Normalizes UTF-8 input to Normalization Form C.
This is a souped-up version of the Normalizer plugin by Torsten Landsiedel.
It adds WP filters to normalize UTF-8 data coming into the system to the de facto web standard Normalization Form C. The Unicode Consortium report is at Unicode Normalization Forms.
For best performance install (if possible) the PHP Internationalization extension Intl
, which includes the PHP class Normalizer
.
However the plugin works without the PHP Intl
extension being installed, as it uses (a modified version of) the Symfony Normalizer
polyfill.
Also text pasted into (most) inputs is normalized immediately using the JavaScript normalize()
method. For browsers without normalization support, the unorm polyfill is used.
For further info, see the WP Trac ticket #30130 Normalize characters with combining marks to precomposed characters and this Make WP Core comment.
The plugin also works around the Safari-specific issues raised by the WP Trac ticket #22363 Accents in attachment filenames should be sanitized.
For existing data, the plugin includes an administration tool to scan and normalize the database. Important: before using this tool to normalize, please backup your database. This is especially important if your database contains non-normalized serialized data, as this plugin uses the same suck-and-see technique as interconnect/it’s Database Search and Replace Script in PHP to deal with serialized data, and is fallible.
A google-cheating schoolboy French translation is supplied.
The plugin should run on PHP 5.2.17 to 7.1, and on WP 3.9.13 to 4.7.1.
The project is on github.
Install in the standard way via the ‘Plugins’ menu in WordPress and then activate. No further action is needed.
Before and after: pasting non-normalized text into a post's title and content.
Before: uploading a file on macOS with Firefox and then viewing with Safari (broken in Safari).
After: uploading a file on macOS with Firefox and then viewing with Safari (whee it works).
Front page of the database scan administration tool.
You can add normalization to anything that passes its content through a filter. The canonical way is to use the unfc_extra_filters
filter which returns an array of filter names – for instance, in your theme’s functions.php
file, add:
function mytheme_unfc_extra_filters( $extra_filters ) { $extra_filters[] = 'myfilter'; return $extra_filters; } add_filter( 'unfc_extra_filters', 'mytheme_unfc_extra_filters' );
Note that the unfc_extra_filters
filter is only called in the administration backend. You can also add a filter directly, to be called in the frontend or backend, by referencing the global PHP variable unfc_normalize
, but you should ensure that the Normalizer
polyfill is loaded if you don’t have the PHP Intl
extension installed:
global $unfc_normalize; if ( $unfc_normalize ) { if ( ! function_exists( 'normalizer_is_normalized' ) ) { // If the "Intl" extension is not installed... $unfc_normalize->load_unfc_normalizer_class(); // ...load the polyfill. } add_filter( 'myfilter', array( $unfc_normalize, 'normalize' ), 6 /* Or whatever priority you choose */ ); }
No, it doesn’t. Alas, such files would have to re-uploaded. However if (enough – or more like any) interest were expressed…