This is the most minimal patch possible to fix the regression. More
can be done in later patches.
Bug: T359290
Change-Id: I9351bdbc2f985e72fb253e424cd7523b73820dcd
MultiGadgetRepo is a wrapper around two or more GadgetRepos so that
they can be used at the same, facilitating migration from one repo to
another.
If a gadget appears in both repos, the definition in the first repo
takes precedence, and a warning is shown on Special:Gadgets.
This can be enabled to wrap GadgetDefinitionNamespaceRepo and
MediaWikiGadgetsDefinitionRepo by setting $wgGadgetsRepo to
'json+definition'. In this configuration, once a new JSON definition
exists for the same name, it is used instead of the old one, and the
old one can then safely be removed at a later time in the safe
knowledge that it is no longer used.
Adapted from If3cc5e22e9812d0fd1a9e8e269ea74a7f667dadd
Bug: T140323
Co-authored-by: Kunal Mehta <legoktm@debian.org>
Change-Id: Ibad53629e63ec8713d7a3b84a19838b94600588e
== What ==
* Remove the empty Gadget and Gadget_definition namespaces.
* Remove the "gadgets-definition-edit" user right.
* Remove need for custom namespace permissions that previously
had to extend editsitejs to apply to NS_GADGET.
== Why ==
Simplify the (unused) "GadgetDefinitionNamespaceRepo" backend for
Gadgets 2.0 by making it less radically different from the status quo.
The experimental 2.0 branch will now make use of the "gadget definition"
content model via "MediaWiki:Gadgets/<id>.json" pages, instead of
through a dedicated namespace.
When I first worked the Gadgets 2.0 branch, content models
were not a thing in MediaWiki, and interface-admin wasn't a thing yet
either. Now that we have per-page permissions and per-page content
models, we don't really need a separate namespace.
This follows the principle of least surprise, and fits well with other
interface admin and site configuration tools such as:
- Citoid, MediaWiki:Citoid-template-type-map.json,
- VisualEditor, MediaWiki:Visualeditor-template-tools-definition.json,
- AbuseFilter, MediaWiki:BlockedExternalDomains.json,
- the upcoming "Community Config" initiative.
If/when we develop the SpecialPage GUI for editing gadget definitions,
this can save its data to these pages the same as it would in
any other namespace. Similar to how Special:BlockedExternalDomains
operates on MediaWiki:BlockedExternalDomains.json.
See also bf1d6b3e93 (I6ffd5e9467), which recently removed the
gadgets-edit user right in favour of the editsite{css,js,json} rights.
Change-Id: I5b04ab251552e839087d0a8a6923d205adc7f771
This makes way for "MediaWiki:Gadgets/<id>.json" to be used free of
any confusion or misinterpretation about the subpage being a
translation of "MediaWiki:Gadgets".
Change-Id: Icb68ce6c26a83d7201a97a5406f8752b3c5b0855
This already returns safe Message objects, which allow Special:Gadgets
to localise it accordingly.
Avoid formatting the list ahead of time with a hardcoded user language,
and instead use Message::listParam() so that the caller controls
the Language object. This makes it safe to use in jobs, CLI, tests,
load.php and elsewhere instead of previously where RequestContext would
cause an exception or use something the caller doesn't control.
Bug: T298334
Change-Id: I77dd3c7fef86b4c8c41837bfb276b019f30baeb0
* Validate that pages have .js, .css or .json suffixes.
* Validate that namespaces are integers, disallowing doubles.
* Validate that type is either general or styles.
* Validate that targets are either desktop or mobile.
* Fix validation of skins: stop allowing `true` as it isn't actually
allowed - it causes an exception in Gadget::isSkinSupported().
Bug: T298334
Change-Id: Ic59c20172f496d281acd59419f5c760ad29ce317
The column user_properties.up_value is a text based type in all rdbms.
postgres does not autoconvert strings to numbers, so SUM( up_value )
gives an error when running on postgres.
The use of SUM( up_value ) is also problematic on mariadb/sqlite, as
up_value can be set to any number via the Action API's options module.
This value goes into the sum and produce a wrong result. Using a
non-number value allows to remove the value from the sum and that also
produce a wrong result.
Change the query to exclude the rows which disables a gadget and just
count the "active" ones. Typically there are only rows to disable a
gadget, when the gadget is marked as default, but default gadget are
displayed with the message 'gadgetusage-default' and does not needs to
be counted. In case a gadget was default in the past, there could always
such rows, so be safe and exclude them. That means it is safe to only
count for "active" gadget option usage.
This also allows to remove the up_value condition on the join to
querycachetwo as only rows to enable a gadget are now selected and get
joined to find the active users.
For the join condition use an assoc array, as for database condition the
database class is quoting the value of that array, if that is a string.
In postgres the " is used to quote columns or tables, but not values,
that gives a query error on postgres about the condition.
fixed when using the array style and not "raw" sql.
Define the condition only once as it should be the same for both
queries. Use the new IDatabase::expr function to write that sql piece.
Change-Id: I8dfc3fd5adc4c4bdabceaab20c4b37ffd48e6bee
This requires 1.42 for some new names
Changes to the use statements done automatically via script
Addition of missing use statement done manually
Change-Id: Iacbea33299995c537a7ef77b524614ad02c6a559
0c61a5a468 subtly affected the
gadget definition parsing, causing issues with definitions containing
HTML comments or certain kinds of whitespace characters. They would get
saved to the pages array with a trailing whitespace but
Gadget::getScripts() would only return pages strictly ending in `.js`.
Trimming the title before in MediaWikiGadgetsDefinitionRepo fixes the
issue.
Bug: T354385
Change-Id: Iacee432756006060217981a534434fd455285793
Remove the test that expects ES6 to fail by default, instead assert
that invalid syntax fails (as it should). Also add various other
test cases, including a commented-out test for valid ES6 by default,
to be enabled after Ie309e761f8b20640f7c0 in MediaWiki core.
Change-Id: If45856c563518255189687e3b1f620973349fb27
As this repo is not used anywhere in production, there is no backward
compatibility. Any scripts, styles, datas specified in existing
definition pages will get ignored.
Gadget::getJSON() no longer checks isPackaged() – that is now left
for callers - only relevant one is in GadgetRLModule. That was done
because validationWarnings() method needs the JSON page list even for
non-package gadgets.
Bug: T353174
Change-Id: Ic820bc4632981369c72f822fb8b9b7195930eaa6
A little efficiency to make getting from Special:GadgetUsage
to the relevant gadget files easier.
Bug: T344255
Change-Id: I0e26d10992bc3f8073bcdc8dbac86726ee6f7c36
Remove DefinitionDeletionUpdate and DefinitionSecondaryDataUpdate –
both call $repo->handlePageUpdate() which is already being called from
onPageSaveComplete/onPageDeleteComplete hooks.
The chain that made this redundant:
1. Ibe2e26d123, change hooks from hardcoded to GadgetRepo singleton.
2. Ieccc1cae8c, migrate from create-only onPageContentInsertComplete
hook to create-and-edit onPageSaveComplete hook.
3. Ib27fd34fbf, further consolidate handling of create and edit
events with handlePageUpdate, and remove any hook-level
conditionals. At this point, the DateUpdate classes became redundant.
Bug: T31272
Change-Id: I20c2759b219c80571237a73e8422f3128047eb87
* Separate out tests for MediaWikiGadgetsDefinitionRepo and
Hooks::onGetPreferences. Removed hack in fetchStructuredList() for
overriding page content for tests.
* Add tests for ES6 gadgets. For this, ResourceLoaderModuleTest is
converted back to an integration test.
* Removed createGadgetDefinitionPage() test helper method. Now using
MediaWikiIntegrationTestCase::editPage().
* Fixed GDNamespaceRepoTest failure when GDNamespaceRepo is the active
repo, by clearing WAN process cache.
Change-Id: I26b84576a91f6cb0ebae64c5fc1408666d767911
* Add tests for onBeforePageDisplay hook. A simple gadget repo
implementation, StaticGadgetRepo, is introduced for this.
* Add integrations tests for various gadget load conditions.
* Add test for onUserGetDefaultOptions hook.
* Add tests for GadgetDefinitionNamespaceRepo.
* Convert ResourceLoaderModuleTest to a unit test.
Change-Id: I275380c2bfcaa44770b3946a0a468eaaabef70c0
With GadgetDefinitionNamespaceRepo, there should be separate rights for
editing gadget json as opposed to editing gadget css/js, so that admins
can be granted the former, while the latter remains restricted to
int-admins.
Reusing the MW core rights for site CSS/JS/JSON seems to make the most
sense here. Protection is now applied via GetUserPermissionsErrors hook
rather than as namespace protection.
For editing any non-CSS/JS/JSON pages in the namespace, editsitejs right
is required.
Bug: T298834
Change-Id: I6ffd5e9467774f1e79ccdce8b6b4739f07be2da8
… e.g. using modern syntax, or avoiding unnecessarily complex
language features. For example, we don't need to call count() when
all we care about is the boolean "is empty/not empty" information.
Change-Id: I13ae802f64627a79b29d1e57ad71486cb2fb977f
This can't do anything for two reasons:
* There are no spaces in the string after the str_replace. There is
nothing to trim.
* The regular expression above guarantees the string starts and ends
with alphanumeric characters. There can't be anything to trim, not
even underscores.
Change-Id: Iea40aafaa2457cdee9d05f105978627d38d9d5b8
* A fixed license header for the file.
* Move class descriptions, doc tags, author/copyright to the
class block. Remove MW-core specific "ingroup" tags such as
"SpecialPage" which has no meaning in an extension and are not
used by Doxygen.
Ref https://gerrit.wikimedia.org/r/q/message:ingroup
Change-Id: I13d9b0ff7a3150180196c9fa58f8a321b14edd22
