[7520] introduced an undocumented feature whereby developers could
register a custom taxonomy with an 'args' parameter, consisting of
an array of config params that, when present, override corresponding
params in the `$args` array passed to `wp_get_object_terms()` when
using that function to query for terms in the specified taxonomy.
The `wp_get_object_terms()` refactor in [38667] failed to respect
this secret covenant, and the current changeset atones for the
transgression.
Props danielbachhuber.
Fixes#40496.
Built from https://develop.svn.wordpress.org/trunk@40513
git-svn-id: http://core.svn.wordpress.org/trunk@40389 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Prior to 4.7, term relationships - as set by `wp_set_object_terms()` or
`wp_remove_object_terms()` - did not affect the term query cache. The
introduction of the 'object_ids' parameter in 4.7 means that the query
cache must be aware of object-term relationships. As such, the
'last_changed' incrementor is now invalidated when term relationships
are modified.
This bug only reared its head when delaying term counting, because term
counting performs its own term query cache invalidation.
Props mboynes.
Fixes#40306.
Built from https://develop.svn.wordpress.org/trunk@40353
git-svn-id: http://core.svn.wordpress.org/trunk@40260 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Prior to [38677], `get_term_by()` would always return false if
an empty string were passed as the queried 'name' or 'slug'. The
refactor to use `get_terms()` broke this behavior; inappropriately
imprecise `empty()` checks caused the 'name' or 'slug' clause to be
discarded altogether when fetching terms, resulting in an incorrect
term being returned from the function.
We fix the regression by special-casing truly empty values passed
to `get_term_by()`, and ensuring that `WP_Term_Query` is properly
able to handle `0` and `'0'` term queries.
Props sstoqnov.
Fixes#21760.
Built from https://develop.svn.wordpress.org/trunk@40293
git-svn-id: http://core.svn.wordpress.org/trunk@40200 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Historically, it has been possible to call `get_term_by()` within
a 'get_terms' filter callback. Since `get_term_by()` was refactored
to use `get_terms()` internally [38677], callbacks of this nature
have resulted in infinite loops.
As a workaround, we introduce a 'suppress_filter' option to `get_terms()`,
and use it when calling the function from within `get_term_by()`.
Props ocean90.
See #21760.
Built from https://develop.svn.wordpress.org/trunk@40275
git-svn-id: http://core.svn.wordpress.org/trunk@40192 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Terms with duplicate names are not allowed at the same level of a
taxonomy hierarchy. The name lookup introduced in [34809] did not
properly account for the 'parent' parameter, with the result that
the duplicate-name restriction was tighter than intended (terms
with duplicate names could not be created at different levels of
a single hierarchy).
Props mikejolley.
Fixes#39984.
Built from https://develop.svn.wordpress.org/trunk@40145
git-svn-id: http://core.svn.wordpress.org/trunk@40084 1a063a9b-81f0-0310-95a4-ce76da25c4cd
`wp_insert_term()` does not allow for terms with the same name to exist
at the same hierarchy level, unless the second term has a unique slug.
When this logic was refactored in [31792] and [34809], a bug was
introduced whereby it was possible to bypass the no-same-named-sibling
check in cases where the first term had a non-auto-generated slug
(ie, where the name was 'Foo' but the slug something other than 'foo',
such that the second term would get the non-matching slug 'foo').
This changeset fixes this issue by ensuring that the duplicate name
check runs both in cases where there's an actual slug clash *and* in
cases where no explicit `slug` has been provided to `wp_insert_term()`.
The result is a more reliable error condition:
`wp_insert_term( 'Foo' ... )` will always fail if there's a sibling
'Foo', regardless of the sibling's slug.
Props mikejolley.
See #39984.
Built from https://develop.svn.wordpress.org/trunk@40144
git-svn-id: http://core.svn.wordpress.org/trunk@40083 1a063a9b-81f0-0310-95a4-ce76da25c4cd
REST API endpoints for your WordPress content. These endpoints provide machine-readable external access to your WordPress site with a clear, standards-driven interface, allowing new and innovative apps for interacting with your site. These endpoints support all of the following:
- Posts: Read and write access to all post data, for all types of post-based data, including pages and media.
- Comments: Read and write access to all comment data. This includes pingbacks and trackbacks.
- Terms: Read and write access to all term data.
- Users: Read and write access to all user data. This includes public access to some data for post authors.
- Meta: Read and write access to metadata for posts, comments, terms, and users, on an opt-in basis from plugins.
- Settings: Read and write access to settings, on an opt-in basis from plugins and core. This enables API management of key site content values that are technically stored in options, such as site title and byline.
Love your REST API, WordPress! The infrastructure says, "Let's do lunch!" but the content API endpoints say, "You're paying!"
Props rmccue, rachelbaker, danielbachhuber, joehoyle, adamsilverstein, afurculita, ahmadawais, airesvsg, alisspers, antisilent, apokalyptik, artoliukkonen, attitude, boonebgorges, bradyvercher, brianhogg, caseypatrickdriscoll, chopinbach, chredd, christianesperar, chrisvanpatten, claudiolabarbera, claudiosmweb, cmmarslender, codebykat, coderkevin, codfish, codonnell822, daggerhart, danielpunkass, davidbhayes, delphinus, desrosj, dimadin, dotancohen, DrewAPicture, Dudo1985, duncanjbrown, eherman24, eivhyl, eliorivero, elyobo, en-alis, ericandrewlewis, ericpedia, evansobkowicz, fjarrett, frozzare, georgestephanis, greatislander, guavaworks, hideokamoto, hkdobrev, hubdotcom, hurtige, iandunn, ircrash, ironpaperweight, iseulde, Japh, jaredcobb, JDGrimes, jdolan, jdoubleu, jeremyfelt, jimt, jjeaton, jmusal, jnylen0, johanmynhardt, johnbillion, jonathanbardo, jorbin, joshkadis, JPry, jshreve, jtsternberg, JustinSainton, kacperszurek, kadamwhite, kalenjohnson, kellbot, kjbenk, kokarn, krogsgard, kuchenundkakao, kuldipem, kwight, lgedeon, lukepettway, mantismamita, markoheijnen, matrixik, mattheu, mauteri, maxcutler, mayukojpn, michael-arestad, miyauchi, mjbanks, modemlooper, mrbobbybryant, NateWr, nathanrice, netweb, NikV, nullvariable, oskosk, oso96_2000, oxymoron, pcfreak30, pento, peterwilsoncc, Pezzab, phh, pippinsplugins, pjgalbraith, pkevan, pollyplummer, pushred, quasel, QWp6t, schlessera, schrapel, Shelob9, shprink, simonlampen, Soean, solal, tapsboy, tfrommen, tharsheblows, thenbrent, tierra, tlovett1, tnegri, tobych, Toddses, toro_unit, traversal, vanillalounge, vishalkakadiya, wanecek, web2style, webbgaraget, websupporter, westonruter, whyisjake, wonderboymusic, wpsmith, xknown, zyphonic.
Fixes#38373.
Built from https://develop.svn.wordpress.org/trunk@38832
git-svn-id: http://core.svn.wordpress.org/trunk@38775 1a063a9b-81f0-0310-95a4-ce76da25c4cd
[38776] introduced a call to `get_term()` using only the term ID. This
causes problems in cases where shared terms have not been split. Since
we have the taxonomy available, there's no harm in passing it along to
`get_term()`.
Props dd32.
See #37291.
Built from https://develop.svn.wordpress.org/trunk@38779
git-svn-id: http://core.svn.wordpress.org/trunk@38722 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Since [37573], `get_object_term_cache()` has expected term IDs to be
stored in the taxonomy relationship cache. The function would then
reach directly into the 'terms' cache to fetch the data corresponding
to a given term, before returning a `WP_Term` object. This caused
problems when, for one reason or another, term data was cached
inconsistently:
* If the 'terms' cache is empty for a given term ID, despite the earlier call to `_prime_term_caches()`, `get_term()` would return an error object.
* If the array of cached term IDs contains an invalid ID, `get_term()` would return an error object.
We avoid these errors by no longer touching the 'terms' cache directly,
but running term IDs through `get_term()` and allowing that function to
reference the cache (and database, as needed). If `get_term()` returns
an error object for any of the cached term IDs, `get_object_term_cache()`
will return that error object alone. This change ensures that upstream
functions, like `get_the_terms()`, return `WP_Error` objects in a
predictable fashion.
Props dd32, michalzuber.
Fixes#37291.
Built from https://develop.svn.wordpress.org/trunk@38776
git-svn-id: http://core.svn.wordpress.org/trunk@38719 1a063a9b-81f0-0310-95a4-ce76da25c4cd
This changes the global `$wp_taxonomies` to an array of `WP_Taxonomy ` objects. `WP_Taxonomy ` includes methods to handle rewrite rules and hooks.
Each taxonomy argument becomes a property of `WP_Taxonomy`. Introducing such a class makes further improvements in the future much more feasible.
Props boonebgorges for review.
Fixes#36224. See #36217.
Built from https://develop.svn.wordpress.org/trunk@38747
git-svn-id: http://core.svn.wordpress.org/trunk@38690 1a063a9b-81f0-0310-95a4-ce76da25c4cd
This introduces the singular `edit_term`, `delete_term`, and `assign_term` meta capabilities for terms, and switches the base capability name for tags from `manage_categories` to `manage_post_tags` and the corresponding `edit_post_tags`, `delete_post_tags`, and `assign_post_tags`.
All of these capabilities ultimately map to `manage_categories` so by default there is no change in the behaviour of the capabilities for categories, tags, or custom taxonomies. The `map_meta_cap` filter and the `capabilities` argument when registering a taxonomy now allow for control over editing, deleting, and assigning individual terms, as well as a separation of capabilities for tags from those of categories.
Fixes#35614
Props johnjamesjacoby for feedback
Built from https://develop.svn.wordpress.org/trunk@38698
git-svn-id: http://core.svn.wordpress.org/trunk@38641 1a063a9b-81f0-0310-95a4-ce76da25c4cd
`WP_Term_Query` already supports querying terms by 'slug', 'name', and 'term_taxonomy_id'. Its additional arguments allow us to generate nearly the same SQL queries as before.
This change has one yuge benefit: the term queries are now cached.
Add tests to increase coverage of `get_term_by()`.
Props spacedmonkey, boonebgorges, johnjamesjacoby, pento, ocean90.
Fixes#21760.
Built from https://develop.svn.wordpress.org/trunk@38677
git-svn-id: http://core.svn.wordpress.org/trunk@38620 1a063a9b-81f0-0310-95a4-ce76da25c4cd
The new 'object_ids' parameter for `WP_Term_Query` allows queries for
terms that "belong to" a given object. This change makes it possible
to use `WP_Term_Query` inside of `wp_get_object_terms()`, rather than
assembling a SQL query.
The refactor has a couple of benefits:
* Less redundancy.
* Better consistency in accepted arguments between the term query functions. See #31105.
* Less redundancy.
* Object term queries are now cached. The `get_object_term_cache()` cache remains, and will be a somewhat less fragile secondary cache in front of the query cache (which is subject to frequent invalidation).
* Less redundancy.
A small breaking change: Previously, if a non-hierarchical taxonomy had
terms that had a non-zero 'parent' (perhaps because of a direct SQL
query), `wp_get_object_terms()` would respect the 'parent' argument.
This is in contrast to `WP_Term_Query` and `get_terms()`, which have
always rejected 'parent' queries for non-hierarchical taxonomies. For
consistency, the behavior of `get_terms()` is being applied across the
board: passing 'parent' for a non-hierarchical taxonomy will result in
an empty result set (since the cached taxonomy hierarchy will be empty).
Props flixos90, boonebgorges.
See #37198.
Built from https://develop.svn.wordpress.org/trunk@38667
git-svn-id: http://core.svn.wordpress.org/trunk@38610 1a063a9b-81f0-0310-95a4-ce76da25c4cd
By adding the `$output` parameter to `get_attachment_taxonomies()`, the
function signature matches that of `get_object_taxonomies()`. The change
also allows for more consistent behavior when passing `output=objects`
to `get_object_taxonomies()` for the 'attachment' object type, since
the `$output` parameter is now passed through the function stack.
Props codemovement.pk.
See #37368.
Built from https://develop.svn.wordpress.org/trunk@38292
git-svn-id: http://core.svn.wordpress.org/trunk@38233 1a063a9b-81f0-0310-95a4-ce76da25c4cd
[37623] used the wrong parameter name (count=true instead of fields=count).
For greater flexibility and forward compatibility with other potential changes
to the return value of `get_terms()`, we now do a looser check: any non-array
value is excluded from the filter.
Fixes#36992.
Built from https://develop.svn.wordpress.org/trunk@37634
git-svn-id: http://core.svn.wordpress.org/trunk@37602 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Use of the 'get_terms' filter was consolidated in [37572], with the
introduction of `WP_Term_Query`. At that time, the result of 'count=true'
queries began being filtered by 'get_terms'. This breaks existing 'get_terms'
callbacks, which often assume that the returned value will be an array or a
`WP_Error` object.
Props JustinSainton.
Fixes#36992.
Built from https://develop.svn.wordpress.org/trunk@37623
git-svn-id: http://core.svn.wordpress.org/trunk@37591 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Prior to [37572], arguments passed to `get_terms()` were passed immediately
through `wp_parse_args()`, which made it possible to pass arguments as a
querystring (`hide_empty=0`) rather than an array
(`array( 'hide_empty' => false )`). [37572] moved default argument parsing
into `WP_Term_Query`, while assuming that arguments passed to `get_terms()`
would be formatted as an array.
To provide compatibility, we now parse all args passed to `get_terms()` into
an array before processing.
See #35381.
Built from https://develop.svn.wordpress.org/trunk@37599
git-svn-id: http://core.svn.wordpress.org/trunk@37567 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Since [37573], object relationship caches (`{$taxonomy}_relationships`)
contain term IDs rather than term objects. See #36814. As such, it's no longer
necessary to clear these caches when a term is updated; none of the data that's
changed on update (name, description, count, etc) is stored in the relationship
cache.
Fixes#36251.
Built from https://develop.svn.wordpress.org/trunk@37593
git-svn-id: http://core.svn.wordpress.org/trunk@37561 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Previously, objects containing all data about a term were stored in each
object's term cache. Besides being wasteful, this approach caused invalidation
issues, as when a modified term count required a flush for all objects
belonging to the term.
Backward compatibility is maintained for plugins that continue to put object
data directly into the `{$taxonomy}_relationships` cache bucket.
Fixes#36814.
Built from https://develop.svn.wordpress.org/trunk@37573
git-svn-id: http://core.svn.wordpress.org/trunk@37541 1a063a9b-81f0-0310-95a4-ce76da25c4cd
`WP_Term_Query` is modeled on existing query classes, such as those used
for comments and users. It provides a more consistent structure for generating
term queries, and should make it easier to add new functionality in the future.
Props flixos90, boonebgorges.
See #35381.
Built from https://develop.svn.wordpress.org/trunk@37572
git-svn-id: http://core.svn.wordpress.org/trunk@37540 1a063a9b-81f0-0310-95a4-ce76da25c4cd
[34529] introduced logic intended to prime the termmeta cache for certain
values of the `fields` parameter. There were a few bugs:
* The `all_with_object_id` param was misspelled.
* `term_id` was used instead of `ids`.
* The values being passed to `update_termmeta_cache()` in the case where `fields=ids` was not correct.
All of these would result in a failure to pre-fetch termmeta in some cases.
Props dlh.
Fixes#36932.
Built from https://develop.svn.wordpress.org/trunk@37567
git-svn-id: http://core.svn.wordpress.org/trunk@37535 1a063a9b-81f0-0310-95a4-ce76da25c4cd
Boone Gorges
Drew Jaynes
Sergey Biryukov
John Blackbourn
Dion Hulse
Helen Hou-Sandí
Gary Pendergast
Peter Wilson
Rachel Baker
Pascal Birchler
Dominik Schilling