The new cdb library is pulled in via composer. Since the
library uses namespaces, a backwards-compatability layer
is provided for the old class names:
* CdbReader
* CdbWriter
* CdbException
The PHP/DBA-specific classes should never have been used directly.
Depends on I98302bdf1 in mediawiki/vendor
Change-Id: I39549ac8540b262cf91f7d1830d36327afb3033d
If the xhprof extension is not present we skip running the test, but
phpunit runs the test provider function before that check so we need to
guard against missing constants to avoid spurious warnings in the test
output.
Change-Id: I5541a062ff0c47ca8802315554b3f32dfd01dcd0
Add a helper class to assist in collecting profiling information using
XHProf <https://github.com/phacility/xhprof> and a Profiler
implementation to hook it into the existing MediaWiki profiling system.
The Profiler includes support for generating tabular reports similar to
the traditional output of ProfilerSimpleText and ProfilerSimpleTrace or
sending data to a udpprofile daemon as supported by ProfilerSimpleUDP.
It also produces data compatible with the debugging toolbar.
Bug: T759
Change-Id: I16a75cb7636cb5dcef3830d738b2dcd2047d0aaa
Do not base64-encode non-binary files (containing only whitespace and
printable ASCII characters, which matches sane SVG files).
For SVG files the percent-encoded URIs are actually slightly longer
than the base64-encoded ones (~10%), but compress a lot better
resulting on 15-20% less data to transfer after gzip compression.
(The effect is best seen on the 'oojs-ui' module, which consists
mostly of SVG icons – especially after commenting out everything other
than 'oojs-ui.svg.css'.)
I tried this for binary files too, just in case; but as expected, they
suffer from a noticeable size increase even with compression (~15%).
Bug: 67341
Change-Id: Iddaf863b6be98570a2bb8e606f13946a96345f65
Remove the local copy and include upstream instead.
Remove local unit tests as it's now a proper upstream lib that
is tested by uptream.
Release:
https://github.com/cssjanus/php-cssjanus/releases/tag/v1.1.0
Changes:
Features
* Support "/*!" syntax for @noflip.
* Flip border-style.
Bug fixes
* Improve flipping of background-position.
* Restrict four_notation_quantity to margin, padding and border-width.
* Flip two values in border-radius rule (in addition to four values).
Change-Id: Ieb6e179d6163f99a9c98fd99c2277d25135871fe
This reverts most of commit 2d842f1425,
leaving only the test added in it, and reimplements the same
functionality better.
Instead of stripping /*@noflip*/ annotations in CSSJanus, which is
incompatible with other implementations that preserve it, extend
CSSMin to allow other CSS comments to be present before the
rule-global @embed annotation. (This required making the regex logic
in it even worse than it was, but it's actually slightly less terrible
than I expected it would be. Good thing we have tests!)
Bug: 69698
Change-Id: I58603ef64f7d7cdc6461b34721a4d6b15f15ad79
Follows-up b36d883.
By far most data providers are static (and PHPUnit expects them
to be static and calls them that way).
Most of these classes already had their data providers static
but additional commits sloppily introduced non-static ones.
* ResourceLoaderWikiModuleTest, 8968d8787f.
* TitleTest, 545f1d3a73.
Odd unused method 'dataTestIsValidMoveOperation' was introduced
in 550b878e63.
* GlobalVarConfigTest, a3e18c3670.
Change-Id: I5da99f7cd3da68c550ae507ffe1f725d31e7666f
To do it, just remove /*@noflip*/ annotations in CSSJanus after
we're done processing. They are not needed anymore and some obscure
interactions with CSSMin logic for preserving comments caused
`/*@noflip*/ /*@embed*/ background-image: url(…)` not to work
correctly (it would not be embedded).
This also requires us to always do CSSJanus processing, even when we
don't need flipping, to consistently handle the annotations.
I'm not entirely sure if this is worth it, but I still greatly prefer
doing it to documenting this stupid limitation. :)
Bug: 69698
Change-Id: I311b12b08b2dff9d45efb584db08cf4a11318f59
Was incorrectly assuming values between -1 and 1 were 0
Fixes an issue with mw-ui-input class in RTL mode which
makes use of box shadow.
Change-Id: I13e52467721bc2534e0a7c5245e3380ac43736ef
It's not only Windows. I have an Ubuntu machine with a PHP setup
limited to 32 bit (which is totally fine, it allows me to develop
and always stay compatible to Windows and other 32 bit environments).
I don't see an other way than to skip these tests on a 32 bit
machine. The loss (PHP clips values larger than 2147483647 to
2147483647) does not happen in the code that is tested but in the
test setup.
Change-Id: Ie9a173c0a92ed84eaaea981a25ba130f2eca169e
Remapping of url()s was thrown off if the CSS contained comments
with curly braces. To fix, replace all block comments (except @embed
directives) with placeholders before url() processing and swap them
back in afterwards. Patch also includes tests.
Bug: 60077
Change-Id: If18a93c17ea9bcd529f6a664aa2dcc51d4a37f38
This adds a new generic library class IPSet, which precomputes
a reasonably-efficient data structure from the input list of
addresses and CIDR net/mask specs for fast runtime matching,
and then uses it to check trusted XFF-setters.
See also: 32b4f19f173fc5cff1029eedee63a39a2d72dd3a
Bug: 57021
Change-Id: Ia3b12fb90c3e7e492374a128943b014481cc2730
Changed closure to capital word Closure in doc and type hint,
also changed callback in docs to callable
Change-Id: I52c8e8f13d38a837052101c38b9986be780ca057
Swapped some "$var type" to "type $var" or added missing types
before the $var. Changed some other types to match the more common
spelling. Makes beginning of some text in captial.
Also added some missing @param.
Change-Id: Ic8aaf0a93796b97d0fa4617c1f86ff59f4b36131
Recognize selectors of the forms [attr=val], [attr*=val], [attr~=val]
and [attr^=val], optionally with single-quoted values.
Because these selectors previously weren't recognized, trying to
apply /* @noflip */ to one of them would be recognized as a
single-property noflip rather than a rule-wide noflip, and so only
the first property in the rule would be noflipped.
The (simplified) input that triggered this bug was:
/* @noflip */ figure[typeof*='mw:Image'].mw-halign-left { clear: left; float: left; }
which became { clear: left; float: right; } when run through CSSJanus.
See also I4cddce80397d8 which is a workaround for this issue
Bug: 50910
Change-Id: If424a1df26bb7a5a18cee4b0318b029392528fc0
* Throw InvalidArgumentException
* Use data provider in unit tests
* Detect associative arrays without copying
Per Tyler's post-merge review of Id2833c5a9.
Change-Id: Iec6b135238ca5da3002944066843102f0ae8d23d
There is nothing about the class that mandates it be used exclusively for
profiling, or exclusively within the context of MediaWiki. The functionality it
provides is generic. So move it to libs/ to make it easier to discover & reuse.
Change-Id: Iac8f2f8e1c3231ebe76318d72cf6225e25a7b891
MessagePack is a space-efficient binary data interchange format. I am going to
use it to encode profiling data in ProfilerSimpleUDP. The official PHP
implementation is provided as a C extension, so using it would further
encumber migration to HHVM. This patch adds MWMessagePack, a class implementing
a pack() method for encoding native PHP values as MessagePack byte strings. The
implementation is based on <https://github.com/onlinecity/msgpack-php>, but
revised for clarity and conformity with MediaWiki coding conventions.
Change-Id: Id2833c5a9da659cb13ec1330de9dd57138ada9c8
Introduce new static function, CSSMin::buildUrlValue.
Actually using such values in CSS does not work well because the
URL_REGEX is nowhere near good enough. :(
Change-Id: I04a7078dd0087bcb461fa5e5168c870d37c255f4
We would previously return the path to the local file on the
filesystem, which is useless in all cases and possibly a security
issue in some. Now we return the URL at which the file would be
accessible had it existed.
Also reordered the code around that part to make the control flow
clearer and added a test.
Change-Id: I1d5befb2ea385ae4d316c5d8c5d1fc092b64c4ff
Each can be selectively embedded by placing the /* @embed */ comment
just before the url() value. /* @embed */ at the beginning of the rule
affects all url() values appearing in it.
Three changes in existing behavior for previously supported syntax:
* /* @embed */ comments are no longer preserved in output
* rules not terminated by semicolons are correctly supported
* spaces within url() values are correctly supported
Bug: 46757
Bug: 56514
Change-Id: If9082f553fa920c606f12093f39f4a163ebacc32
Also added commented-out tests which should work, but don't.
Making them work in subsequent patch If9082f55.
Change-Id: I65f62493e6d10e7e90af8844f8a26e3982d75f51
We just need to negate the horizontal offset value in both of them.
This only supports *a single shadow* per element; multiple shadows
are not supported (only the first will be flipped).
Second attempt. First, reverted one: I14822955.
Bug: 45677
Change-Id: I97ee7431e1a5acb35d594076a88a0f9acf290402
The values are not "top right bottom left" here, but
"top-left top-right bottom-right bottom-left".
Bug: 49074
Change-Id: I22bc777b59e667aeb36727fdc8e41e8681979128
The 'color' rule, and by extension 'four_notation_color' too,
only understood #rrggbb and named colors. Extend it to match
rgb[a](...) and hsl[a](...) syntaxes as well.
This makes usage of rgb(a)/hsl(a) syntax in declarations like
'border-color: a b c d;' be interpreted and flipped correctly.
Change-Id: I218a9aa55a86b3d955b92375c1a209fdde312138
Don't mangle 5+ consecutive numeric values in the
'four_notation_quantity' and 'four_notation_color' rules. This
prevents box-shadow rules from being utterly broken in RTL (but still
doesn't flip them properly).
Bug: 45677
Change-Id: I16cb9e171a79c6f299b40fa02908b0c045e3c474
