| Age | Commit message (Collapse) | Author |
|---|
|
Use a shadowDom for the Vomnibar (and all UIComponents)
|
|
If we use contentText, then it is possible for a child and its to
have exactly the same content text, in which case it is arbitrary as to
which is sorted first, hence which we choose (where we should be
choosing the child).
Using innerHTML instead, we guess that even if the contentText lengths
are the same, the innerHTML of the parent will be longer (since it
contains that of the child).
|
|
|
|
We need to sort elements by their text length. This allows us to
exclude the text of descendants from the hint text generated for their
ancestors.
Note: Tests not yet fixed.
|
|
|
|
Since we don't use this huge regexp on most pages, lets just compile it
when we need it.
|
|
This reverts commit 6e7b04692aceeed6544510f853eb58cdad564f9e.
|
|
Shadow DOM doesn't support <link>s, so we have to load the stylesheet
manually and inject it into a <style> element.
|
|
|
|
|
|
This reverts commit 26cbeba9b9301652d15e01ad4dac9248637ba3eb.
|
|
This insulates them from page CSS, so we don't have to compete to style
elements correctly.
|
|
Fix over-riding of vomnibar display CSS.
|
|
This turned out to be trickier than expected. It turns out there's a
period after document.body is defined but before the DOM is ready when
we can't use the HUD (but previously thought we could, ie. HUD.isReady
returned true).
|
|
|
|
This reverts commit cfb137d515e6f214322538fdd9a8e236445b63b0.
Conflicts:
content_scripts/vimium_frontend.coffee
There are HUD initialization issues with this.
|
|
|
|
Fixes #1594.
|
|
The tests are passing, but they are triggering an exception in the
initialization of the HUD. It's proving difficult to get the
initialization in the right order for both the tests and live.
This is a temporary fix...
|
|
This reverts commit ed7533b48f2f0c7cdd0272f9c6c61c44cfc5c463.
This was somehow(?) causing the HUD to not be displayed.
Need to come back and look at the tests issue.
The tests are passing, they just generate an exception.
|
|
smblott-github/fix-grab-back-focus-after-web-navigation
Fix grab back focus after web navigation
|
|
Make getVisibleClientRect behave as expected by default
|
|
|
|
Fixes #1592.
|
|
|
|
|
|
This requires passing of an extra truthy argument in order to access the
(generally) unexpected behaviour of sometimes returning the rects of
child elements. All locations in the code that *actually* wanted this
behaviour have been updated to continue using it.
Also add a comment about the unexpected behaviour in the function
description.
|
|
Fix bounds and amounts in Scroller.scrollIntoView
|
|
|
|
Don't focus <input type="submit" /> elements from link hints
|
|
|
|
This fixes #1596.
|
|
|
|
Fixes #1588. (??)
|
|
Keys aren't passkeys with Ctrl, Alt or Meta.
|
|
Some extensions (eg. Hacker Vision) override the background colour of
iframes to make the screen dark, even when the frame is already set to
transparent. This gives our rule higher priority, so that the frame is
still rendered as transparent, and doesn't obscure the content beneath.
|
|
Fixes #1586.
|
|
The UX around this is not quite right yet. It's better to disable it
for now.
|
|
|
|
|
|
|
|
- Simplify the settings logic.
- Send a single request for all required settings (instead of 12
inidividual requests for the 12 settings values we need in the front
end).
|
|
|
|
We track the URL of the active frame in each tab so that we can
correctly populate the rules in the page popup.
Previously, we tracked URL changes only on focus events. This commit
moves that to isEnabledForURL, which is also called when the URL changes
as a result of web navigation.
|
|
|
|
https://github.com/mrmr1993/vimium into mrmr1993-exclusion-rules-pushState-and-hash
Conflicts:
content_scripts/vimium_frontend.coffee
|
|
focus-main-frame-merge
Conflicts:
content_scripts/vimium_frontend.coffee
Also rework to remove dependency on Chrome 41+ sendMessage API
(specifically, relating to optional target frameId argument).
|
|
Conflicts:
background_scripts/main.coffee
content_scripts/vimium_frontend.coffee
|
|
The event listeners were registered late, potentially allowing the page
to get priority over us for key events, etc., when:
* the original URL was disabled by an exclusion rule
* the URL was changed
- without a page load (by history.pushState or modifying
location.hash), and
- the new URL isn't (completely) disabled by any exclusion rules.
This forces the event listeners to be registered even when the
current URL is disabled, to avoid this problem.
|
|
This uses the chrome.webRequest API to detect changes to page URL which
*do not* cause the content script to refresh.
|
