There are two counters (ext.echo.unseen and ext.echo.unseen.click) that
are incremented in two different places. It's hard to see how they're
connected unless you know where to look.
Add comments to each of these counters pointing to the other one and
explaining what they're for.
Change-Id: I0699cba85bf797e4f7ef42cc6a7a996aa35510f0
The following sniffs are failing and were disabled:
* MediaWiki.Commenting.FunctionComment.MissingDocumentationPrivate
* MediaWiki.Commenting.FunctionComment.MissingReturn
* MediaWiki.Usage.ForbiddenFunctions.isset
* MediaWiki.Usage.PHPUnitDeprecatedMethods.AssertInternalTypeGeneric
Additional changes:
* Also sorted "composer fix" command to run phpcbf last.
Change-Id: I29416247ff3736799543926813beaf4afd569a6e
This seems strange, because markseen sounds like a write action, but it
writes to the seentime cache rather than the database. For multi-DC
support, we need writes to the seentime cache to happen in the local
data center, and the easiest way to do that is to make it a GET request
rather than a POST request.
It would be nice if marking as seen could be consolidated into the GET
request for fetching notifications, but I didn't do that because the
code for those fetches is pretty complicated, and some fetches (like
polling) should not mark as seen.
Bug: T222851
Change-Id: If4c504a9dc562b1d4e626e155fba8ebb5cdb0579
Allows users to opt out of receiving daily or weekly digests containing notifications
they have already marked as read on the web.
Bug: T169386
Change-Id: Ib47248678f88095492fb6896530be5a9f5bb43ca
Instead of returning false when no links in the user namespace are
found, return an empty array. They're both falsey, but this way we don't
break the type hint in generateMentionEvents(), which is the only thing
that getUserLinks() is used for anyway.
Bug: T239275
Change-Id: I93b320be07cfdae68c5e296b2caa62ea4fae5ff2
Now that seen times are TTL'ed, there is little sense/incentive to set
them in the cache to the default value, since we assume the default
value if they're not in cache (which is being set here). Therefore,
setting them only increases the cache size without any gains.
Bug: T222851
Change-Id: I78a0045e4a53653196114da81c2d60328e78c269
These entries weren't TTLed at all, and so we had millions of records
for inactive users that were being stored forever. If a user doesn't
view their notifications for a year, it's OK for their notifications
badge to go back to the unseen state until they click it.
(The fallback behavior on a cache miss is to act as if the seentime is
the UNIX epoch, which means that any unread notification they have is
more recent than it, and is considered unseen.)
Bug: T222851
Change-Id: I99230d2351b40751a3f2f5123c5f38693120259e
Defaults to MainStash, but can be configured to use a different cache
backend by setting $wgEchoSeenTimeCacheType to a string that keys into
$wgObjectCaches.
Bug: T222851
Change-Id: Ifb935cc8be4618f7794ee79a234fc66d5cc5728a
This code will be enabled when Iba1d7863171268066bf7597182c57a0a2041497f
relinquishes the responsibility for rendering the Echo notification badge
and wiring up of the related JS.
It makes 3 assumptions:
1) Minerva will expose a VERSION property on the skins.minerva.scripts module
to tell Echo it can begin control of the functionality
2) A new hook `SkinMinervaReplaceNotificationsBadge` will run on the server side
allowing Echo extension to render the Notifications badge in Minerva.
3) A new client side hook (echo.mobile) will fire whenever the Echo dialog is opened or
closed.
All code relating to Echo inside MobileFrontend and Minerva is
moved here.
CSS for the modules is kept in Minerva as skinStyles
This code remains dormant until Iba1d7863171268066bf7597182c57a0a2041497f lands.
It pre-registers a "to-be-created" hook SkinMinervaReplaceNotificationsBadge that
substitutes the Minerva badge.
It also watches the export value of skins.minerva.scripts for a VERSION value - when
this appears it will take the signal that it should manage the frontend code.
In the new system the mobile specific code is limited to the mobile version of
Minerva. The desktop version of Echo loads on Minerva desktop - presenting an
opportunity in future to consolidate both implementations to use the same component.
The mobile version of Vector and Timeless for example will load the mobile overlay
(with existing styling issues that we don't need to worry about right now given
we don't officially support skins other than Minerva as mobile)
Testers:
* Check require( 'ext.echo.mobile' )(); inside initMobile
inside ext.echo.init does not fire until
Iba1d7863171268066bf7597182c57a0a2041497f is checked out.
Depends-On: I1a66939d2b596094b419de40b370e79f09c85581
Bug: T221007
Change-Id: I09c27a084100b223662f84de6cbe01bebe1fe774
Update usages of the variable for the capitalization difference.
Depends-On: I4d289267991f1f9a8e0710ec6ee5a2131306c510
Change-Id: Ie49e898b33765aa0723501be8c0c30622e7cbd2f
As noted in Ic22075bb5e, UserNotificationGatewayTest relies on the presence of
globals. When Ic22075bb5e is merged, UserNotificationGatewayTest will fail in
CI. This patch injects the configuration object into UserNotificationGateway.
Change-Id: I9c15a588a76a41a3ebfb59dac7f0761f756008ff
New preference added so that user can set
* Displaying (n) total unread count in the title
* Displaying notification snippet for incoming notification
Bug: T229732
Change-Id: I35eb68dedf1e087b4668bfec404935f1244b3d0b
There's no real reason for these to be in MakeGlobalVariablesScript
because they don't really depend on the request context: they were being
omitted if the user was not logged in, but we can just check the user's
logged-in status in JS instead.
Bug: T221151
Change-Id: I2df6d7e061545d342bc3068dccd5ce2f6e85fe78
Remove wgEchoMaxNotificationCount and wgEchoPollForUpdates from the
startup module and put them in the ext.echo.init module.
Change-Id: I03f9a3953aa97ead1a29c13a992a02404a6d0b68
The hook (SpecialMuteModifyFormFields) is used to append
the option to mute/unmute echo notifications from a specified user.
Special:Mute handles posting and saving the fields, the only
requirement is that the field name is the same as the property
that wants to be modified, in this case 'echo-notifications-blacklist'
Bug: T220163
Depends-On: I2b3eee0802cb086091f35ecce13ae77a8e7d518d
Change-Id: I77b3ccfdce9b501eb8ecd58c0d7bbecb78029a7e
RevisionRecord::getComment() can return null. When it does, accessing
->text on it fails.
Bug: T226681
Change-Id: I069cd5d5bef51a3e2b8f4e7b50d478cf47f65e00
* MVC is used to change the count of notifications on header icons.
* The Widgets are loaded after 60secs or icon click, whichever occurs first.
Bug: T219222
Change-Id: I2d034a76d5c3abe55894dd1dd10e28713344c5bb
PHP doesn't care much about the name (in terms of case sensitivity)
but I think we should make sure the names of the method should be as
they're in their definition.
Change-Id: I6e38d8be64efaec4200471f2d3007275d7ddecec
Now that OOUI's "alerts" icon pack includes all the icons we need, we can use it and get rid of the badgeicons module.
Bug: T139779
Change-Id: I8218530ed2cdd2d81c1fc24509f36ea2b6742bd9
Follow-up Icfc8516 and I3d997f9 (migrating from deprecated Revision
to RevisionRecord)
When generating 'edit-user-talk' events, if we don't have the section-text,
we use the edit summary instead.
Revision->getComment() returns a string but RevisionRecord->getComment()
returns a CommentStoreComment. As a result of the migration, an instance
of CommentStoreComment was put into the event 'extra_data' and later
added as a plain text params during notification presentation.
This patch proposes to
1) Make sure to add the comment's TEXT to the event
2) Test that the 'section-text' is a string before using it
as a message parameter. This will cause faulty notifications to
render without a body instead of erroring out. We could write a
maintenance script to fix those notifications but I don't
think it's worth it.
Bug: T223741
Change-Id: Ie3a1c59a43b0981ec42fc77e7ebb0e8bd132dbe2
This reverts commit 13956265c1.
The Wikimedia CI jobs running phan were installing mediawiki development
dependencies which might overlap with the dependencies from the
extension. Ultimately phan failed with:
includes/RemoteSchema.php:9
PhanRedefinedInheritedInterface
\RemoteSchema inherits abstract Interface \JsonSerializable declared at
internal:0 which is also declared at
../../vendor/jakub-onderka/php-parallel-lint/src/JsonSerializable.php:4.
This may lead to confusing errors.
If7171a6d46473b0bb05e3adaeb40229e7881c2b6 adjusts the CI configuration
to no more install MediaWiki core development dependencies and solely
relies on mediawiki/vendor.git. That prevents the duplicate interface
definition.
Bug: T223397
Depends-On: If7171a6d46473b0bb05e3adaeb40229e7881c2b6
Change-Id: I794fbffe59104f138802bc6ded1a9d9a58ba9437
Follow up on Icfc85167a636bef95daab2.
Bug: T221163
Depends-On: I31d819f4453032da8deb117254b954fba058e0a3
Change-Id: I3d997f9e8d4bd1566ec6fd8fe2d33be5161458b3
To my knowledge in all the places I'm touching in this patch the new
code is functionally identical to the old one.
Change-Id: I0ffa96d2f9cb9bf932f68b689244051c96c17ad9
Class Revision is deprecated and in this patch, replaces
usage with appropriate classes; RevisionRecord, RevisionStore,
etc.
Bug: T221163
Change-Id: Icfc85167a636bef95daab236ab80113c1a3cf41b
When creating the various notification sources,
the server was indexing them with wfWikiID() but
the client was using wgDBname to find which one
is local and which are remote. On some wikis,
like TWN, wfWikiID() includes a db suffix so
the JS app on Special:Notifications is failing to
find the local source and errors.
Bug: T167336
Change-Id: Id60f723b615fb7db54a6f17b1c1be20dfe98e36c
When deleting echo_notification or echo_email_batch rows, also delete
the corresponding echo_event rows if no other echo_notification rows
or echo_email_batch rows refer to them.
Bug: T221262
Change-Id: I416ff107bd000a0cfac102408c993f8bec2d0286
$wgEchoSectionTransition was introduced when we moved some notification
types between sections (alert, message).
$wgEchoBundleTransition was introduced when we made bundles dynamic
and expandable.
Both flags have been OFF for years and are not needed anymore.
This patch removes all traces of them.
Bug: T140710
Change-Id: I16a5d54b09e71997f80208db6f4fbdb040d03ab1
EventMapper->fetchById() has retry on master built-in but it doesn't
kick in in a Job because the load balancer is a new instance and doesn't
know about previous writes.
This change makes the job always read from master to make
sure the event is found. It is going to write to master
right after anyway.
Bug: T204894
Change-Id: I9a2873234f1dd5416e6c2bedeb904880d1f79562
It was sort of checked in an implicit and roundabout way, in the sense
that the preferences page checks this and forces the relevant
preferences to on or off. But this was difficult to discover, and it's
not clear if this works correctly when defaults change.
Instead, explicitly check whether the requested notify type is available
for the event's category. Also rewrite the entire method to use
array_filter().
Change-Id: I9502a42fc705b2e56ef5edab433f1804f2924359
When pages are deleted or undeleted, their associated
Echo notifications also need to be marked as deleted
or undeleted. This is done by looking up the
echo_target_page table.
With this change, echo_event.event_page_id is also included
in the search. This way, all the events that don't use
target pages are also moderated when needed.
Bug: T217452
Change-Id: I277fca68ce088ab564e76ed9dfb2134ab3be4c4a
When the location on disk of an icon is known,
load its content and turn it into a data uri
for the <img> tag.
Otherwise (when it's only configured with a url),
use the url in the <img> tag.
Bonus: the rasterized icons were a bit blurry
and they are now crisp.
Bug: T55479
Change-Id: I8b5d7f09d8181f22927b8d4712505212ea1a10b1
This trait was using private fields from its host classes
($this->event, $this->language). It created a weird coupling
situation where the host class uses the trait and the trait
uses the host class. Effectively a circular dependency.
Also, phan is complaining about it in
I65ae6adc10941c05a2646e551b1baa829e4e8654
Change-Id: Ib2796b7ca62ecd7ece19583d7ca83e4252a5d878
Do this by default in EventPresentationModel so it automatically works
for all email subject messages.
Update messages to use this parameter. Also update email subject
messages that were trying to use parameters that didn't exist, and fix
message documentation that was completely wrong about what the
parameters were.
Also update translated messages to account for parameter shifts (in many
cases, these messages were using the wrong parameter or a nonexistent
one).
Bug: T219620
Change-Id: Ia420c4db13c95d91e1bfc4c7950942df5858379b
It is already called with a context in SpecialNotifications
Set context on SpecialNotificationsMarkRead and pass title to html form
to pass texts
Change-Id: I24eb85648580ac1a4df5e7587806bad57c141785
Also remove fetchNewestByUserBundleHash() because it's now unused,
and remove the echo_notification_user_hash_timestamp index which
existed specifically to support that function's DB query.
Bug: T143763
Change-Id: I74be8f156bc14d0e189d328953d17dc26cdb697b
This code improvements seeks to improve on code readability, consistency,
maintainability and efficiency.
Change-Id: I4f07886044e9a75824f9e7ddad039f3112b1c4a1
Incrementing user edit count was moved to a deferred update in
I0d6d7ddd91bbb21995142808248d162e05696d47
That causes the notification to lag behind, and the thank you notification is
sent for your first edit on the second, you 10th on the 11th, etc.
This change simply assumes one more edit than the current count unless it is
running from the CLI (this is needed for the test to continue to work).
Since the update job is mergeable, it is possible that a certain value will
be skipped and the notification for it never sent. See task for moe details.
Bug: T209541
Change-Id: Iea61b0f525be25f63f50582933a16a79a52e141f
No longer used in the new bundling system.
Also removes indexes that contain bundle_base.
Bug: T143763
Bug: T131415
Change-Id: Ibf94cdc471a11cb14995fee6a55af0d227b50aa5
By running updateEchoSchemaForSuppression.php as an updater.
The patch-*.sql files already existed, they were added
in 2013(!) by 34fbeaf8c but never applied.
Bug: T136427
Bug: T50059
Change-Id: Ied049681df4bab325f069c3a68cf704ee9a8f2c9
This fixes tests, because EchoHooks has "echo_event.*" which is invalid
with table prefixes, like done in unit tests
Added EchoEvent::selectFields() for future use and to replace more
SELECT *
Bug: T217487
Change-Id: I51cb46812431635d11780633dc7d807cd04f813d
This codebase already used the …::class feature a lot. This is more
about making the code consistent and easier to refactor in the future.
Change-Id: If5b2456b825aae753ed97445160ec191c18df8e3
* preg_match_all already returns the number of matches. We can just use
this number instead of counting it.
* Checking for a strlen() of 0 is a little tooo expressive, because we
don't really care about the actual length of the string.
Change-Id: I0537a7740e5d369b79364f24aecf71c4e8fa7db1
Since the correct transformation method is unclear, change them to plain message keys.
Bug: T211876
Change-Id: I055c089574e7dcf5d634da402290a2ad024126f9
Several links that link to mediawiki.org still use
"http" protocol which isn't a safe protocol. This commit
changes http links that link to mediawiki.org to use
"https" protocol instead.
Bug: T189687
Change-Id: Ie8a8fca4148181c9d1fe4379993aae3aaacf5ed9
Specify which notification types allow notifying the event agent in
$wgEchoNotifications, and stop specifying it in the event_extra data.
Putting 'notifyAgent' => true in event_extra will still work, but is
discouraged.
Change-Id: I4f558654ec23757dd4ecd6986eb3e9a5593f5386
For the algorithms in question this does not make any difference. Excess
elements are thrown away anyway. This is for performance.
Change-Id: I645e389b3f993bc8015fd729b9302aa25471f833
If all the code needs to know is if an array is empty or not, using
count() creates more confusion than anything.
Change-Id: I253308505fe5af5b5c56d85169789271967298ba
Note that the array of matches returned by preg_match changes when
PREG_OFFSET_CAPTURE is set. That's why the additional [0] (for the
match) and [1] (for the offset) are needed.
Bug: T203930
Change-Id: I89f9ea3bef49fe9128fd42805695982f012ecba9
Some Echo events include 'revid' in their extra info.
It can be used to check if a revision is minor in order
to respect the 'enotifminoredits' preference.
This patch ensures that it is not trying to call a function on
a null revision reference and it logs to debug to help
troubleshooting.
Also encapsulate the "is minor" check to a private function
to (hopefully) make the relationship with the preference
more clear.
Bug: T204795
Change-Id: I28a4c54f610dccc1356f6af0de9c2623d7bf94f0
See, this last part of the compiled regular expression is wrapped in
an (…)*, which means it is entirely optional. It does not make any
difference if this part is found or not. The compiled regular
expression matches with or without any of these "line ending"
fragments being present.
I can not really figure out what the intention of this was. A line
ending anchor ($) is not missing – I'm pretty sure about this.
Otherwise it could not detect signatures that are wrapped in more
than a single HTML tag, for example.
Instead of fixing it I decided to remove it. The tests should show
this code was not needed.
The motivation for this patch is to improve performance. This part of
the regular expression is quite heavy and can cause a lot of
backtracking for literally zero benefit.
Bug: T203930
Bug: T204291
Change-Id: Ia5323b401b947edeb7094d7eec131ba6c80edf70
\h matches only horizontal whitespace, but no newlines. This is what
we want in all these cases, because nothing of this (headlines,
signatures, timestamps) is even allowed to span multiple lines in
wikitext. The tests should show this still succeeds.
The idea is to make these regular expressions more strict so they
don't run in so much expensive backtracking.
Bug: T203930
Bug: T204291
Change-Id: I805f8cb082edcd26713ef41d3ae5b61194c131e5
In double quoted strings PHP tries to understand all kinds of escape
sequences, but \A is not one of them. Such sequences are left untouched,
including the backslash.
In single quoted strings, there are no escape sequences. All are left
untouched, which is what we want in case of a regular expression.
TL;DR: The resulting string is the same in both cases. I'm touching this
because my IDE shows a warning about the unknown \A escape sequence.
It must be either turned into "\\A" or '\A'.
Change-Id: Ie1e84c67c344faf77bc86a0b28dc82d31c3a7dbe
false ?? null evaluates to false, so for non-write modules we were
passing false as the token type instead of null, which breaks.
Bug: T204758
Change-Id: Ief25150ce8f4b4b64a224f97f3fd528883b2f326
Disable some errors related to different members of an array
having different taints, and phan-taint-check conflating them.
Bug: T202383
Change-Id: Ic6c2c5bb7c6092d581e646358d836f55d5cf3222
Add a markasreadwiki parameter, and use cross-wiki API proxying to mark
the specified notification(s) as read on the originating wiki.
This allows notifications to be marked as read when the primary link is
followed, even if the primary link points to a different wiki.
Bug: T179765
Change-Id: Id7e1e11997173e1578e33cd189dc0f93a5e4ba63
Add support for POST requests and tokens to EchoForeignRequest
and ApiCrossWiki, and add the ApiCrossWiki trait to ApiEchoMarkRead.
Change-Id: Idadaacd0d0c4a957bf2499049fc105a60c73bc52
That way we'll be able to mix it into non-query modules as well.
Unfortunately, PHP traits don't let overridden methods call their
original versions, which had to be worked around in a few places:
- $this->foreignNotifications can't be initialized in the constructor
any more, so it's now lazy-created through $this->getForeignNotifications()
- Adding the 'wikis' parameter to getAllowedParams() now happens by
calling getCrossWikiParams() rather than calling the parent
- Overriding getForeignQueryParams() can't call the parent anymore, so
instead we just inline it
Change-Id: I415e6d921819fc1f7869c7d2f8bb62830a84c2a1
This is what MediaWiki itself does. It's not great, and isn't any more
translatable than a config setting, but consistency isn't valueless. In
future, with T202326 we may wish to vary this, but for now this is
simpler.
Change-Id: I76fca8ee255c65ab9b7e988d44de0d0fbd3c84b7
Instead just use the system message `emailsender` like we do for all other
e-mails, which is thus translatable.
Change-Id: I7e58a3a4e224d551c6fae4a76fcba19fe838d3e9
These were broken because e8632ab0f6 in MW core stopped passing a
triggering user to LinksUpdate.
This commit takes the user from the Revision object instead. In weird
cases that might be different from what LinksUpdate says, but
page-linked notifications don't use the agent in their rendering anyway.
Also remove the code that refused to route events to a page creator if
they didn't have an agent. It's perfectly legitimate for events not to
have an agent, that shouldn't preclude them from using non-agent-related
locators.
Bug: T200119
Change-Id: Ia31131b1d1b2640d962ab7f3e573599c43ae50d4
Followup I85452d0f0afe974d26a575e000f6ae2ceeddf06c
* initialize "$success = true" before using it in
a loop to keep track of the overall success of
all batches.
* Add check for readOnly db in markUnRead() and
markAllRead(), like it was done in markRead()
Bug: T202672
Change-Id: Ifdfa93059268d5b02ba3e0e885661ce593845791
php7 gives:
count(): Parameter must be an array or an object that implements
Countable
null or [] are both evaluate to false, so just remove the count
Change-Id: Id92dbd48f308d1e9dffa086699e0e944744aeeaa
This makes clear which escaping should be used.
In all cases it needs Message::text, because there are used inside
Html::element. There is no visual different.
Change-Id: I17474a7d5f057321e8c759d4bf94c8234c7a89c7
Instead just use $wgPasswordSender explicitly. This also allows us to drop the
extension registration callback function, which is a performance improvement.
Bug: T200390
Change-Id: I08d2f040c5ad8feb395a2e8e176f91636efe1d3d
* Reduce responsibility of resolve() methods to only supplying
the resolves values.
Moved logic for populating the cache and clearing the queue
to the base class, and made 'lookups' private.
* The second parameter to LocalCache::add() is unused, and never passed.
Removed to avoid confusion.
* The getTargets() method is unused. Removed.
* The getLookups() method is unused. Removed.
* The internal 'lookups' member was being used both for its keys and its
values, but never at the same time. This seemed risky, especially in
EchoRevisionLocalCache::resolve() where the associative array was passed
directly to the 'where' clause of IDatabase::select(), which shouldn't
espect keys when creating the 'IN' clause.
Using only values would keep value types flexible, but would require
use of the less efficient in_array().
Keeping both keys and values and calling array_values() would work.
Using only keys also works and is simpler, so long only ints are used.
* The tests were swapping 'targets' MapCacheLRU with a HashBagOStuff.
Following-up 4939bff7, this was forgotten, but works because the two
called methods (get and set) exist in both, but still seems odd.
Fixed by using TestingAccessWrapper to act on the existing object
instead of swapping it out.
* Improved tests by asserting more of the observed behaviour and impact.
Change-Id: I530eeac8bf3b407b8c633e0e20c7d35cc49f7a9f
This fixes some issues I found while updating this code base, e.g.
this removes types a method really does not return.
Change-Id: I19457e7bf88945eec958bf53e0b76a7585715a45
