content:'\00bb';}
#logmsg table { border-spacing: 0px; border-collapse: collapse; border-top: 4px solid #fa0; border-bottom: 1px solid #fa0; background: #fff; }
#logmsg table th { text-align: left; font-weight: normal; padding: 0.2em 0.5em; border-top: 1px dotted #fa0; }
#logmsg table td { text-align: right; border-top: 1px dotted #fa0; padding: 0.2em 0.5em; }
#logmsg table thead th { text-align: center; border-bottom: 1px solid #fa0; }
#logmsg table th.Corner { text-align: left; }
#logmsg hr { border: none 0; border-top: 2px dashed #fa0; height: 1px; }
#header, #footer { color: #fff; background: #636; border: 1px #300 solid; padding: 6px; }
#patch { width: 100%; }
#patch h4 {font-family: verdana,arial,helvetica,sans-serif;font-size:10pt;padding:8px;background:#369;color:#fff;margin:0;}
#patch .propset h4, #patch .binary h4 {margin:0;}
#patch pre {padding:0;line-height:1.2em;margin:0;}
#patch .diff {width:100%;background:#eee;padding: 0 0 10px 0;overflow:auto;}
#patch .propset .diff, #patch .binary .diff {padding:10px 0;}
#patch span {display:block;padding:0 10px;}
#patch .modfile, #patch .addfile, #patch .delfile, #patch .propset, #patch .binary, #patch .copfile {border:1px solid #ccc;margin:10px 0;}
#patch ins {background:#dfd;text-decoration:none;display:block;padding:0 10px;}
#patch del {background:#fdd;text-decoration:none;display:block;padding:0 10px;}
#patch .lines, .info {color:#888;background:#fff;}
--></style>
<div id="msg">
<dl class="meta" style="font-size: 105%">
<dt style="float: left; width: 6em; font-weight: bold">Revision</dt> <dd><a style="font-weight: bold" href="https://core.trac.wordpress.org/changeset/32364">32364</a><script type="application/ld+json">{"@context":"http://schema.org","@type":"EmailMessage","description":"Review this Commit","action":{"@type":"ViewAction","url":"https://core.trac.wordpress.org/changeset/32364","name":"Review Commit"}}</script></dd>
<dt style="float: left; width: 6em; font-weight: bold">Author</dt> <dd>pento</dd>
<dt style="float: left; width: 6em; font-weight: bold">Date</dt> <dd>2015-05-06 02:59:50 +0000 (Wed, 06 May 2015)</dd>
</dl>
<pre style='padding-left: 1em; margin: 2em 0; border-left: 2px solid #ccc; line-height: 1.25; font-size: 105%; font-family: sans-serif'>WPDB: When checking that a string can be sent to MySQL, we shouldn't use `mb_convert_encoding()`, as it behaves differently to MySQL's character encoding conversion.
Props mdawaffe, pento, nbachiyski, jorbin, johnjamesjacoby, jeremyfelt.
See <a href="https://core.trac.wordpress.org/ticket/32165">#32165</a>.</pre>
<h3>Modified Paths</h3>
<ul>
<li><a href="#trunksrcwpadminincludesupgradephp">trunk/src/wp-admin/includes/upgrade.php</a></li>
<li><a href="#trunksrcwpincludescommentphp">trunk/src/wp-includes/comment.php</a></li>
<li><a href="#trunksrcwpincludescompatphp">trunk/src/wp-includes/compat.php</a></li>
<li><a href="#trunksrcwpincludesversionphp">trunk/src/wp-includes/version.php</a></li>
<li><a href="#trunksrcwpincludeswpdbphp">trunk/src/wp-includes/wp-db.php</a></li>
<li><a href="#trunktestsphpunittestscommentphp">trunk/tests/phpunit/tests/comment.php</a></li>
<li><a href="#trunktestsphpunittestscompatphp">trunk/tests/phpunit/tests/compat.php</a></li>
<li><a href="#trunktestsphpunittestsdbcharsetphp">trunk/tests/phpunit/tests/db/charset.php</a></li>
<li><a href="#trunktestsphpunittestsdbphp">trunk/tests/phpunit/tests/db.php</a></li>
</ul>
</div>
<div id="patch">
<h3>Diff</h3>
<a id="trunksrcwpadminincludesupgradephp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/src/wp-admin/includes/upgrade.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/src/wp-admin/includes/upgrade.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/src/wp-admin/includes/upgrade.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -527,7 +527,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> if ( $wp_current_db_version < 31351 )
</span><span class="cx" style="display: block; padding: 0 10px"> upgrade_420();
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( $wp_current_db_version < 32308 )
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( $wp_current_db_version < 32364 )
</ins><span class="cx" style="display: block; padding: 0 10px"> upgrade_430();
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> maybe_disable_link_manager();
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -1446,17 +1446,33 @@
</span><span class="cx" style="display: block; padding: 0 10px"> function upgrade_430() {
</span><span class="cx" style="display: block; padding: 0 10px"> global $wp_current_db_version, $wpdb;
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( $wp_current_db_version < 32308 ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( $wp_current_db_version < 32364 ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> $content_length = $wpdb->get_col_length( $wpdb->comments, 'comment_content' );
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( ! $content_length ) {
- $content_length = 65535;
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( false === $content_length ) {
+ $content_length = array(
+ 'type' => 'byte',
+ 'length' => 65535,
+ );
+ } elseif ( ! is_array( $content_length ) ) {
+ $length = (int) $content_length > 0 ? (int) $content_length : 65535;
+ $content_length = array(
+ 'type' => 'byte',
+ 'length' => $length
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( 'byte' !== $content_length['type'] ) {
+ // Sites with malformed DB schemas are on their own.
+ return;
+ }
+
+ $allowed_length = intval( $content_length['length'] ) - 10;
+
</ins><span class="cx" style="display: block; padding: 0 10px"> $comments = $wpdb->get_results(
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- "SELECT comment_ID FROM $wpdb->comments
- WHERE comment_date_gmt > '2015-04-26'
- AND CHAR_LENGTH( comment_content ) >= $content_length
- AND ( comment_content LIKE '%<%' OR comment_content LIKE '%>%' )"
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ "SELECT `comment_ID` FROM `{$wpdb->comments}`
+ WHERE `comment_date_gmt` > '2015-04-26'
+ AND LENGTH( `comment_content` ) >= {$allowed_length}
+ AND ( `comment_content` LIKE '%<%' OR `comment_content` LIKE '%>%' )"
</ins><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> foreach ( $comments as $comment ) {
</span></span></pre></div>
<a id="trunksrcwpincludescommentphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/src/wp-includes/comment.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/src/wp-includes/comment.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/src/wp-includes/comment.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2118,17 +2118,7 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $compacted = compact( 'comment_post_ID', 'comment_author', 'comment_author_email', 'comment_author_url', 'comment_author_IP', 'comment_date', 'comment_date_gmt', 'comment_content', 'comment_karma', 'comment_approved', 'comment_agent', 'comment_type', 'comment_parent', 'user_id' );
</span><span class="cx" style="display: block; padding: 0 10px"> if ( ! $wpdb->insert( $wpdb->comments, $compacted ) ) {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $fields = array( 'comment_author', 'comment_author_email', 'comment_author_url', 'comment_content' );
-
- foreach( $fields as $field ) {
- if ( isset( $compacted[ $field ] ) ) {
- $compacted[ $field ] = $wpdb->strip_invalid_text_for_column( $wpdb->comments, $field, $compacted[ $field ] );
- }
- }
-
- if ( ! $wpdb->insert( $wpdb->comments, $compacted ) ) {
- return false;
- }
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return false;
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $id = (int) $wpdb->insert_id;
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2252,6 +2242,8 @@
</span><span class="cx" style="display: block; padding: 0 10px"> * @return int|bool The ID of the comment on success, false on failure.
</span><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> function wp_new_comment( $commentdata ) {
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ global $wpdb;
+
</ins><span class="cx" style="display: block; padding: 0 10px"> if ( isset( $commentdata['user_ID'] ) ) {
</span><span class="cx" style="display: block; padding: 0 10px"> $commentdata['user_id'] = $commentdata['user_ID'] = (int) $commentdata['user_ID'];
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2295,7 +2287,22 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $comment_ID = wp_insert_comment($commentdata);
</span><span class="cx" style="display: block; padding: 0 10px"> if ( ! $comment_ID ) {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return false;
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $fields = array( 'comment_author', 'comment_author_email', 'comment_author_url', 'comment_content' );
+
+ foreach( $fields as $field ) {
+ if ( isset( $commentdata[ $field ] ) ) {
+ $commentdata[ $field ] = $wpdb->strip_invalid_text_for_column( $wpdb->comments, $field, $commentdata[ $field ] );
+ }
+ }
+
+ $commentdata = wp_filter_comment( $commentdata );
+
+ $commentdata['comment_approved'] = wp_allow_comment( $commentdata );
+
+ $comment_ID = wp_insert_comment( $commentdata );
+ if ( ! $comment_ID ) {
+ return false;
+ }
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> /**
</span></span></pre></div>
<a id="trunksrcwpincludescompatphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/src/wp-includes/compat.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/src/wp-includes/compat.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/src/wp-includes/compat.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -13,23 +13,85 @@
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+/**
+ * Returns whether PCRE/u (PCRE_UTF8 modifier) is available for use.
+ *
+ * @ignore
+ * @since 4.2.2
+ * @access private
+ *
+ * @param bool $set - Used for testing only
+ * null : default - get PCRE/u capability
+ * false : Used for testing - return false for future calls to this function
+ * 'reset': Used for testing - restore default behavior of this function
+ */
+function _wp_can_use_pcre_u( $set = null ) {
+ static $utf8_pcre = 'reset';
+
+ if ( null !== $set ) {
+ $utf8_pcre = $set;
+ }
+
+ if ( 'reset' === $utf8_pcre ) {
+ $utf8_pcre = @preg_match( '/^./u', 'a' );
+ }
+
+ return $utf8_pcre;
+}
+
</ins><span class="cx" style="display: block; padding: 0 10px"> if ( ! function_exists( 'mb_substr' ) ) :
</span><span class="cx" style="display: block; padding: 0 10px"> function mb_substr( $str, $start, $length = null, $encoding = null ) {
</span><span class="cx" style="display: block; padding: 0 10px"> return _mb_substr( $str, $start, $length, $encoding );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> endif;
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+/*
+ * Only understands UTF-8 and 8bit. All other character sets will be treated as 8bit.
+ * For $encoding === UTF-8, the $str input is expected to be a valid UTF-8 byte sequence.
+ * The behavior of this function for invalid inputs is undefined.
+ */
</ins><span class="cx" style="display: block; padding: 0 10px"> function _mb_substr( $str, $start, $length = null, $encoding = null ) {
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( null === $encoding ) {
+ $encoding = get_option( 'blog_charset' );
+ }
+
</ins><span class="cx" style="display: block; padding: 0 10px"> // The solution below works only for UTF-8,
</span><span class="cx" style="display: block; padding: 0 10px"> // so in case of a different charset just use built-in substr()
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $charset = get_option( 'blog_charset' );
- if ( ! in_array( $charset, array( 'utf8', 'utf-8', 'UTF8', 'UTF-8' ) ) ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( ! in_array( $encoding, array( 'utf8', 'utf-8', 'UTF8', 'UTF-8' ) ) ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> return is_null( $length ) ? substr( $str, $start ) : substr( $str, $start, $length );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Use the regex unicode support to separate the UTF-8 characters into an array
- preg_match_all( '/./us', $str, $match );
- $chars = is_null( $length ) ? array_slice( $match[0], $start ) : array_slice( $match[0], $start, $length );
- return implode( '', $chars );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+ if ( _wp_can_use_pcre_u() ) {
+ // Use the regex unicode support to separate the UTF-8 characters into an array
+ preg_match_all( '/./us', $str, $match );
+ $chars = is_null( $length ) ? array_slice( $match[0], $start ) : array_slice( $match[0], $start, $length );
+ return implode( '', $chars );
+ }
+
+ $regex = '/(
+ [\x00-\x7F] # single-byte sequences 0xxxxxxx
+ | [\xC2-\xDF][\x80-\xBF] # double-byte sequences 110xxxxx 10xxxxxx
+ | \xE0[\xA0-\xBF][\x80-\xBF] # triple-byte sequences 1110xxxx 10xxxxxx * 2
+ | [\xE1-\xEC][\x80-\xBF]{2}
+ | \xED[\x80-\x9F][\x80-\xBF]
+ | [\xEE-\xEF][\x80-\xBF]{2}
+ | \xF0[\x90-\xBF][\x80-\xBF]{2} # four-byte sequences 11110xxx 10xxxxxx * 3
+ | [\xF1-\xF3][\x80-\xBF]{3}
+ | \xF4[\x80-\x8F][\x80-\xBF]{2}
+ )/x';
+
+ $chars = array( '' ); // Start with 1 element instead of 0 since the first thing we do is pop
+ do {
+ // We had some string left over from the last round, but we counted it in that last round.
+ array_pop( $chars );
+
+ // Split by UTF-8 character, limit to 1000 characters (last array element will contain the rest of the string)
+ $pieces = preg_split( $regex, $str, 1000, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY );
+
+ $chars = array_merge( $chars, $pieces );
+ } while ( count( $pieces ) > 1 && $str = array_pop( $pieces ) ); // If there's anything left over, repeat the loop.
+
+ return join( '', array_slice( $chars, $start, $length ) );
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> if ( ! function_exists( 'mb_strlen' ) ) :
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -38,16 +100,54 @@
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> endif;
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+/*
+ * Only understands UTF-8 and 8bit. All other character sets will be treated as 8bit.
+ * For $encoding === UTF-8, the $str input is expected to be a valid UTF-8 byte sequence.
+ * The behavior of this function for invalid inputs is undefined.
+ */
</ins><span class="cx" style="display: block; padding: 0 10px"> function _mb_strlen( $str, $encoding = null ) {
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( null === $encoding ) {
+ $encoding = get_option( 'blog_charset' );
+ }
+
</ins><span class="cx" style="display: block; padding: 0 10px"> // The solution below works only for UTF-8,
</span><span class="cx" style="display: block; padding: 0 10px"> // so in case of a different charset just use built-in strlen()
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $charset = get_option( 'blog_charset' );
- if ( ! in_array( $charset, array( 'utf8', 'utf-8', 'UTF8', 'UTF-8' ) ) ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( ! in_array( $encoding, array( 'utf8', 'utf-8', 'UTF8', 'UTF-8' ) ) ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> return strlen( $str );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Use the regex unicode support to separate the UTF-8 characters into an array
- preg_match_all( '/./us', $str, $match );
- return count( $match[0] );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+ if ( _wp_can_use_pcre_u() ) {
+ // Use the regex unicode support to separate the UTF-8 characters into an array
+ preg_match_all( '/./us', $str, $match );
+ return count( $match[0] );
+ }
+
+ $regex = '/(?:
+ [\x00-\x7F] # single-byte sequences 0xxxxxxx
+ | [\xC2-\xDF][\x80-\xBF] # double-byte sequences 110xxxxx 10xxxxxx
+ | \xE0[\xA0-\xBF][\x80-\xBF] # triple-byte sequences 1110xxxx 10xxxxxx * 2
+ | [\xE1-\xEC][\x80-\xBF]{2}
+ | \xED[\x80-\x9F][\x80-\xBF]
+ | [\xEE-\xEF][\x80-\xBF]{2}
+ | \xF0[\x90-\xBF][\x80-\xBF]{2} # four-byte sequences 11110xxx 10xxxxxx * 3
+ | [\xF1-\xF3][\x80-\xBF]{3}
+ | \xF4[\x80-\x8F][\x80-\xBF]{2}
+ )/x';
+
+ $count = 1; // Start at 1 instead of 0 since the first thing we do is decrement
+ do {
+ // We had some string left over from the last round, but we counted it in that last round.
+ $count--;
+
+ // Split by UTF-8 character, limit to 1000 characters (last array element will contain the rest of the string)
+ $pieces = preg_split( $regex, $str, 1000 );
+
+ // Increment
+ $count += count( $pieces );
+ } while ( $str = array_pop( $pieces ) ); // If there's anything left over, repeat the loop.
+
+ // Fencepost: preg_split() always returns one extra item in the array
+ return --$count;
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> if ( !function_exists('hash_hmac') ):
</span></span></pre></div>
<a id="trunksrcwpincludesversionphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/src/wp-includes/version.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/src/wp-includes/version.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/src/wp-includes/version.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -11,7 +11,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> *
</span><span class="cx" style="display: block; padding: 0 10px"> * @global int $wp_db_version
</span><span class="cx" style="display: block; padding: 0 10px"> */
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">-$wp_db_version = 32308;
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+$wp_db_version = 32364;
</ins><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> /**
</span><span class="cx" style="display: block; padding: 0 10px"> * Holds the TinyMCE version
</span></span></pre></div>
<a id="trunksrcwpincludeswpdbphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/src/wp-includes/wp-db.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/src/wp-includes/wp-db.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/src/wp-includes/wp-db.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -1809,6 +1809,8 @@
</span><span class="cx" style="display: block; padding: 0 10px"> * @return int|false The number of rows affected, or false on error.
</span><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> function _insert_replace_helper( $table, $data, $format = null, $type = 'INSERT' ) {
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $this->insert_id = 0;
+
</ins><span class="cx" style="display: block; padding: 0 10px"> if ( ! in_array( strtoupper( $type ), array( 'REPLACE', 'INSERT' ) ) ) {
</span><span class="cx" style="display: block; padding: 0 10px"> return false;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -1829,7 +1831,6 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $sql = "$type INTO `$table` ($fields) VALUES ($formats)";
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $this->insert_id = 0;
</del><span class="cx" style="display: block; padding: 0 10px"> $this->check_current_query = false;
</span><span class="cx" style="display: block; padding: 0 10px"> return $this->query( $this->prepare( $sql, $values ) );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2021,17 +2022,11 @@
</span><span class="cx" style="display: block; padding: 0 10px"> // We can skip this field if we know it isn't a string.
</span><span class="cx" style="display: block; padding: 0 10px"> // This checks %d/%f versus ! %s because it's sprintf() could take more.
</span><span class="cx" style="display: block; padding: 0 10px"> $value['charset'] = false;
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- } elseif ( $this->check_ascii( $value['value'] ) ) {
- // If it's ASCII, then we don't need the charset. We can skip this field.
- $value['charset'] = false;
</del><span class="cx" style="display: block; padding: 0 10px"> } else {
</span><span class="cx" style="display: block; padding: 0 10px"> $value['charset'] = $this->get_col_charset( $table, $field );
</span><span class="cx" style="display: block; padding: 0 10px"> if ( is_wp_error( $value['charset'] ) ) {
</span><span class="cx" style="display: block; padding: 0 10px"> return false;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">-
- // This isn't ASCII. Don't have strip_invalid_text() re-check.
- $value['ascii'] = false;
</del><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $data[ $field ] = $value;
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2064,10 +2059,6 @@
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( false !== $value['length'] && mb_strlen( $value['value'] ) > $value['length'] ) {
- return false;
- }
-
</del><span class="cx" style="display: block; padding: 0 10px"> $data[ $field ] = $value;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2406,14 +2397,16 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> /**
</span><span class="cx" style="display: block; padding: 0 10px"> * Retrieve the maximum string length allowed in a given column.
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ * The length may either be specified as a byte length or a character length.
</ins><span class="cx" style="display: block; padding: 0 10px"> *
</span><span class="cx" style="display: block; padding: 0 10px"> * @since 4.2.1
</span><span class="cx" style="display: block; padding: 0 10px"> * @access public
</span><span class="cx" style="display: block; padding: 0 10px"> *
</span><span class="cx" style="display: block; padding: 0 10px"> * @param string $table Table name.
</span><span class="cx" style="display: block; padding: 0 10px"> * @param string $column Column name.
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- * @return mixed Max column length as an int. False if the column has no
- * length. WP_Error object if there was an error.
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ * @return mixed array( 'length' => (int), 'type' => 'byte' | 'char' )
+ * false if the column has no length (for example, numeric column)
+ * WP_Error object if there was an error.
</ins><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> public function get_col_length( $table, $column ) {
</span><span class="cx" style="display: block; padding: 0 10px"> $tablekey = strtolower( $table );
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2446,27 +2439,47 @@
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> switch( $type ) {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- case 'binary':
</del><span class="cx" style="display: block; padding: 0 10px"> case 'char':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- case 'varbinary':
</del><span class="cx" style="display: block; padding: 0 10px"> case 'varchar':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return $length;
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return array(
+ 'type' => 'char',
+ 'length' => (int) $length,
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> break;
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ case 'binary':
+ case 'varbinary':
+ return array(
+ 'type' => 'byte',
+ 'length' => (int) $length,
+ );
+ break;
</ins><span class="cx" style="display: block; padding: 0 10px"> case 'tinyblob':
</span><span class="cx" style="display: block; padding: 0 10px"> case 'tinytext':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return 255; // 2^8 - 1
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return array(
+ 'type' => 'byte',
+ 'length' => 255, // 2^8 - 1
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> break;
</span><span class="cx" style="display: block; padding: 0 10px"> case 'blob':
</span><span class="cx" style="display: block; padding: 0 10px"> case 'text':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return 65535; // 2^16 - 1
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return array(
+ 'type' => 'byte',
+ 'length' => 65535, // 2^16 - 1
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> break;
</span><span class="cx" style="display: block; padding: 0 10px"> case 'mediumblob':
</span><span class="cx" style="display: block; padding: 0 10px"> case 'mediumtext':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return 16777215; // 2^24 - 1
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return array(
+ 'type' => 'byte',
+ 'length' => 16777215, // 2^24 - 1
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> break;
</span><span class="cx" style="display: block; padding: 0 10px"> case 'longblob':
</span><span class="cx" style="display: block; padding: 0 10px"> case 'longtext':
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- return 4294967295; // 2^32 - 1
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ return array(
+ 'type' => 'byte',
+ 'length' => 4294967295, // 2^32 - 1
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> break;
</span><span class="cx" style="display: block; padding: 0 10px"> default:
</span><span class="cx" style="display: block; padding: 0 10px"> return false;
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2572,50 +2585,55 @@
</span><span class="cx" style="display: block; padding: 0 10px"> * remove invalid characters, a WP_Error object is returned.
</span><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> protected function strip_invalid_text( $data ) {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Some multibyte character sets that we can check in PHP.
- $mb_charsets = array(
- 'ascii' => 'ASCII',
- 'big5' => 'BIG-5',
- 'eucjpms' => 'eucJP-win',
- 'gb2312' => 'EUC-CN',
- 'ujis' => 'EUC-JP',
- 'utf32' => 'UTF-32',
- );
-
- $supported_charsets = array();
- if ( function_exists( 'mb_list_encodings' ) ) {
- $supported_charsets = mb_list_encodings();
- }
-
</del><span class="cx" style="display: block; padding: 0 10px"> $db_check_string = false;
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> foreach ( $data as &$value ) {
</span><span class="cx" style="display: block; padding: 0 10px"> $charset = $value['charset'];
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Column isn't a string, or is latin1, which will will happily store anything.
- if ( false === $charset || 'latin1' === $charset ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( is_array( $value['length'] ) ) {
+ $length = $value['length']['length'];
+ } else {
+ $length = false;
+ }
+
+ // There's no charset to work with.
+ if ( false === $charset ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> continue;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ // Column isn't a string.
</ins><span class="cx" style="display: block; padding: 0 10px"> if ( ! is_string( $value['value'] ) ) {
</span><span class="cx" style="display: block; padding: 0 10px"> continue;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // ASCII is always OK.
- if ( ! isset( $value['ascii'] ) && $this->check_ascii( $value['value'] ) ) {
- continue;
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $truncate_by_byte_length = 'byte' === $value['length']['type'];
+
+ $needs_validation = true;
+ if (
+ // latin1 can store any byte sequence
+ 'latin1' === $charset
+ ||
+ // ASCII is always OK.
+ ( ! isset( $value['ascii'] ) && $this->check_ascii( $value['value'] ) )
+ ) {
+ $truncate_by_byte_length = true;
+ $needs_validation = false;
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Convert the text locally.
- if ( $supported_charsets ) {
- if ( isset( $mb_charsets[ $charset ] ) && in_array( $mb_charsets[ $charset ], $supported_charsets ) ) {
- $value['value'] = mb_convert_encoding( $value['value'], $mb_charsets[ $charset ], $mb_charsets[ $charset ] );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( $truncate_by_byte_length ) {
+ mbstring_binary_safe_encoding();
+ if ( false !== $length && strlen( $value['value'] ) > $length ) {
+ $value['value'] = substr( $value['value'], 0, $length );
+ }
+ reset_mbstring_encoding();
+
+ if ( ! $needs_validation ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> continue;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> // utf8 can be handled by regex, which is a bunch faster than a DB lookup.
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( 'utf8' === $charset || 'utf8mb3' === $charset || 'utf8mb4' === $charset ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( ( 'utf8' === $charset || 'utf8mb3' === $charset || 'utf8mb4' === $charset ) && function_exists( 'mb_strlen' ) ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> $regex = '/
</span><span class="cx" style="display: block; padding: 0 10px"> (
</span><span class="cx" style="display: block; padding: 0 10px"> (?: [\x00-\x7F] # single-byte sequences 0xxxxxxx
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2625,7 +2643,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> | \xED[\x80-\x9F][\x80-\xBF]
</span><span class="cx" style="display: block; padding: 0 10px"> | [\xEE-\xEF][\x80-\xBF]{2}';
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( 'utf8mb4' === $charset) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( 'utf8mb4' === $charset ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> $regex .= '
</span><span class="cx" style="display: block; padding: 0 10px"> | \xF0[\x90-\xBF][\x80-\xBF]{2} # four-byte sequences 11110xxx 10xxxxxx * 3
</span><span class="cx" style="display: block; padding: 0 10px"> | [\xF1-\xF3][\x80-\xBF]{3}
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2638,6 +2656,11 @@
</span><span class="cx" style="display: block; padding: 0 10px"> | . # anything else
</span><span class="cx" style="display: block; padding: 0 10px"> /x';
</span><span class="cx" style="display: block; padding: 0 10px"> $value['value'] = preg_replace( $regex, '$1', $value['value'] );
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+
+ if ( false !== $length && mb_strlen( $value['value'], 'UTF-8' ) > $length ) {
+ $value['value'] = mb_substr( $value['value'], 0, $length, 'UTF-8' );
+ }
</ins><span class="cx" style="display: block; padding: 0 10px"> continue;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2654,8 +2677,14 @@
</span><span class="cx" style="display: block; padding: 0 10px"> $queries[ $value['charset'] ] = array();
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- // Split the CONVERT() calls by charset, so we can make sure the connection is right
- $queries[ $value['charset'] ][ $col ] = $this->prepare( "CONVERT( %s USING {$value['charset']} )", $value['value'] );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ // We're going to need to truncate by characters or bytes, depending on the length value we have.
+ if ( 'byte' === $value['length']['type'] ) {
+ // Split the CONVERT() calls by charset, so we can make sure the connection is right
+ $queries[ $value['charset'] ][ $col ] = $this->prepare( "CONVERT( LEFT( CONVERT( %s USING binary ), %d ) USING {$value['charset']} )", $value['value'], $value['length']['length'] );
+ } else {
+ $queries[ $value['charset'] ][ $col ] = $this->prepare( "LEFT( CONVERT( %s USING {$value['charset']} ), %d )", $value['value'], $value['length']['length'] );
+ }
+
</ins><span class="cx" style="display: block; padding: 0 10px"> unset( $data[ $col ]['db'] );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2674,16 +2703,19 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $this->check_current_query = false;
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $row = $this->get_row( "SELECT " . implode( ', ', $query ), ARRAY_N );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $sql = array();
+ foreach ( $query as $column => $column_query ) {
+ $sql[] = $column_query . " AS x_$column";
+ }
+
+ $row = $this->get_row( "SELECT " . implode( ', ', $sql ), ARRAY_A );
</ins><span class="cx" style="display: block; padding: 0 10px"> if ( ! $row ) {
</span><span class="cx" style="display: block; padding: 0 10px"> $this->set_charset( $this->dbh, $connection_charset );
</span><span class="cx" style="display: block; padding: 0 10px"> return new WP_Error( 'wpdb_strip_invalid_text_failure' );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $cols = array_keys( $query );
- $col_count = count( $cols );
- for ( $ii = 0; $ii < $col_count; $ii++ ) {
- $data[ $cols[ $ii ] ]['value'] = $row[ $ii ];
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ foreach ( array_keys( $query ) as $column ) {
+ $data[ $column ]['value'] = $row["x_$column"];
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2725,6 +2757,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => $query,
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => $charset,
</span><span class="cx" style="display: block; padding: 0 10px"> 'ascii' => false,
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => false,
</ins><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $data = $this->strip_invalid_text( array( $data ) );
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2747,7 +2780,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> * @return string|WP_Error The converted string, or a WP_Error object if the conversion fails.
</span><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> public function strip_invalid_text_for_column( $table, $column, $value ) {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- if ( ! is_string( $value ) || $this->check_ascii( $value ) ) {
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ if ( ! is_string( $value ) ) {
</ins><span class="cx" style="display: block; padding: 0 10px"> return $value;
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2764,7 +2797,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> $column => array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => $value,
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => $charset,
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'ascii' => false,
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => $this->get_col_length( $table, $column ),
</ins><span class="cx" style="display: block; padding: 0 10px"> )
</span><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span></span></pre></div>
<a id="trunktestsphpunittestscommentphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/tests/phpunit/tests/comment.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/tests/phpunit/tests/comment.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/tests/phpunit/tests/comment.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -121,7 +121,8 @@
</span><span class="cx" style="display: block; padding: 0 10px"> $_SERVER['REMOTE_ADDR'] = '';
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $post_id = $this->factory->post->create();
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $u = $this->factory->user->create();
+ $post_id = $this->factory->post->create( array( 'post_author' => $u ) );
</ins><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $data = array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'comment_post_ID' => $post_id,
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -136,8 +137,10 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $id = wp_new_comment( $data );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- $this->assertFalse( $id );
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $comment = get_comment( $id );
</ins><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ $this->assertEquals( strlen( $comment->comment_content ), 65535 );
+
</ins><span class="cx" style="display: block; padding: 0 10px"> // Cleanup.
</span><span class="cx" style="display: block; padding: 0 10px"> if ( isset( $remote_addr ) ) {
</span><span class="cx" style="display: block; padding: 0 10px"> $_SERVER['REMOTE_ADDR'] = $remote_addr;
</span></span></pre></div>
<a id="trunktestsphpunittestscompatphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/tests/phpunit/tests/compat.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/tests/phpunit/tests/compat.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/tests/phpunit/tests/compat.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -2,15 +2,168 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> /**
</span><span class="cx" style="display: block; padding: 0 10px"> * @group compat
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ * @group security-153
</ins><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> class Tests_Compat extends WP_UnitTestCase {
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- function test_mb_substr() {
- $this->assertEquals('баб', _mb_substr('баба', 0, 3));
- $this->assertEquals('баб', _mb_substr('баба', 0, -1));
- $this->assertEquals('баб', _mb_substr('баба', 0, -1));
- $this->assertEquals('I am your б', _mb_substr('I am your баба', 0, 11));
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ function utf8_string_lengths() {
+ return array(
+ // string, character_length, byte_length
+ array( 'баба', 4, 8 ),
+ array( 'баб', 3, 6 ),
+ array( 'I am your б', 11, 12 ),
+ array( '1111111111', 10, 10 ),
+ array( '²²²²²²²²²²', 10, 20 ),
+ array( '３３３３３３３３３３', 10, 30 ),
+ array( '𝟜𝟜𝟜𝟜𝟜𝟜𝟜𝟜𝟜𝟜', 10, 40 ),
+ array( '1²３𝟜1²３𝟜1²３𝟜', 12, 30 ),
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ function utf8_substrings() {
+ return array(
+ // string, start, length, character_substring, byte_substring
+ array( 'баба', 0, 3, 'баб', "б\xD0" ),
+ array( 'баба', 0, -1, 'баб', "баб\xD0" ),
+ array( 'баба', 1, null, 'аба', "\xB1аба" ),
+ array( 'баба', -3, null, 'аба', "\xB1а" ),
+ array( 'баба', -3, 2, 'аб', "\xB1\xD0" ),
+ array( 'баба', -1, 2, 'а', "\xB0" ),
+ array( 'I am your баба', 0, 11, 'I am your б', "I am your \xD0" ),
+ );
+ }
+
+ /**
+ * @dataProvider utf8_string_lengths
+ */
+ function test_mb_strlen( $string, $expected_character_length ) {
+ $this->assertEquals( $expected_character_length, _mb_strlen( $string, 'UTF-8' ) );
+ }
+
+ /**
+ * @dataProvider utf8_string_lengths
+ */
+ function test_mb_strlen_via_regex( $string, $expected_character_length ) {
+ _wp_can_use_pcre_u( false );
+ $this->assertEquals( $expected_character_length, _mb_strlen( $string, 'UTF-8' ) );
+ _wp_can_use_pcre_u( 'reset' );
+ }
+
+ /**
+ * @dataProvider utf8_string_lengths
+ */
+ function test_8bit_mb_strlen( $string, $expected_character_length, $expected_byte_length ) {
+ $this->assertEquals( $expected_byte_length, _mb_strlen( $string, '8bit' ) );
+ }
+
+ /**
+ * @dataProvider utf8_substrings
+ */
+ function test_mb_substr( $string, $start, $length, $expected_character_substring ) {
+ $this->assertEquals( $expected_character_substring, _mb_substr( $string, $start, $length, 'UTF-8' ) );
+ }
+
+ /**
+ * @dataProvider utf8_substrings
+ */
+ function test_mb_substr_via_regex( $string, $start, $length, $expected_character_substring ) {
+ _wp_can_use_pcre_u( false );
+ $this->assertEquals( $expected_character_substring, _mb_substr( $string, $start, $length, 'UTF-8' ) );
+ _wp_can_use_pcre_u( 'reset' );
+ }
+
+ /**
+ * @dataProvider utf8_substrings
+ */
+ function test_8bit_mb_substr( $string, $start, $length, $expected_character_substring, $expected_byte_substring ) {
+ $this->assertEquals( $expected_byte_substring, _mb_substr( $string, $start, $length, '8bit' ) );
+ }
+
+ function test_mb_substr_phpcore(){
+ /* https://github.com/php/php-src/blob/php-5.6.8/ext/mbstring/tests/mb_substr_basic.phpt */
+ $string_ascii = 'ABCDEF';
+ $string_mb = base64_decode('5pel5pys6Kqe44OG44Kt44K544OI44Gn44GZ44CCMDEyMzTvvJXvvJbvvJfvvJjvvJnjgII=');
+
+ $this->assertEquals( 'DEF', _mb_substr($string_ascii, 3) );
+ $this->assertEquals( 'DEF', _mb_substr($string_ascii, 3, 5, 'ISO-8859-1') );
+
+ // specific latin-1 as that is the default the core php test opporates under
+ $this->assertEquals( 'peacrOiqng==' , base64_encode( _mb_substr($string_mb, 2, 7, 'latin-1' ) ) );
+ $this->assertEquals( '6Kqe44OG44Kt44K544OI44Gn44GZ', base64_encode( _mb_substr($string_mb, 2, 7, 'utf-8') ) );
+
+ /* https://github.com/php/php-src/blob/php-5.6.8/ext/mbstring/tests/mb_substr_variation1.phpt */
+ $start = 0;
+ $length = 5;
+ $unset_var = 10;
+ unset ($unset_var);
+ $heredoc = <<<EOT
+hello world
+EOT;
+ $inputs = array(
+ /*1*/ 0,
+ 1,
+ 12345,
+ -2345,
+ // float data
+ /*5*/ 10.5,
+ -10.5,
+ 12.3456789000e10,
+ 12.3456789000E-10,
+ .5,
+ // null data
+ /*10*/ NULL,
+ null,
+ // boolean data
+ /*12*/ true,
+ false,
+ TRUE,
+ FALSE,
+ // empty data
+ /*16*/ "",
+ '',
+ // string data
+ /*18*/ "string",
+ 'string',
+ $heredoc,
+ // object data
+ /*21*/ new classA(),
+ // undefined data
+ /*22*/ @$undefined_var,
+ // unset data
+ /*23*/ @$unset_var,
+ );
+ $outputs = array(
+ "0",
+ "1",
+ "12345",
+ "-2345",
+ "10.5",
+ "-10.5",
+ "12345",
+ "1.234",
+ "0.5",
+ "",
+ "",
+ "1",
+ "",
+ "1",
+ "",
+ "",
+ "",
+ "strin",
+ "strin",
+ "hello",
+ "Class",
+ "",
+ "",
+ );
+ $iterator = 0;
+ foreach($inputs as $input) {
+ $this->assertEquals( $outputs[$iterator] , _mb_substr($input, $start, $length) );
+ $iterator++;
+ }
+
+ }
+
</ins><span class="cx" style="display: block; padding: 0 10px"> function test_hash_hmac_simple() {
</span><span class="cx" style="display: block; padding: 0 10px"> $this->assertEquals('140d1cb79fa12e2a31f32d35ad0a2723', _hash_hmac('md5', 'simple', 'key'));
</span><span class="cx" style="display: block; padding: 0 10px"> $this->assertEquals('993003b95758e0ac2eba451a4c5877eb1bb7b92a', _hash_hmac('sha1', 'simple', 'key'));
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -34,3 +187,10 @@
</span><span class="cx" style="display: block; padding: 0 10px"> $this->assertEquals( array( 'foo' ), $json->decode( '["foo"]' ) );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+/* used in test_mb_substr_phpcore */
+class classA {
+ public function __toString() {
+ return "Class A object";
+ }
+}
</ins></span></pre></div>
<a id="trunktestsphpunittestsdbcharsetphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/tests/phpunit/tests/db/charset.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/tests/phpunit/tests/db/charset.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/tests/phpunit/tests/db/charset.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -6,6 +6,7 @@
</span><span class="cx" style="display: block; padding: 0 10px"> * Test WPDB methods
</span><span class="cx" style="display: block; padding: 0 10px"> *
</span><span class="cx" style="display: block; padding: 0 10px"> * @group wpdb
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ * @group security-153
</ins><span class="cx" style="display: block; padding: 0 10px"> */
</span><span class="cx" style="display: block; padding: 0 10px"> class Tests_DB_Charset extends WP_UnitTestCase {
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -28,57 +29,227 @@
</span><span class="cx" style="display: block; padding: 0 10px"> // latin1. latin1 never changes.
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'latin1',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "\xf0\x9f\x8e\xb7",
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => "\xf0\x9f\x8e\xb7"
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => "\xf0\x9f\x8e\xb7",
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'latin1_char_length' => array(
+ // latin1. latin1 never changes.
+ 'charset' => 'latin1',
+ 'value' => str_repeat( 'A', 11 ),
+ 'expected' => str_repeat( 'A', 10 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'latin1_byte_length' => array(
+ // latin1. latin1 never changes.
+ 'charset' => 'latin1',
+ 'value' => str_repeat( 'A', 11 ),
+ 'expected' => str_repeat( 'A', 10 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'ascii' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> // ascii gets special treatment, make sure it's covered
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'ascii',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => 'Hello World',
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => 'Hello World'
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => 'Hello World',
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'ascii_char_length' => array(
+ // ascii gets special treatment, make sure it's covered
+ 'charset' => 'ascii',
+ 'value' => str_repeat( 'A', 11 ),
+ 'expected' => str_repeat( 'A', 10 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'ascii_byte_length' => array(
+ // ascii gets special treatment, make sure it's covered
+ 'charset' => 'ascii',
+ 'value' => str_repeat( 'A', 11 ),
+ 'expected' => str_repeat( 'A', 10 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'utf8' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> // utf8 only allows <= 3-byte chars
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'utf8',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "H€llo\xf0\x9f\x98\x88World¢",
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => 'H€lloWorld¢'
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => 'H€lloWorld¢',
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'utf8_23char_length' => array(
+ // utf8 only allows <= 3-byte chars
+ 'charset' => 'utf8',
+ 'value' => str_repeat( "²３", 10 ),
+ 'expected' => str_repeat( "²３", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8_23byte_length' => array(
+ // utf8 only allows <= 3-byte chars
+ 'charset' => 'utf8',
+ 'value' => str_repeat( "²３", 10 ),
+ 'expected' => "²３²３",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
+ 'utf8_3char_length' => array(
+ // utf8 only allows <= 3-byte chars
+ 'charset' => 'utf8',
+ 'value' => str_repeat( "３", 11 ),
+ 'expected' => str_repeat( "３", 10 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8_3byte_length' => array(
+ // utf8 only allows <= 3-byte chars
+ 'charset' => 'utf8',
+ 'value' => str_repeat( "３", 11 ),
+ 'expected' => "３３３",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'utf8mb3' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> // utf8mb3 should behave the same an utf8
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'utf8mb3',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "H€llo\xf0\x9f\x98\x88World¢",
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => 'H€lloWorld¢'
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => 'H€lloWorld¢',
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'utf8mb3_23char_length' => array(
+ // utf8mb3 should behave the same an utf8
+ 'charset' => 'utf8mb3',
+ 'value' => str_repeat( "²３", 10 ),
+ 'expected' => str_repeat( "²３", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8mb3_23byte_length' => array(
+ // utf8mb3 should behave the same an utf8
+ 'charset' => 'utf8mb3',
+ 'value' => str_repeat( "²３", 10 ),
+ 'expected' => "²３²３",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
+ 'utf8mb3_3char_length' => array(
+ // utf8mb3 should behave the same an utf8
+ 'charset' => 'utf8mb3',
+ 'value' => str_repeat( "３", 11 ),
+ 'expected' => str_repeat( "３", 10 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8mb3_3byte_length' => array(
+ // utf8mb3 should behave the same an utf8
+ 'charset' => 'utf8mb3',
+ 'value' => str_repeat( "３", 10 ),
+ 'expected' => "３３３",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'utf8mb4' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> // utf8mb4 allows 4-byte characters, too
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'utf8mb4',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "H€llo\xf0\x9f\x98\x88World¢",
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => "H€llo\xf0\x9f\x98\x88World¢"
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => "H€llo\xf0\x9f\x98\x88World¢",
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'utf8mb4_234char_length' => array(
+ // utf8mb4 allows 4-byte characters, too
+ 'charset' => 'utf8mb4',
+ 'value' => str_repeat( "²３𝟜", 10 ),
+ 'expected' => "²３𝟜²３𝟜²３𝟜²",
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8mb4_234byte_length' => array(
+ // utf8mb4 allows 4-byte characters, too
+ 'charset' => 'utf8mb4',
+ 'value' => str_repeat( "²３𝟜", 10 ),
+ 'expected' => "²３𝟜",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
+ 'utf8mb4_4char_length' => array(
+ // utf8mb4 allows 4-byte characters, too
+ 'charset' => 'utf8mb4',
+ 'value' => str_repeat( "𝟜", 11 ),
+ 'expected' => str_repeat( "𝟜", 10 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'utf8mb4_4byte_length' => array(
+ // utf8mb4 allows 4-byte characters, too
+ 'charset' => 'utf8mb4',
+ 'value' => str_repeat( "𝟜", 10 ),
+ 'expected' => "𝟜𝟜",
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'koi8r' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'koi8r',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "\xfdord\xf2ress",
</span><span class="cx" style="display: block; padding: 0 10px"> 'expected' => "\xfdord\xf2ress",
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'koi8r_char_length' => array(
+ 'charset' => 'koi8r',
+ 'value' => str_repeat( "\xfd\xf2", 10 ),
+ 'expected' => str_repeat( "\xfd\xf2", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'koi8r_byte_length' => array(
+ 'charset' => 'koi8r',
+ 'value' => str_repeat( "\xfd\xf2", 10 ),
+ 'expected' => str_repeat( "\xfd\xf2", 5 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'hebrew' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'hebrew',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "\xf9ord\xf7ress",
</span><span class="cx" style="display: block; padding: 0 10px"> 'expected' => "\xf9ord\xf7ress",
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'hebrew_char_length' => array(
+ 'charset' => 'hebrew',
+ 'value' => str_repeat( "\xf9\xf7", 10 ),
+ 'expected' => str_repeat( "\xf9\xf7", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'hebrew_byte_length' => array(
+ 'charset' => 'hebrew',
+ 'value' => str_repeat( "\xf9\xf7", 10 ),
+ 'expected' => str_repeat( "\xf9\xf7", 5 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'cp1251' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'cp1251',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "\xd8ord\xd0ress",
</span><span class="cx" style="display: block; padding: 0 10px"> 'expected' => "\xd8ord\xd0ress",
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'cp1251_char_length' => array(
+ 'charset' => 'cp1251',
+ 'value' => str_repeat( "\xd8\xd0", 10 ),
+ 'expected' => str_repeat( "\xd8\xd0", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'cp1251_byte_length' => array(
+ 'charset' => 'cp1251',
+ 'value' => str_repeat( "\xd8\xd0", 10 ),
+ 'expected' => str_repeat( "\xd8\xd0", 5 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'tis620' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'tis620',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => "\xccord\xe3ress",
</span><span class="cx" style="display: block; padding: 0 10px"> 'expected' => "\xccord\xe3ress",
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'tis620_char_length' => array(
+ 'charset' => 'tis620',
+ 'value' => str_repeat( "\xcc\xe3", 10 ),
+ 'expected' => str_repeat( "\xcc\xe3", 5 ),
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ ),
+ 'tis620_byte_length' => array(
+ 'charset' => 'tis620',
+ 'value' => str_repeat( "\xcc\xe3", 10 ),
+ 'expected' => str_repeat( "\xcc\xe3", 5 ),
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ ),
</ins><span class="cx" style="display: block; padding: 0 10px"> 'false' => array(
</span><span class="cx" style="display: block; padding: 0 10px"> // false is a column with no character set (ie, a number column)
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => false,
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => 100,
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => 100
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => 100,
+ 'length' => false,
</ins><span class="cx" style="display: block; padding: 0 10px"> ),
</span><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -94,8 +265,23 @@
</span><span class="cx" style="display: block; padding: 0 10px"> $fields['big5'] = array(
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => 'big5',
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => $big5,
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'expected' => $big5
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'expected' => $big5,
+ 'length' => array( 'type' => 'char', 'length' => 100 ),
</ins><span class="cx" style="display: block; padding: 0 10px"> );
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+ $fields['big5_char_length'] = array(
+ 'charset' => 'big5',
+ 'value' => str_repeat( $big5, 10 ),
+ 'expected' => str_repeat( $big5, 3 ) . 'a',
+ 'length' => array( 'type' => 'char', 'length' => 10 ),
+ );
+
+ $fields['big5_byte_length'] = array(
+ 'charset' => 'big5',
+ 'value' => str_repeat( $big5, 10 ),
+ 'expected' => str_repeat( $big5, 2 ) . 'a',
+ 'length' => array( 'type' => 'byte', 'length' => 10 ),
+ );
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> // The data above is easy to edit. Now, prepare it for the data provider.
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -170,14 +356,14 @@
</span><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $all_ascii_fields = array(
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'post_content' => array( 'value' => 'foo foo foo!', 'format' => '%s', 'charset' => false ),
- 'post_excerpt' => array( 'value' => 'bar bar bar!', 'format' => '%s', 'charset' => false ),
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'post_content' => array( 'value' => 'foo foo foo!', 'format' => '%s', 'charset' => $charset ),
+ 'post_excerpt' => array( 'value' => 'bar bar bar!', 'format' => '%s', 'charset' => $charset ),
</ins><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> // This is the same data used in process_field_charsets_for_nonexistent_table()
</span><span class="cx" style="display: block; padding: 0 10px"> $non_ascii_string_fields = array(
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'post_content' => array( 'value' => '¡foo foo foo!', 'format' => '%s', 'charset' => $charset, 'ascii' => false ),
- 'post_excerpt' => array( 'value' => '¡bar bar bar!', 'format' => '%s', 'charset' => $charset, 'ascii' => false ),
</del><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+ 'post_content' => array( 'value' => '¡foo foo foo!', 'format' => '%s', 'charset' => $charset ),
+ 'post_excerpt' => array( 'value' => '¡bar bar bar!', 'format' => '%s', 'charset' => $charset ),
</ins><span class="cx" style="display: block; padding: 0 10px"> );
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> $vars = get_defined_vars();
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -544,4 +730,16 @@
</span><span class="cx" style="display: block; padding: 0 10px">
</span><span class="cx" style="display: block; padding: 0 10px"> self::$_wpdb->query( $drop );
</span><span class="cx" style="display: block; padding: 0 10px"> }
</span><ins style="background-color: #dfd; text-decoration:none; display:block; padding: 0 10px">+
+ function test_strip_invalid_test_for_column_bails_if_ascii_input_too_long() {
+ global $wpdb;
+
+ // TEXT column
+ $stripped = $wpdb->strip_invalid_text_for_column( $wpdb->comments, 'comment_content', str_repeat( 'A', 65536 ) );
+ $this->assertEquals( 65535, strlen( $stripped ) );
+
+ // VARCHAR column
+ $stripped = $wpdb->strip_invalid_text_for_column( $wpdb->comments, 'comment_agent', str_repeat( 'A', 256 ) );
+ $this->assertEquals( 255, strlen( $stripped ) );
+ }
</ins><span class="cx" style="display: block; padding: 0 10px"> }
</span></span></pre></div>
<a id="trunktestsphpunittestsdbphp"></a>
<div class="modfile"><h4 style="background-color: #eee; color: inherit; margin: 1em 0; padding: 1.3em; font-size: 115%">Modified: trunk/tests/phpunit/tests/db.php</h4>
<pre class="diff"><span>
<span class="info" style="display: block; padding: 0 10px; color: #888">--- trunk/tests/phpunit/tests/db.php 2015-05-06 02:29:37 UTC (rev 32363)
+++ trunk/tests/phpunit/tests/db.php 2015-05-06 02:59:50 UTC (rev 32364)
</span><span class="lines" style="display: block; padding: 0 10px; color: #888">@@ -746,7 +746,6 @@
</span><span class="cx" style="display: block; padding: 0 10px"> 'value' => '¡foo foo foo!',
</span><span class="cx" style="display: block; padding: 0 10px"> 'format' => '%s',
</span><span class="cx" style="display: block; padding: 0 10px"> 'charset' => $expected_charset,
</span><del style="background-color: #fdd; text-decoration:none; display:block; padding: 0 10px">- 'ascii' => false,
</del><span class="cx" style="display: block; padding: 0 10px"> 'length' => $wpdb->get_col_length( $wpdb->posts, 'post_content' ),
</span><span class="cx" style="display: block; padding: 0 10px"> )
</span><span class="cx" style="display: block; padding: 0 10px"> );
</span></span></pre>
</div>
</div>
</body>
</html>