kassens (Jan Kassens)

forgot to remove the simple default flag removal

change the select benchmark

SPEEDUP set parsed.simple to simple Attribute Operators test attribute operator

SPEEDUP don't assume that all pseudos are not simple. let QSA figure that our for us

fixed spaces

more links to specific selector lists

updated SlickLast to latest subtlegradient/master

moar links to focused benchmarks

rearrange links

changed parsed.simple logic to figure it out dynamically when a selector fails

FIXED bug when the context is the document

moved QSA back to where it was after parsing

disable the old QSA

WIP moving QSA before parsing

almost working for everything and almost as fastas without it :/

rm tmproj

changed old idGetsName to be slightly shorter and more stable?

experimental new `idGetsName` bug checking code. More stable and reproducable in IE8's various doctypes

experimental use of QSA in the " " combinator

Refining parsed.type specs

Removing extra comma

Merge commit 'FabianBeiner/master' into tmp

(Tried to) improve(d) the translations.

Fixed typo in "Sonntag"

use context.document not document for specs

fixed syntax error

fix for the nth logics, all specs are passing now. Some little fixes and style on the "all doctype" specs too.

fix for the specs, they were giving incorrect results.

fixed whitespace and added modeline

uncommenting previously freezing tests. Thanks JDalton

Merge commit 'subtleGradient/master'

disable test in xml

use parsed.raw in QSA, not the expression so that we can accept a parsed selector object

must accept a parsed selector object QSA

Merge commit 'subtleGradient/master'

Slick must accept a parsed selector object

remove empty spec

comments

better handling of the document input in setDocument

don't return Slick for local.* methods

more css

run automatically

more SlickSpec branding

begin SlickSpec branding

FIXED specs for escaped chars

FIXED support for escaped characters in selectors

Uncomment and convert all the rest of the Google Closure Specs

FIXED specs: Should NOT include the "\" for escaped characters in the parsed selector

added ids with escaped colons to the parser specs

FIXED support for attributes without quotes with spaces, improved parser performance

better handling of random input, support for using the window as a context

added attribute selectors without quotes with spaces to the specs `[ blah *= foobar ]`

stop auto-creating a TinyURL

support detecting fake JSSpec Helper windows

slick input specs

Welcome Jan Kassens to the Slick team!

updated specs proving that the spaces in attributes bug exists

new failing parser specs for attributes with spaces around the attribute operator

Revert "convert the RegExp from being a single giant string to multiple concatenated strings. this isn't a problem since compression combines the strings together again anyway but then it doesn't save needless whitespace in the compressed version of the page."

rearranged things as per IRC discussion. no implicit declaration. explicit only

CHANGED setDocument now accepts any node or document

clean up the setDocument "feature detection"

updated SlickLast

added todo note about [attrib=square[]]

oops, bad name :/

move nth select specs from match to select specs runner and made it XML compliant

removing :index pseudo and casing for index

rewrite of nth-child logic

adding nth-child specs

Squashed commit of the following:

removing meaningless block adding missing ;

Merge commit 'fabiomcosta/fixes'

only process these setDocument tests for HTML documents

fix for the google specs, from 35 failing specs we got 21 failing now.

updated frameworks/slick.js

Merge branch 'master' of git://github.com/mootools/mootools-more

fixing a typo

updating dependencies; depender.client is no longer in -more

Merge branch 'master' of git://github.com/mootools/mootools-more

keyboard updates:

depender client is now packaged with depender app

updating unit tester

Merge branch 'master' into trunk

rewriting keyboard to allow nested controls. trust me, it's awesome.

mask now checks against document body for resizing.

adding a check on the offset parent calc in Element.position

slight tweak on the path calculation for depender.

moving the previous keyboard tests to a different local as there's another one coming in a sec.

making sure the waiter target wins over the update target

Few style tweaks in Date and URI

updating dependencies for HtmlTable

fixing a typo in the Assets docs

- Small cleanup of Request.Periodical.

picking up this fix for the overflow: http://github.com/limodou/mootools-more/commit/604f6012a7129a9a0093bcae1dca2f86b572ff78

rewrite of nth-child logic

adding nth-child specs

speedup getByTag

uidOf benchmark. We're already using the best option. FF makes the diff clear

added profiling to bench_selector to better investigate what's going on in the code and optimize

NEW set displayName for every function; allows better debugging / perf testing

fixed whitespace

moved lookupAttribute down

removed the custom getByTag since match:selector should handle that function and this therefore eliminates a function call

added YASS to frameworks

remove the feature specs from the parser spec runner

move namespace specs into a separate Describe from tagName

fixed somespecs that weren't firing in testswarm

updated test landing page

specs: document.search → Slick

run tests by default onload

make a test so it doesn't sit there doing nothing

remove google test from mocks test. since it has its own runner

temporarily remove tests which cause ∞ loop

Merge commit 'fabiomcosta/xtra-specs'

* new specs showing that the id selector still returns inconsistent values. The QSA will return every matched node with the passed id, while browsers that dont have it will return just the first one (using GEById). Browsers without QSA will still return more than one element if the element matched by the GEById function is a named node, so it will fallback to the GEBTagName and will return more than one node.

removing space between ) and {

Oops, order of operations.

.contains() doesn't exist on XML elements in IE, make its use optional.

Forgot to pass along seed and matches when qSA is used.

Forgot to pass along the seed argument to Sizzle.

Switched jQuery's multiFilter to use Sizzle.matches instead (which is much more efficient).

Missed adding in support for :animated in jQuery.

Fixed :not() functionality, made it work with :not(div #a) and with :not(:first).

Fixed an issue with multiple-depth filters failing with an incorrect parser error.

Updated jQuery to jQuery 1.3b2pre and disabled the core tests in the test suite.

Resolved the conflict.

Add failing test for attribute selector with a "." (e.g. a[href='http://www.google.com/']).

Don't consider selector expressions with a comma (",") in them as "simple". This fixes a bug where Sizzle is incorrectly passed a "not()" non-simple selector such as "not:(#mark, code)".

Run core.js unit tests too, and add a failing test for not('ID, tag selector').

Fix test broken by addition of form#name-tests in a previous commit.

Fix selector for grouped name attributes (e.g. "input[name='types[]']") not being recognized as a valid selector expression.

Add missing semi-colons.

Call getElementsByName conditionally in NAME selector function so that we don't incorrectly call it on a context that doesn't support the getElementsByName method (at this time, only the DOM Document object supports getElementsByName).

Allow name selectors with quotes (e.g. "input[name='foo']" vs "input[name=foo]") to use the NAME selector strategy (rather than falling back to the TAG strategy) and add a test suite for name selectors.

Add a note about how to run tests for jQuery with Sizzle.

I'm not crazy about it, but removeClass should run String.clean() if addClass does, that's for consistency. Another option I'm thinking of is running clean() only on Element.get('class'), that seems to make sense to me!

No more slice argument for Array.from. It's not being used and makes the code much harder than it needs to be for the HTMLCollection case, Fabio proved that. So if you wanna slice, then call slice after you've made your array.

* Possible fix for Array.from for HTMLCollection in IE. This needs testing once the Moo2 API is final.