This patch introduces a namespace declaration for the
MediaWiki\Json to FormatJson and establishes a class
alias marked as deprecated since version 1.43.
Bug: T353458
Change-Id: I5e1311e4eb7a878a7db319b725ae262f40671c32
Use
$this->getServiceContainer()
instead of
MediaWikiServices::getInstance()
in tests where possible.
Change-Id: I798b2941f37a43b9073072935b54c3ea0cfe70dd
To implement source maps, we want FileModule::getScript() to return
an array with path information, but that would break subclasses of
FileModule which concatenate to the return value of parent::getScript().
So allow scripts to be generated by a callback, eliminating the need for
concatenation in subclasses.
* Factor out most of the loop body of expandPackageFiles() into
expandFileInfo(). Add 'name' to the return value. Ensure 'filePath'
is always a FilePath object since we need that to safely return it
to ResourceLoader later. Document the return value.
* Make sure the base path is always set in FilePath objects returned by
expandFileInfo().
* Factor out the loop body of the final stage of file info expansion
into readFileInfo(). Retain filePath, do not unset it.
* Assert that $fileInfo['content'] is definitely set.
* Convert array_map() in getDefinitionSummary() to a loop.
* Migrate LanguageDataModule.
Bug: T47514
Change-Id: I97d61b5793159cea365740e0563f7b733e0f16de
Move ResourceLoader classes to their own namespace. Strip the
"ResourceLoader" prefix from all except ResourceLoader itself.
Move the tests by analogy.
I used a namespace alias "RL" in some callers since RL\Module is less
ambiguous at the call site than just "Module".
I did not address DependencyStore which continues to have a non-standard
location and namespace.
Revert of a241d83e0a.
Bug: T308718
Change-Id: Id08a220e1d6085e2b33f3f6c9d0e3935a4204659
This reverts commit e08ea8ccb9.
Reason for revert: Breaks Phan in extensions, and as far as I’m aware,
this change isn’t urgently needed for anything, so the simplest fix is
to revert it again for now. After PHP 7.4 it should be safer to try this
again (we hopefully won’t need the two “hack” classes by then).
Bug: T308443
Change-Id: Iff3318cbf97a67f821f78e60da62a583f63e389e
Move ResourceLoader classes to their own namespace. Strip the
"ResourceLoader" prefix from all except ResourceLoader and
ResourceLoaderContext.
Move the tests by analogy.
I used a namespace alias "RL" in some callers since RL\Module is less
ambiguous at the call site than just "Module".
I did not address DependencyStore which continues to have a non-standard
location and namespace.
Change-Id: I92998ae6a82e0b935c13e02a183e7c324fa410a3
This test was an awful, awful, mess. (And I take full responsibility.)
Changing the global user language mid-way into execution is in no way
supported by ResourceLoader and this test was going through great
lengths to fool mw.loader about what's really going on.
Basically, all we're doing is get a list of instructions on what tests
to run, get comparison values based how the PHP side proceses these
(for parity comparison), and then run the assertions.
The mw.language singleton already has support for multiple languages,
and mediawiki.jqueryMsg already supports injecting data and constructing
new instances for each test case. Make use of that :)
Instead of calling ResourceLoaderLanguageDataModule::getData by
trying to hot load the same module repeatedly from load.php,
just export this data using 'packageFiles'.
The mediawiki.jqueryMsg QUnit suite previously took 4s to run locally
and now only 0.1s (145ms). This is not only significant for this
particular module, but also for QUnit in general as Headless Chrome
only took about 7s to run all of MediaWiki core's test suites prior
to this change, which is now down to ~3s. (MacBook Pro)
For WMF CI:
* Before (master commit):
- mediawiki.jqueryMsg.test: ~2.1s (2135ms)
- MediaWiki core total: ~4.8s
* After (this patch):
- mediawiki.jqueryMsg.test: ~0.015s (15ms)
- MediaWiki core total: ~3.6s
Bug: T250045
Bug: T225730
Change-Id: I5f1067feb0a43d63bfc5e7fff5110285a5e433c8
This class is marked @private, but nonetheless, there is at least
one use in MobileFrontend, so keeping a back-compat alias for now.
Also rename the HtmlEmitter utility class (also private, not used
anywhere outside this file), and fix its broken documentation by
repeating the @class and @private tags on its own constructor.
It previously had all its methods indexed by JSDuck as part of the
previous class in that file (Parser), which created doc pages
with examples that don't work.
Change-Id: I02d851d9b6eac89f1a2b85b438b982bd055bedee
Follows-up c0fb8a8836, I890e6e49b.
* Disable 'qunit' env in general source code. And re-declare
locally in the few src files that use it properly.
* Create separate eslint config for tests/qunit with various
rules disabled (e.g. valid-jsdoc and es3-keywords).
Change-Id: I37ccec2019de55edfee92697eb80478df7cb6220
QUnit used to have bad state management (a few years ago) at which point it
became useful to verify the number of assertions in case an asynchronous
failure happened, as it would likely go unnoticed.
* Errors outside testStart/testEnd weren't caught.
QUnit now monitors window.onerror.
* Assertions could be attributed to the wrong test.
QUnit no longer does this since the assert object is associated with
the current test through lexical scope.
* assert.async()/done() replaced global semaphore (QUnit.start).
* A test could forget to be marked as async and make no assertions.
QUnit now marks a test as failed if it makes 0 assertions.
QUnit also has built-in async tracking for promises.
If a test is not reaching all assertions for some reason, this
will cause an error of some sort that is tracked. If in some
specific scenario this isn't the case, assert.expect() can still
be called (e.g. when expecting 0 assertions), but it'd be worthwhile
to file an upstream bug report in that case.
Follows-up 7c363752, which removed 'QUnit.config.requireExpects' from
our test configuration.
Follows-up c4c7007de6 and various other commits that already removed
the test counts from a subset tests. This commit removes the remainder.
Change-Id: Ie58396ba9c83d27220508481cb97c0fa74487756
* jsonlint now includes docs/, includes/ (api and installer i18n),
maintenance/, and tests/.
539 files -> 864 files.
- Continue to exclude JSDuck artefacts in docs/js/.
- Continue to exclude vendor/.
* jshint now includes mw-config/ and maintenance/.
177 files -> 179 files.
* jscs now includes everything jshint includes.
172 files -> 179 files.
- The -skip.js files no longer need excluding.
Use the native exclude syntax for jshint and jscs so that other
software and services with JSHint support use these as well.
Change-Id: Idebf30275f9c93483069367f923ed290c38e0b26
* Set up Grunt via package.json (run `npm install` in mediawiki-core)
* Add grunt task for node-jscs (NEW)
This is a style checker (as opposed to jshint, which is for
code quality). There are a few small style-related things that
JSHint can check (camelcase, onevar etc.) but those are being
deprecated in JSHint v3, people should use more sophisticated
tools like node-jscs for this instead. As such this commit
removes moves of those options from our jshint configuration.
See: http://jshint.com/blog/jshint-3-plans/
* Add grunt task for jshint
This will use the same jshint configuration as we use on
Jenkins but makes it easier to run locally from the command
line by being part of the same `$ grunt test` task list.
Also:
* Changed jshintignore to use "dir/**"" instead of "/dir" or "dir"
because the latter is not compatible with Grunt for some reason.
See also https://github.com/gruntjs/grunt-contrib-jshint/issues/126.
Examples of coding style rules that were being violated that we
can now catch in node-jscs:
* Operator "," should stick to preceding expression
* Missing space after "if" keyword
* Multiple line break
* Empty block (in jquery.textSelection and mediawiki.language)
Bug: 54218
Change-Id: Ib9d7eab9f0d5cea5fb33f0b9f82e5554897fdfe0
Follows-up I1343872de7, Ia533aedf63 and I2df2f80b81.
Also updated usage in text in documentation and the
installer LocalSettingsGenerator.
Most of them were handled by this regex:
- find: (require|include|require_once|include_once)\s*\(\s*(.+?)\s*\)\s*;$
- replace: $1 $2;
Change-Id: I6b38aad9a5149c9c43ce18bd8edbab14b8ce43fa
Squiz.WhiteSpace.LanguageConstructSpacing:
Language constructs must be followed by a single space;
expected "require_once expression" but found
"require_once(expression)"
It is a keyword (e.g. like `new`, `return` and `print`). As
such the parentheses don't make sense.
Per our code conventions, we use a space after keywords like
these. We appeared to have an unwritten exception for `require`
that doesn't make sense. About 60% of require/include usage
was missing the space and/or had superfluous parentheses.
It is as silly as print("foo") or return("foo"), it works
because keywords have no significance for whitespace between
it and the expression that follows, and since experessions can
be wrapped in parentheses for clarity (e.g. when doing string
concatenation or mathematical operations) the parenthesis
before and after basiclaly just ignored.
Change-Id: I2df2f80b8123714bea7e0771bf94b51ad5bb4b87
· Use env(1) in shebangs instead of hardcoding paths.
· $IP is already set in the constructor of Maintenance classes.
· Add sapi guard to some phpunit files.
Change-Id: I6c6fd6c61e2861b5992f2ccd67a4e3f62e2c445e
There were still some files not passing jshint, and for files
that did, we managed to screw 'em up again.
Added more explicit settings in .jshintrc to avoid relying on a
kind of default somewhere. There are too many default-factors:
closest(.jshintrc), ~/.jshintrc, IDE/editor, node-jshint..
Added node_modules/ and extensions/ to .jshintignore.
Previously "$ jshint ." would recurse over all kinds of
unrelated code. Extensions should have their own jshint
dotfiles. When linting from Jenkins this won't be a problem as
those will be ran per repo (so when linting core it will skip extensions and when in an extension dir, the core dotfiles
don't apply as they'll be out of scope).
Some of our modules are really messy and should be refactored
to be less spaghetti-ish and have more descriptive variable
names and more manageable function-level complexity.
But for this commit, I'm keeping it as much as-is as possible,
because its hard/large enough to review as it is.
A few errors are cited below to give an impression of the kind
of warnings I addressed (for cases where the diff isn't
so obvious):
* jquery.hidpi.js: line 110, col 15, Empty block.
* mediawiki.jqueryMsg.js: line 34, col 17, Too many var statements.
* mediawiki.jqueryMsg.js: line 145, col 33, Strings must use singlequote.
* mediawiki.action.edit.js: line 74, col 73, 'selectText' is defined but never used.
* startup.js: line 19, col 22, 'isCompatible' is defined but never used.
* jquery.byteLength.test.js: line 26, col 9, Identifier 'U_00A2' is not in camel case.
* jquery.localize.test.js: line 63, col 29, 'attrs' is defined but never used.
* mediawiki.cldr.test.js: line 72, col 27, 'mw' is not defined.
* mediawiki.jscompat.test.js: line 6, col 17, Strings must use singlequote.
* mediawiki.api.parse.test.js: line 9, col 17, Strings must use singlequote.
* mediawiki.api.parse.test.js: line 7, col 15, 'mw' is not defined.
* mediawiki.api.parse.test.js: line 14, col 24, '$' is not defined.
* mediawiki.api.test.js: line 43, col 28, 'data' is defined but never used.
Other fixes:
* Add closures fix implied global errors ($, mw and more),
and prevents local variables from becoming globals.
* Avoid jQ magic map arg etc. (per code conventions).
* Fix incorrect usage of jQuery methods (prop instead of attr,
prop false instead of removeProp, ..).
* Unquote keys in object literals for consistency, and
enforce single quotes (no magic quotes in javascript, as much
as we might think "\n" and '/n' are really exactly the same).
Chose single quotes over double quotes since that's what most
code already had and what conventions seemed to prefer
(both the old generic ones and the new per-lang ones since
2011/2012).
* Enforce camelCase variable names with jshint per code
conventions.
* $foo.on('x', fn).trigger('x') -> $foo.on('x', fn); fn()
(No event simulation overhead, unless intended of course)
* Incorrect indentation (ignore whitespace in the diff!).
* Avoid proprietary selectors like ':first' when .eq(0)
afterwards is just as possible (significantly faster in
jQuery due to mostly avoiding the Sizzle engine and going
native in modern browsers).
* When at it, convert deprecated jQuery methods to new ones.
Mostly just .delegate(sel, type, fn) -> .on(type, sel, fn).
* Addressed whitespace here and there.
Interesting:
* mediawiki.js: local function "compare" wasn't used anymore
(hasn't been in a while!) removed per jshint warning.
* mediawiki.special.recentchanges.js: Was a mess, only a few
lines of code, rewritten.
Pfew, let's hope it's the last one before we lint from Jenkins!
Change-Id: I23ad60a1d804c542d9b91454aaa20ce7be4ff289
Updated script for generating mediawiki.jqueryMsg sample
data to be generic instead of Jasmine specific
Removed mediawiki.jqueryMsg.spec.js
* The bulk of the tests were already in QUnit (did that a
while ago)
* The little $.each loop over the sample data is now in
QUnit as well.
* Made it so that it doesn't need a hardcoded copy of
languageClasses but instead pulls it from load.php
on-demand and then restores mw.language later on.
The mediawiki.jqueryMsg.test module now has several failures
which is annoying but should not block the merge for now
because they were failing under Jasmine as well, this is a
known bug in jqueryMsg, a few cases don't work in js yet.
To be investigated (12/66 fail currently).
Change-Id: I243d055d6f5129fd9fd760943d05c7cd210d84bf
