From 06b672cb340f1774fb3abe6ed2486aaaf7d841d1 Mon Sep 17 00:00:00 2001 From: "Dominik Schilling (ocean90)" Date: Mon, 1 Feb 2016 14:57:53 +0000 Subject: [PATCH] Media: In `wp_read_image_metadata()` make sure that IPTC keywords are UTF8 encoded. Prevents missing `_wp_attachment_metadata` when an image contains keywords with latin extended characters. Merges [36429] to the 4.4 branch. See #35316. git-svn-id: https://develop.svn.wordpress.org/branches/4.4@36430 602fd350-edb4-49c9-b593-d223f7449a82 --- src/wp-admin/includes/image.php | 8 +++++--- src/wp-includes/kses.php | 13 +++++++++++++ 2 files changed, 18 insertions(+), 3 deletions(-) diff --git a/src/wp-admin/includes/image.php b/src/wp-admin/includes/image.php index 139d405070..2faccfe71a 100644 --- a/src/wp-admin/includes/image.php +++ b/src/wp-admin/includes/image.php @@ -408,12 +408,14 @@ function wp_read_image_metadata( $file ) { } } - foreach ( $meta as &$value ) { - if ( is_string( $value ) ) { - $value = wp_kses_post( $value ); + foreach ( $meta['keywords'] as $key => $keyword ) { + if ( ! seems_utf8( $keyword ) ) { + $meta['keywords'][ $key ] = utf8_encode( $keyword ); } } + $meta = wp_kses_post_deep( $meta ); + /** * Filter the array of meta data read from an image's exif data. * diff --git a/src/wp-includes/kses.php b/src/wp-includes/kses.php index d34608b89f..4827b0de2c 100644 --- a/src/wp-includes/kses.php +++ b/src/wp-includes/kses.php @@ -1580,6 +1580,19 @@ function wp_kses_post( $data ) { return wp_kses( $data, 'post' ); } +/** + * Navigates through an array, object, or scalar, and sanitizes content for + * allowed HTML tags for post content. + * + * @since 4.4.2 + * + * @param mixed $value The array or string to filter. + * @return mixed $value The filtered content. + */ +function wp_kses_post_deep( $data ) { + return map_deep( $data, 'wp_kses_post' ); +} + /** * Strips all of the HTML in the content. *