Why:
* Protected variables were introduced to support temporary accounts
so that temporary users could be filtered based on their IP address.
* Filters that use protected variables are protected in order to
preserve privacy. This can't be undone.
* It is mistakenly possible to protect a filter that does not use
protected variables (T378553). We need a mechanism to fix these
mistakes.
What:
* Introduce a maintenance script that takes a filter ID and, if the
filter is protected, sets it to unprotected while maintaining any
other existing privacy levels.
Bug: T378551
Change-Id: I4dfe3970221397d5be5ea0697490d8c8e3726adf
Why:
* Protected variables were introduced to support temporary accounts
so that temporary users could be filtered based on their IP address.
* Filters that use protected variables are protected in order to
preserve privacy. This can't be undone.
* The current workflow for protecting filters is to have a 'protect'
checkbox that can be checked for any filter, even one without
protected variables, to the discretion of the editor. This has
led to mistakes (see T37765).
What:
* Do not show the 'protect' checkbox on the form by default. Instead,
show it when saving a filter with protected variables, after form
submission. The user must check it to save the filter.
* Still show the checkbox in a disabled state with a warning message
if a filter is already protected, so that the editor can easily see
it is protected.
Bug: T377765
Change-Id: Ie5c94ac1399860ccdca4482508dd37ff07309764
Why:
* For improvements to the protected filters workflow (T377765),
a checkbox needs to be displayed conditionally on the status
of a previously submitted form.
What:
* Pass a status object into ::buildFilterEditor, and build the
HTML for any warnings or errors in that function, instead of
first building the HTML and passing it.
* This also allows more than one error and/or warning to be
shown if present, though this may not happen in practice yet.
Change-Id: I37da49711e7fc192856f013cd931c05379f8e8c6
Remove the use statement for MockAuthorityTrait, which is unused.
We can't use it because SpecialAbuseFilter checks permissions on
the user, not the authority.
Change-Id: I2baca5753dd9e21aa90b6856c5850d1100eda4d5
Why:
* The ProtectedVarsAccessLogger::logViewProtectedVariableValue
method creates a debounced log entry for accessing protected
variable values.
* This log can be created when a user calls the 'abuselog' query
API. This API query is made using a GET request.
* Because the ProtectedVarsAccessLogger creates the log
when it is called, the TransactionProfiler raises a warning
about writing to the primary DB on a GET request.
* We should address this warning by creating the log entry on
POSTSEND, in a similar way to how the CheckUser extension
creates CheckUser log entries.
What:
* Wrap the call to ProtectedVarsAccessLogger::log in
::logViewProtectedVariableValue with a DeferredUpdate callback
that is run POSTSEND.
** As part of this, the expectations that the code only uses the
replica DBs are silenced inside the DeferredUpdate as the
PostSend-GET expectations do not usually allow for writes.
The only other method would to be create the log via a job,
but we want the log creation to occur quickly and reliably.
Bug: T379083
Change-Id: If14e65b27b0bdd4fd0b99319024ffa281bf44656
Implicitly marking parameter $... as nullable is deprecated in PHP
8.4. The explicit nullable type must be used instead.
Bug: T376276
Change-Id: I303342cf1a002d5f0afc77ce147ce9453ea5282e
Why:
- For account creations and account autocreations, the user_name
property is deliberately unset, to avoid displaying the IP address of
an unregistered user. Instead, `accountname` is set with the newly
created account name
- For logging that someone has seen a protected variable value, we need
to record the username that was seen
What:
- Use `accountname` as a fallback in case `user_name` is not set, when
logging protected variable access
- Update tests to cover this case.
Bug: T376885
Change-Id: I688a3529fac0ad8455977a0cfdb950f0105f550d
Changes to the use statements done automatically via script
Addition of missing use statement done manually
Change-Id: If80031678a474157e4cc78a3d3621dab53aded67
Like CheckUser, AbuseFilter should also log when specific protected
logs are viewed.
- Add support for debouncing logs to reduce log spam
- Log when AbuseFilterViewExamine with protected variables available
is accessed
- Log when SpecialAbuseLog with protected variables available is
accessed
- Log when QueryAbuseLog with protected variables available is accessed
Bug: T365743
Change-Id: If31a71ea5c7e2dd7c5d26ad37dc474787a7d5b1a
CVE-2024-PENDING
Why:
* The 'abusefiltercheckmatch' API allows callers to match
arbitary filter conditions against existing AbuseFilter logs
* The API does not check if the performer has the ability to
see the log details for the given filter, so can allow a user
to bypass hidden and protected visibility settings.
What:
* Call AbuseFilterPermissionManager::canSeeLogDetailsForFilter
before attempting to match a filter against a given AbuseFilter
log.
* Add a test to verify that this security fix works.
Bug: T372998
Change-Id: I4a2467dc4e0d1f8401d5428a89c7f6d6ebcdfa70
- Fix an issue where if a user didn't have view permissions they could
get the preference check error (a preference they wouldn't have) on
SpecialAbuseLog
- Fix an issue where the `change-access` hadn't been updated to the used
disabled/enabled log types
- Fix an issue where a ProtectedVarsAccessLoggerTest test wasn't
correctly using the data provider data
- Improve naming since ProtectedVarsAccessLogger exists in its own test
file instead of being a subset of tests on AbuseLoggerTest
Bug: T371798
Change-Id: I53f22855e63d9e1339361a5c9ee7886e0f74714a
This change is needed to unblock a change in core to the markup.
(I0195d4b0f790f6595cc626a6db96b4fc6380a0f4). The current markup
in core is loading additional CSS styles to support legacy
markup.
Bug: T360668
Change-Id: I4bd1a8a9d4eda1b3e89d067d6671d3f8bad4f584
Write logs related to temporary accounts to CheckUser if the extension
is available so that logs are topically centralized.
Bug: T373525
Depends-On: I35d50df7cd6754e29d964cc716fb3c42406272df
Change-Id: Ic95f211f4db7ce6dc2d769d2f3af206f4a3935e4
Similar to how CheckUser logs access to IP information about temporary
accounts, AbuseFilter needs to log whenever protected variables are
accessed.
- Implement ProtectedVarsAccessLogger which handles access logging
- Log whenever a user changes their ability to access protected
variables via Special:Preferences
Bug: T371798
Change-Id: Ic7024d9c5f369eb33c4198a59638de9a1d58b04b
Users need to enable a preference before gaining access to the IPs
from `user_unnamed_ip`, a protected variable.
- Add a preference that the user can check to toggle their access
- Check for the preference and the view right for logs that reveal
protected variables on:
+ AbuseFilterViewExamine
+ SpecialAbuseLog
+ QueryAbuseLog
Bug: T371798
Change-Id: I5363380d999118982b216585ea73ee4274a6eac1
Small performance benefit by just one db call instead of multiple
Most test cases only use one filter, but some 2 to 4
Change-Id: I498c447e3873d2138e21541467115c9a67bb909e
AF rules don't support associative arrays, so the named capturing groups are provided in the array only by their numeric keys.
Bug: T374294
Change-Id: I53b39917e6677f3a5b8f68bcf0faebf48668ea27
Why:
* SearchFilters.php allows the caller to search by a regex that
is applied to the pattern.
* This script can be expanded to allow callers to specify what
consequence should be associated with the filters that are
outputted.
What:
* Add a 'consequence' option to the SearchFilters.php maintenance
script, which is applied through a LIKE query on the
af_actions column.
** This can be specified with or without the pattern option.
** Instead of making pattern required, the script now requires
that one of consequence or pattern is provided.
* Expand the tests for the script for this new code, along with
using the new ::expectCallToFatalError method to be able to
test previously untestable code.
Bug: T373148
Change-Id: I1b507d8f9dc1f4cf91ee4f83ccde745eb6d46d6d
A filter using a protected variable can be loaded via filter id
using testing tools even though the user might not have the right
to view protected variables. This can potentially leak PII and as
such, testing tools should check for the right before allowing
protected filters to be seen.
- Unload a filter asap if it uses protected variables and the
requestor doesn't have viewing rights. This:
+ disallows loading of existing protected filters on page load
+ disallows testing against rules that use protected variables
+ disallows subsequent requests for protected filters (via API)
There is a known bug (see T369620) where no user feedback is
provided if an API request for a filter returns no result (typically
when no filter matches the requested id). This commit adds another
pathway to that bug (the filter exists but is protected and not
returned by the API) but does not update this UI/UX.
Bug: T364834
Change-Id: I6a572790edd743596d70c9c4a2ee52b4561e25f3
Why:
- Ie13181b78b8e2903c6cc0f0f778689bcc8b8ce2e modifies the status message
returned
What:
- Loosen the check for the status error message such that 'captcha-edit'
and 'captcha-edit-fail' are both valid; we can revert this after
Ie13181b78b8e2903c6cc0f0f778689bcc8b8ce2e is merged
Change-Id: I5a0698d84932a474800a68dba9b76b3433b19290
Why:
- These checks are no longer needed, now that Idc47bda has been merged
What:
- Remove the `method_exists()` checks
Change-Id: I6e428df6b6e036146ae4cc57374cde8810d3f5f7
Search is restricted to users with the right to view private variables
but not necessarily the right to view protected variables. Users who
don't have the right to view protected variables shouldn't be able to
search against protected variables, as this might leak the PII.
- Filter out filters using protected variables in search results
if the user doesn't have the right to view protected variables
Bug: T367390
Change-Id: I7412112c9cc676f29d706b116b779bc17183a952
Alternative approach to fixing the regression proposed by
Daimona in I78d3a2cd7bada962d7ef9b0f2c39d898bf8987ce.
Bug: T368203
Change-Id: I637367c3b3850f7988d890379fef7f4753159953
- Rename `$hidden` to `$privacyLevel` in Flags::__construct for
consistency with other places.
- Rename `shouldProtectFilter` and simplify its return value to always
be an array, since that's how it's currently used. Rename a variable
that is assigned the return value of this method.
- Add a missing message key to a list of dynamic message keys.
- Rename a property from 'hidden' to 'privacy' in FilterStoreTest for
consistency. Add a test for removing the protected flag.
- Update old comment referencing `filterHidden`; the method was removed
in I40b8c8452d9df.
- Use ISQLPlatform::bitAnd() instead of manual SQL in
AbuseFilterHistoryPager.
- Update mysterious reference to "formatRow" in SpecialAbuseLog.
- Update other references to the very same method in two other places,
this time credited as "SpecialAbuseLog".
- Add type hints to a few methods; this not only helps with type safety,
but it also allows PHPUnit to automatically use the proper type in
mocks.
Change-Id: Ib0167d993b761271c1e5311808435a616b6576fe
`canSeeLogDetails` should also be checked when a filter is protected, as
it is the base right for being able to see abuselog entries. With this
in mind, check that immediately at the beginning of the method, instead
of repeating calls. Also merge the conditionals, and return early when a
permission check fails.
Move a test up so that it comes immediately after its data provider, and
add test cases for a few combinations of rights.
Change-Id: Ic3cf58f43803bef8bf2d65566434baff145b3fd5
This test globally set `setForceShowCaptcha` to true, which caused
problems for following tests.
Bug: T368705
Change-Id: I5077e4b874c1bf1c6b68895349af0c9ecd4094ed
Why:
- The previous attempt to integrate AbuseFilter with ConfirmEdit set
a flag on the request object
(I110a5f5321649dcf85993a0c209ab70b9886057c) didn't work in WMF
production because in WMF, we load ConfirmEdit first, followed by
AbuseFilter. Therefore any flag set in an AbuseFilter hook is ignored
by ConfirmEdit
What:
- Remove implementation of ConfirmEditTriggersCaptchaHook, as this does
not work when AbuseFilter is loaded after ConfirmEdit.
- Repurpose onConfirmEditTriggersCaptcha to handle non-edit actions only
- Implement the EditFilterMergedContent hook and call SimpleCaptcha's
public confirmEditMerged method if CaptchaConsequence has specified
that a CAPTCHA should be displayed, and if the CAPTCHA has not already
been solved
Soft-Depends-On: Idc47bdae8007da938f31e1c0f33e9be4813f41d7
Bug: T20110
Change-Id: I7dd3a7c41606dcf5123518c2d3d0f4355f5edfd3
