By default the API uses location.host as the host, however during
development it is useful to test against production wikis
For example to test against English Wikipedia:
$wgVectorSearchHost = 'en.wikipedia.org';
Note: Links when clicked will not take the user to the target page, and
instead will take the user to the search results page with a link to
create the page.
The following config can be used to workaround that page:
$wgDisableTextSearch = true;
$wgSearchForwardUrl = "/w/index.php?title=$1";
Change-Id: I5fbac7f54844d7a9d6976007bc0d0ff9938b9f2b
Uses ResourceLoader's virtual config feature to get the config and pass
it down to Wvui's typeahead search component.
Disclaimer: I'm a typescript noob and am not sure if the
config.json.d.ts is correct although it seems to make tsc happy.
Bug: T260167
Change-Id: I2eced14c7df3b795b4de0e5149c2ca9fd598c7be
Creates a new skins.vector.search module that
replaces the searchSuggest module from MediaWiki core.
This module creates a new Vue app using the WVUI
search widget for the new search experience.
The legacy search input form is still retains on pageload,
and the new search kicks on search input focus.
In order to manage that transition, the legacy search
input is styled to resemble the new WVUI input, and the
new input is manually focused after the component mounts.
Vue is also added as a dev-dependency to help with
type-checking.
Other changes:
* the entry in skin.json is reordered alphabetically after
skins.vector.js
Bug: T264355
Change-Id: Ibb9561a77a14734297cb4d0ddcd415fc0750b45d
Updating 'search' icon to latest WVUI/OOUI optimized, reduced path. Also
- amending size of search input and position of icon in input slightly to
align it closer to standard text inputs
- simplify CSS by getting rid of selector which is targeting both, input and
button and applying rules only where needed
- fixing code comments
- increasing icon size to `16px` equivalent `em` to enable user text zooming
preference applied to search icon as well. `16px` is a compromise towards the
old look and feel of the previous icon bringing it closer to standard icon size in
legacy Vector.
This change affects both modern and legacy versions of Vector.
Bug: T266166
Change-Id: Ib4c0c74d3cac30e1893f4c76e56e1197652d41ba
Replaces vector search icon with an update version that is
copied from the OOUI icon set and used in WVUI search.
This update will help us unify the visual appearance of the
existing search input with the one in WVUI.
Bug: T264355
Change-Id: I34792ee80e711b10b441668cc4ae18cc0cc9daa6
There are popovers such as the discussion tools "Add a link" popover
that need to stack on top of the footer. This replaces the
mw-footer-container's `relative` positioning with `static` positioning
which is one possible solution to this problem.
Alternatively, the popover could be appended to the end of the body
after the footer and positioned absolutely.
Bug: T264679
Change-Id: I34168c181a1e05c33cd42f664fcccb25abd4519b
An only-child will match first- and last-child selectors
so this additional selector is not required.
Change-Id: I17d468ba2bda37edc4e54f42f42d50e4b8831b46
Following-up Iecb0d6c4c80cee1d1684597a6dedf7323e2ec54d.
It isn't set anywhere inherited, hence we don't need to set it here.
Change-Id: Ife24aa95e6d1830b8d50001976277dbb2535c47f
Dropped all usages of the no longer applicable `skin-vector-search-header-legacy`
class.
Bug: T258116
Change-Id: I16a5cf8dda2ab84ff4b505d5a368587190c409cd
Unifying blockquote spacing and border style with Design Style Guide.
Also ensuring that blockquote children don't apply inherited whitespace.
Bug: T265947
Change-Id: I43a6951c17376fe3fdd4d24966ec4661c18ffd87
Uses flexbox alignment to resolve an issue where the
clickable area of the notification and alert icons were
overlapping with the links next to them.
Bug: T264339
Change-Id: I2afc12504d7184583fa8331479125474c68017dc
Now the header is no longer absolutely positioned the height is not
fixed and the personal tools can safely spill onto a new line
Bug: T264206
Change-Id: Ib2a3cf1907c7d06c2c92ccbd902a98a3f8242f37
Follow up to 6d967ed4a8c
Add the missing ! to check that the portal is hidden rather
than visible to restore the correct logic.
Bug: T71729
Change-Id: I29f6f5e1e6adf3d6d9795cbdabcc152c5d5ac28f
On the long term we want to provide access to these variables in other
extensions. e.g. VisualEditor
Bug: T259331
Change-Id: Ibbc6f1905ea384a3d159088e3f5eca947eb6ec08
As part of comparing Vue search with legacy search, we need to track how
long it takes to lazy load the wvui library. A similar metric was added
to measuring the mediawiki.searchSuggest module in
I0fa6b8904bd43c87a68e9161f00d686a0e588966.
This commit adds the following metrics which will only be used in our
synthetic tests. We are not doing RUM tests at this time.
To test locally, add the following to your LocalSettings.php and append
the query param `useskinversion=2` e.g.
(http://localhost:8181/wiki/Test?useskinversion=2):
```
$wgVectorUseCoreSearch = false;
```
Marks:
* mwVectorVueSearchLoadStart: Marks the start of loading the search
module.
* mwVectorVueSearchLoadEnd: Marks the end of loading the search
module.
Measures:
* mwVectorVueSearchLoadStartToLoadEnd: Measures the time it takes to
load the search module.
Bug: T251544
Change-Id: I14e44b45a66213821d69cd22395fedbae747da88
Drop support for vectorMenu, vectorTabs and
vectorMenuCheckbox, body, menu selectors in preference
for standard selectors.
This change will impact a large amount of user scripts/styles but should
not impact any gadgets.
These classes were kept around for user scripts and styles however are not
needed internally. As we transition to a more maintainable skin menu
system, it is time to lose these selectors even though this will cause
disruption.
Vector now will use the mw-portlet class rather than the vector-menu
class in its own CSS styling, however it keeps the other classes to
allow differentiation of the different types of menu.
Changes to test: Previously the tests assumed all portlets were empty
when checking the classes. This is very rare, so its better to check
the classes of non-empty portlets, so several tests are updated
accordingly to drop the emptyPortlet class.
Bug: T262092
Change-Id: I1824335eb47d613c2a4804ec1f1106c0f4c16101
Kept as simple as possible for now. The new class is added but no classes
are removed. This will be done in a follow up.
Bug: T256897
Bug: T253938
Change-Id: Ib31a9d8f2ac14e63b63e82abd4a9aa1fcb956f45
* The margin-left should only be auto at small resolutions - otherwise on large
monitors it becomes visually separated from the logo
* the margin left should be larger
* and max width should be smaller.
Bug: T261686
Change-Id: Ia1331f51764a34f113e3336735e6cfd5fde1d49d
Minor visual fix for IE9 for modern mode with search not in header.
i.e. with these settings:
$wgVectorIsSearchInHeader = false;
$wgVectorDefaultSkinVersion = 2;
Slight refactor of CSS. Removes block of styles that "reset" floats.
These were uneccessary because as flex-children, the floats were
ignored on those elements anyway.
Moves flex-wrap on .mw-header from layout-search-header.less
to layout-default.less since that's where display:flex is defined
and the rule is general enough that it should apply for different layouts.
Remove the @supports block with the `float:none;`. That is unnecessary
because flex-children are not affected by floats.
Change-Id: Ida3d2a7bc2b2f70238129df876714228fe5cdf84
This moves the header, navigation, sidebar, and article toolbar to be
before the content in the DOM. As a result, a lot of absolute
positioning logic can be removed and styles can be simplified.
Note that although the sidebar was moved from the header into the
workspace container allowing it to de-absolutely positioned, its
absolute positioning was kept intact as it has a fair amount of
complexity that should be handled in a separate task.
To activate, set `$wgVectorIsSearchInHeader = true;`
Changes that could cause concern:
* The "jump to search" link was removed as the search is now much
earlier in the DOM and I questioned the value of keeping this. However,
it can be added back in if this change is contentious.
* A "jump to content" link was added to account for the new DOM order.
* Because the sidebar was taken out of the header, users will not be
able to tab from the sidebar button into the sidebar without additional
tweaking (e.g. should we add JS to enable this?). It was deemed that
this work can be saved as a follow-up task.
* I applied `overflow-y: auto` to the `mw-page-container` because the
header's top margin was collapsing and caused whitespace to appear
between the viewport and the header. Alternatively, we could apply a top
padding to the page container and remove the header's top margin. I went
for the simplest solution but am open to alternatives.
* I left the footer as-is in this patch to minimize risk. It might be
cleaner later on to move the footer inside the workspace container which
would leave only one workspace container.
Bug: T261802
Change-Id: Ic553fab3bde25769b103d899b92b3b694c00c384
Provides a loading indicator to show while the new Vue.js based
search widget loads. Given that the new widget will pull down the
entire Vue.js runtime, it's likely that there will be a delay
before the search suggestions appear. This loader is meant to
improve the perceived loading experience of the new widget.
Adds:
- New searchLoader.js file containing loading behaviour.
- This overrides the code searchSuggest loading behaviour.
- New SearchBoxLoader.less file containing the loader styles.
- i18n message: 'vector-search-loader'.
- The Event type to jsdoc.json
Bug: T254695
Change-Id: I6b5f0a60018954e10b9e80792030b67b2ec33e5a
The star should spin its current background image when transitioning between different
watch requests.
Bug: T259053
Depends-On: I1e11f0e129c53b405a2ffa8
Change-Id: Id2f9b2e25761c052aeaa410edead65ec298209a2
The width should apply at all resolutions. Note, because of the nature
of flex box and flex-grow the personal tools can grow larger than this
value.
This avoids Alex Hollender (WMF) and similarly long usernames from every
running to the next line before they are allowed to.
Bug: T249363
Change-Id: I4640947aaaf7ab764cb17b911af7085ac291b7d1
Carved out from I8e6e283db5d4034861f8d50d9b3f211df1a78c6c.
It makes sense to have a variable defined for that value.
Change-Id: Iaaca1fd7c15a0c121e60b5ae6debf4bf6920168e
I overlooked the horizontal padding on the page container. This needs to be
included in the decision on whether to make the header 2 lines on 1.
Bug: T249363
Change-Id: I4fabac7d57e37db87d2363073317109f582de883
Following up I6bc4cf541eefd00e2e42 we also need to remove the
linear-gradient hack that only made sense in combination with the
now gone PNG fallback background-image.
Change-Id: I0e7ed0451884a6bd612cb1082555338a26129e2d
These are no longer needed. The classes remain where necessary for
gadgets but the CSS rules no longer need to apply to them.
Change-Id: I18afa15ddab75128463dc83c916e11436db0575a
The calculations were a little incorrect as I failed to consider the
sidebar button correctly and how the search's min width and max width
impact layout.
I also move rules from Sidebar.less regarding the placement of the button
into layout where I believe they belong. We do not have a header component, so the
positioning (margin) of the sidebar button in current form should be here.
This can be revisited if we introduce a header component.
Bug: T249363
Change-Id: I4ff640380eafc8beedb2c3c8fb00a56c71c5cb45
The width comfortable should consider the max-width of the search
not the min-width.
This fixes the bug documented in T249363#6391041
Bug: T249363
Change-Id: I3e216a3705730092f88d1dcbb5193e411945a083
As part of moving search into header work, a min-width (via
@min-width-supported variable) was introduced on the body and takes
effect when the search feature is enabled.
However, given a min-width already exists on the page container from the
max-width work, I'm wondering if it makes sense to replace that one with
the @min-width-supported variable as it seems like we should only have
one min-width vs. having two.
Note: As a bonus, this has the (unintended but helpful) side-effect of
mitigating the sidebar button being blocked by the personal menu at
small viewport widths (T258465).
[1] I7f8059d43eaab49de362405784b34a4fe502c7b0
Bug: T258465
Change-Id: I920cd0e9d1564c82bcdc89b721352620158073c6
* Remove the PNG fallbacks for chevronHorizontal-….svg and menu.svg.
As of T248061, these are no longer needed.
* Added the one line of trivial CSS directly to skins.vector.styles
instead of through its own module.
This helps recovers the module cost of vue module deployed this
week (from Ib6c8f890fb3d6e7), which is currently empty and unused.
With T253582, we'll be able to recover a lot more budget in
this area.
Bug: T258766
Change-Id: I6bc4cf541eefd00e2e428f918664a26da331c1a9
To support roll out and avoid issues with cached HTML the new
styles for the new search feature are restricted to HTML where
the body tag has `skin-vector-search-header` class.
For legacy mode, we introduce a new class
`skin-vector-search-header-legacy` and temporarily use a CSS3 `:not()`
selector to ensure the styles ship during the phase where cached
HTML can be served. While this will create some display issues in
browsers that do not support CSS3 selectors, all grade A browsers in
our compatability matrix support this.
Bug: T249363
Change-Id: I7f8059d43eaab49de362405784b34a4fe502c7b0
Keep variables concerned with layout in the master file but pull out
the default layout into a separate file.
Change-Id: I4acc2937f8e8a76274a3ffb76e3729dc89ce1ad7
Add onSkinPageReadyConfig hook that overrides module after page loaded
The new module is currently empty pending further work in the
feature branch.
Depends-On: I0dc38e74052027f26a70d58b5f520e5830e0d55d
Bug: T257706
Change-Id: Ib6c8f890fb3d6e751f5f01a6576614b9cc9b440c
`padding: 0` was either a rule targeting very early Operas or
Netscape/Firefox or came out of Eric Meyer's reset.css as
misappropriation.
normalize.css haven't had it in v1.0.0, only relied on `margin: 0`
normalization for IE 6 & 7.
Change-Id: I3d2894a1e68414b64751bd6ebe7e1af77d260ee7
In PHP we add collapsible classes to all elements except watchstar
so that certain tabs can be collapsed under the more menu in JS.
This adds unnecessary complexity to our codebase and is not used
if JS is disabled.
To simplify this and bring Vector's PHP consistency with core this
logic is moved to JavaScript.
Bug: T259372
Change-Id: I2acbf7089198118626368ee8a37615d2de062f83
Per discussions, its proposed that the target of all media queries
is defined in the entry points skin and skin-legacy
Please verify with `git diff HEAD^ -w` that no changes to print
styles have occurred
Bug: T253842
Change-Id: Id7d1c806d77ee50335a1c9985acc7e4406e64ccf
All those have been lingering while not being used for a while.
- arrow* files are part of core 'mediawiki.icon' module
- 'link-icon.png' is without any reference nor clear original usage
- 'magnify-clip.png' is part of core 'mediawiki.skinning' module
- 'search-fade.png' was used in #simpleSearch before it got unified with
Design Style Guide components
Change-Id: I985d3ec25b26eea359ce3dbf5abbe4647bc37ab4
Use normal `background-image` properties with SVGs only now
that IE 8 and Android 2.1 is removed from Grade C.
Also removing all PNG fallback images.
Bug: T248062
Change-Id: Ib91cd0514d331ab6a0f8b668aef6991cf3267fe2
In future we will be displaying the header and logo in the print
display - to avoid adding unnecessary borders scope this rule.
Bug: T253842
Change-Id: I123b73fcaa09c0213914ca6fd8074a1305814529
$.collapsibleTabs.getSettings can return undefined so the code
should take this into account
Bug: T177108
Change-Id: I2630a2ba2884542869748047ab84752543072647
This will help mitigate the sidebar overflowing the page container
(related to T257518). Note that this does not prevent the overflow from
happening as the sidebar can still be longer than the viewport, but it
should reduce the number of times it can happen.
Bug: T257518
Change-Id: Id7138b4d4459242772bee8e11dc7edeaf76b3ca0
Follow up to I340b9e7e91960713c0ebb4d3d26e2ae2d5628f37
The layout styles reference internal CSS classes within
Vector components that may change at any time.
For legacy layout, I leave the styles the same (they have a FIXME)
As Aron noted on code review the impact such a change
could have on user styles. For modern however I simplify the styles
as follows:
* The top margin on the form is promoted to the main element - this has
the same end result.
* Likewise the width dimensions are moved from child elements to
the parent
Bug: T249363
Change-Id: If923a5dddaac6217462e75d476e07d923ee1743f
No changes made to the CSS rules.
* 'SearchBox.less' is the common part that won't change in modern.
* layout styles copied to both 'layout.less'
Bug: T249363
Change-Id: I340b9e7e91960713c0ebb4d3d26e2ae2d5628f37
* `mw-content-container` now wraps the footer (as well as the content)
because we want the footer to match the content width at all times and
to expand with the content when the sidebar is closed (at small viewport
widths or when on history/special pages)
* `mw-footer-container` margins were replaced with padding to avoid
issues with margin collapsing.
* Applied a white background to sidebar to handle the case of the
sidebar overflowing the `mw-page-container`. When that happens, we at
least want the text in the sidebar to be legible.
* Closely related, `mw-page-container`'s `overflow: hidden` style was
removed to prevent `mw-page-container` from cutting off the sidebar. The
purpose of this style was make it appear as if the sidebar was being
hidden by `mw-page-container`, but tweaking the sidebar's translation
animation to achieves this effect as well.
Bug: T257518
Change-Id: I89edf89b2ac4abe2053f0c9b366f143133ff420f
- Creates a new user-preference called 'VectorSidebarVisible'
which stores the sidebar hidden/collapsed state for logged-in
users.
- Updates that user-preference on the client whenever the sidebar
is expanded or collapsed.
- Refactors the sidebar related javascript into a separate file.
Bug: T255727
Change-Id: Ib1ce934f3646cd8feebf0d3b15c38b5b969ec957
This patch closely follows the desired guidelines/desired
styles Alex Hollender has put forth in his prototype, but uses
multiple containers to achieve this look since our DOM order/structure
is different than the DOM structure in the prototype. The following
containers are used, but unlike his prototype, they are sometimes used
more than once:
* Page Container: Contains every other container and limits the overall
max-width of the white part of the page.
* Workspace Container: Contains the sidebar and content container. The
sidebar is displaced ~30 pixels to the start (left) of the workspace
container at all times.
* Content Container: Contains the content. The max-width of this changes
depending on whether you are on a special page/history page vs. other
pages.
* Article Toolbar Container: Contains the article toolbar. The max-width
of this is always the same as the max-width of the article content as we
don't want the toolbar to move when going from the article page to the
history/special page.
Changes to be aware:
* To test locally, `$wgVectorLayoutMaxWidth = true;`. This design is
temporarily feature flagged and defaults to being "off".
* Note that layout-max-width.less is a temporary file made to meet the
feature flag requirement of T246420 (intended to derisk the deployment).
After the deploy, we should merge most if not all of the rules into
layout.less where the max-width design will become the default.
* Per Jon's code review comment, I have relaxed the indenting of
skin.mustache to make the diff easier to reason about. If desired, the
correct indenting can be achieved in a (much less risky) follow-up
commit.
Bug: T246420
Bug: T153043
Change-Id: Ie49f629bc705850c6996164a516957476c034048
This explains a few issues that were occuring in the storybook
relating to modern/legacy.
Follow up to 5195f5fd6 which doesn't reference why.
Components should not import styles relating to one of the Vector
versions. They should all be self contained.
Change-Id: I6f9ef974be97f8be593cd2b79ce37c3e517d549f
Removing presentational `<div class="visualClear"></div>` in process.
Only leaving in for legacy/layout.less and adding comment for future
evaluation if clearfix is continuously needed. In modern we've
got all parts of Vector in our awareness without featuring `float`s.
Bug: T254195
Depends-On: Iddf8f1dffc2d30299b89b3f4966b7fe8ee63090b
Change-Id: I673c28c2d7da4f96c31121d9aec6558e390de24e
Currently, the `aria-expanded` attribute is placed on the checkbox
element. However, since Ife287fc8c6e0d2aee5facf42d5d4308dea918ee3, the
checkbox is excluded from the accessibility tree, and this attribute
should be placed on the label button instead.
This commit prepares for future changes [1] to the checkbox hack interface
in a backwards compatible way:
* Passes button to `updateAriaExpanded` function
* Passes button to `bindUpdateAriaExpandedOnInput` function
[1] Ia2755e189babbd70945b66a1a812fc3ece40b577
Change-Id: Icc6ba994d57ea1f8050aa408aebc8c81f03d8783
* Moves the sidebar into the header so that user can tab directly from
the sidebar button into the sidebar (when open).
* Because the sidebar is absolutely positioned inside the header and the
header applies a top-margin, we need to adjust the top position
calculation for the sidebar.
* Removes the checkbox from the accessibility tree through CSS `display:
none;` instead of HTML attributes.
* Given that the checkbox is not able to receive focus, the
`#mw-sidebar-checkbox:focus ~ .mw-header .mw-checkbox-hack-button` CSS
rule is obsolete and moved instead to the label button. An additional
outline: 0 rule was added to remove the dotted outline that Firefox
applies.
* Makes the "Skip to navigation" jump link point to the sidebar now.
After the sidebar has been tabbed through, the rest of the navigation
can be tabbed to.
Bug: T246420
Change-Id: I981da3650854568bb9cfbf3c6c59e7625e7d094c
This separation introduced in I8dbc29b7a19f7f doesn't work as well
as expected. All but strictly `.mw-jump-link` rule are typographic
rules.
In the current mish-mash of when to apply typographic styles only to
article content (`.mw-body`, `mw-body-content`, `.mw-parser-output`,
`#mw-data-after-content`?) inherited from core, it makes looking up
and finding issues between typographic styles harder, and leads to
unnecessary complexity instead of clean separation of concerns.
This question surfaced in where would link styles belong originated in
T213778.
Change-Id: I521185d505d8688f076dd09acbedb22e801f772e
jdlrobson
jenkins-bot
Nicholas Ray
Jan Drewniak
Volker E
Ed Sanders
Timo Tijhof
Stephen Niedzielski
hmonroy
Peter Ovchyn
AronDemian