This includes a new test in the emoji loader, to see if the browser supports rendering Unicode 9.0 emoji, as well as an update to `twemoji.js`, to load Unicode 9.0 twemoji images.
The default CDN location for loading Twemoji images has changed, moving to a versioned subdirectory. This allows updates without needing to purge the CDN.
Fixes#37361.
git-svn-id: https://develop.svn.wordpress.org/trunk@38087 602fd350-edb4-49c9-b593-d223f7449a82
When `$size` is passed to `image_get_intermediate_size()` as an array of width
and height values and an exact image size matching those values isn't available,
the function loops through the available attachment sizes and returns the
smallest image larger than the requested dimensions with the same aspect ratio.
The aspect ratio check is skipped for the 'thumbnail' size to provide a fallback
for small sizes when no other image option is available. This resulted in a poor
selection when the size requested was smaller than the 'thumbnail' dimensions
but a larger size matching the requested ratio existed.
This refactors the internals of `image_get_intermediate_size()` to ensure the
'thumbnail' size is only returned as a fallback to small sizes once all other
options have been considered, and makes the control flow easier to follow.
This also introduces a new helper function, `wp_image_matches_ratio()` for
testing whether the aspect ratios of two sets of dimensions match. This function
is also now used in `wp_calculate_image_srcset()` during the selection process.
Props flixos, joemcgill.
Fixes#34384, #34980.
git-svn-id: https://develop.svn.wordpress.org/trunk@38086 602fd350-edb4-49c9-b593-d223f7449a82
Improves the `containFocus()` function to always get the correct first and last
focusable elements, even when the theme browser shows the active theme details.
Also, when on the first and last theme, adds a `disabled` attribute on the
previous and next navigation buttons to make them not focusable.
Fixes#37383.
git-svn-id: https://develop.svn.wordpress.org/trunk@38084 602fd350-edb4-49c9-b593-d223f7449a82
This ensures that the `wp.customize.selectiveRefresh` object will be available when in the Customizer preview.
Fixes#37390.
Props westonruter, celloexpressions.
git-svn-id: https://develop.svn.wordpress.org/trunk@38083 602fd350-edb4-49c9-b593-d223f7449a82
* Remove title attributes.
* Show "Install Now" and "Details" links if the importer isn't installed yet.
* Show a "Run Importer" link if the importer is installed. It also handles activation if the plugin isn't activated.
* Add `aria-label` attributes to each link.
* Unify the importer descriptions to make them independent from the plugin state. The API was changed in [meta3690].
* Adjust JavaScript callbacks for ajaxified importer installs.
Props afercia, swissspidy, ocean90.
See #24766.
Fixes#35191.
git-svn-id: https://develop.svn.wordpress.org/trunk@38075 602fd350-edb4-49c9-b593-d223f7449a82
[36874] changed the `get_term()` call so that no `$taxonomy` parameter
was passed, as 4.4 made the parameter optional. This change made it
impossible to access a shared term that has not yet been splitr, since
passing an ambiguous `$term_id` to `get_term()` results in an error.
Restoring the `$taxonomy` parameter fixes the regression.
Props alleynoah, dlh.
Fixes#37205.
git-svn-id: https://develop.svn.wordpress.org/trunk@38069 602fd350-edb4-49c9-b593-d223f7449a82
Previously, when inserting an image from a URL, leaving the `alt`
field blank in the media modal would result in an image being
inserted into the editor without an `alt` attribute, rather than
an empty `alt`. This happened because the `props.type` would not
get set in `wp.media.string.props()` — because `attachment` is
undefined in this case — causing the image fallbacks to get
skipped.
This fixes the issue by explicitly setting `props.type` to 'image'
in `wp.media.string.image()` before filling out the rest of the
properties.
Props ambrosey, dabnpits.
Fixes#36735.
git-svn-id: https://develop.svn.wordpress.org/trunk@38065 602fd350-edb4-49c9-b593-d223f7449a82
Previously, `wp_mail()` implemented Reply-To as a generic header, using
PHPMailer's `addCustomHeader()`. As such, the email address portion of
the header was being incorrectly encoded when the name portion
contained UTF-8 characters. Switching to PHPMailer's more specific
`addReplyTo()` method fixes the issue.
For greater readability, the handling of all address-related headers
(To, CC, BCC, Reply-To) has been standardized.
Props szepe.viktor, iandunn, bpetty, stephenharris.
Fixes#21659.
git-svn-id: https://develop.svn.wordpress.org/trunk@38058 602fd350-edb4-49c9-b593-d223f7449a82
`WP_HTTP_Proxy()` is used directly in `WP_Http_Curl()` and `WP_Http_Streams()`. Since `WP_Http::request()` doesn't use them anymore we have to move the proxy handling into `WP_Http::request()` so the proxy data can be passed to `Requests::request()`.
Props rmccue.
See #33055.
Fixes#37107.
git-svn-id: https://develop.svn.wordpress.org/trunk@38054 602fd350-edb4-49c9-b593-d223f7449a82
In some cases, images in the media library may contain spaces in
their filenames. This results in an invalid `srcset` attribute,
causing broken images on the front end. This change fixes the issue
by replacing spaces in URLs with URL encoded '%20' characters before
returning the `srcset` string.
Props underdude, joemcgill.
Fixes#36549.
git-svn-id: https://develop.svn.wordpress.org/trunk@38052 602fd350-edb4-49c9-b593-d223f7449a82
Also adds a missing initial `@since` version for `wp_xmlrpc_server::_prepare_post_type()`.
See [37890]. See #36217.
git-svn-id: https://develop.svn.wordpress.org/trunk@38051 602fd350-edb4-49c9-b593-d223f7449a82
After [38047], also include the comment_content with html in the preg_match against blacklist keys to match urls.
Props ocean90.
Fixes#37208.
git-svn-id: https://develop.svn.wordpress.org/trunk@38048 602fd350-edb4-49c9-b593-d223f7449a82
Use `wp_kses()` to clean comment_content for preg_match against the blacklist_keys. Also includes some initial unit tests for `wp_blacklist_check()`.
Previously, if a blacklisted key was used in comment_content split by an html tag the regex in `wp_blacklist_check()` would not find a match. Example: Where "springfield" was a blacklisted word, if the content of a comment included `spring<i>field</i>" `wp_blacklist_check()` would not return true.
Props cfinke.
Fixes#37208.
git-svn-id: https://develop.svn.wordpress.org/trunk@38047 602fd350-edb4-49c9-b593-d223f7449a82
The empty string was not incorrect. Using `array()` here instead makes things a bit more consistent by aligning with `get_sites()`, `get_users()`, and `get_terms()`.
See #32504.
git-svn-id: https://develop.svn.wordpress.org/trunk@38042 602fd350-edb4-49c9-b593-d223f7449a82
When using `register_meta()` with the function signature from 4.5 and earlier, the `auth_{$type}_meta_{$key}` and `sanitize_{$type}_meta_{$key}` filters are used. Any calls to `register_meta()` expecting this behavior should continue to work. The new filters, which take advantage of object subtypes, should not be added unless the proper `$args` array is passed.
See #35658.
git-svn-id: https://develop.svn.wordpress.org/trunk@38041 602fd350-edb4-49c9-b593-d223f7449a82
