This commit adds concept of "really bad list" to the
badlists infrastructure. Really bad lists won't be
fetched from a remote server, while plain bad list
will be fetched but won't be compiled.
A really bad list is denoted by the `nofetch` token
following the URL.
Really bad lists can cause more serious issues such
as causing undue launch delays because the remote
server where a really bad list is hosted fails to
respond properly and times out.
Such an example of really bad list is hpHosts which
original server no longer exist.
Many filter lists are known to cause serious filtering
issues in uBO and are not meant to be used in uBO.
Unfortunately, unwitting users keep importing these
filter lists and as a result this ends up causing
filtering issues for which the resolution is always
to remove the incompatible filter list.
Example of inconpatible filter lists:
- Reek's Anti-Adblock Killer
- AdBlock Warning Removal List
- ABP anti-circumvention filter list
uBO will use the following resource to know
which filter lists are incompatible:
- https://github.com/uBlockOrigin/uAssets/blob/master/filters/badlists.txt
Incompatible filter lists can still be imported into
uBO, useful for asset-viewing purpose, but their content
will be discarded at compile time.
Cloud storage is a limited resource, and thus it
makes sense to support data compression before
sending the data to cloud storage.
A new hidden setting allows to toggle on
cloud storage compression:
name: cloudStorageCompression
default: false
By default, this hidden setting is `false`, and a
user must set it to `true` to enable compression
of cloud storage items.
This hidden setting will eventually be toggled
to `true` by default, when there is good confidence
a majority of users are using a version of uBO
which can properly handle compressed cloud storage
items.
A cursory assessment shows that compressed items
are roughly 40-50% smaller in size.
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/1196
For yet unknown reasons, in Chromium 84 the rendering of
`svg/use`-based icons in the dashboard causes page load
stall and undue high memory usage when the dashboard is
opened, especially the Filter list pane.
The fix in this commit is to avoid using `svg/use` and to
directly create and insert the `path` element defining an
icon.
Added a dotted box around found text occurrences,
as just pale yellow to highlight the text is not
enough to visually distinguish from surrounding text.
Iterating through found text occurrences will now
ensure they are vertically positioned in the middle
of the editor.
Code review following latest changes.
Also, move the input field to the left so that it
renders properly on smaller displays and does not
jump around when the result position/count numbers
change.
This also makes it easier to add more functionality
to the editor's toolbar in the future.
Related commit:
- 23332400f5
Since the search worker can go away after its time-to-live
elapsed, we may need to pass again the haystack on which
search operations are performed.
Before this commit, CodeMirror's add-on for search occurrences
was limited to find at most 1000 first occurrences, because of
performance considerations.
This commit removes this low limit by having the search
occurrences done in a dedicated worker. The limit is now
time-based, and highly unlikely to ever be hit under normal
condition.
With this change, all search occurrences are gathered,
and as a result:
- All occurrences are reported in the scrollbar instead of
just the 1,000 first
- The total count of all occurrences is now reported, instead
of capping at "1000+".
- The current occurrence rank at the cursor or selection
position is now reported -- this was not possible to report
this before.
The number of occurrences is line-based, it's not useful to
report finer-grained occurences in uBO.
Since it's possible to execute specific code paths according
to whether the context is that of a worker or not, it's possible
to keep the main/thread code in a single file. Keeping the
main/worker code paths into a single file is more convenient
for both code maintenance and code review.
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/1118
Usage of space in network filter patterns will now be
strictly interpreted as the filter being a hosts file
entry.
Usage of space in any other scenario will cause the
pattern of the network filter to be rejected as
erroneous.
The pseudo user styles code served only browsers based
on Chromium 65 and earlier -- Chromium 66 supports
native user styles and was first released more than two
years ago.
In Chromium-based browsers, the pseudo user styles code
is being unconditionally injected in every page/frame
just in case the browser is version 65 or earlier.
Removing pseudo user styles reduce uBO's main content
script in Chromium-based browsers by more than 20K.
Related thread:
- https://github.com/NanoAdblocker/NanoCore/issues/348#issuecomment-653646507
Related commit:
- aed850978e
No need to mind rejected promise after all,
vAPI.tabs.executeScript() is designed to
not fail -- I had forgotten about this.
Related discussion:
- https://bugzilla.mozilla.org/show_bug.cgi?id=1652925
It's not clear the code here will fix the reported
issue, but I did identify that the subframe
dictionary of a very long-lived web page can
theoretically grow unbound.
Regression from:
- 16727d68c8
The issue was causing the element picker to being
unable to select elements with no valid `srcset`
property.
Test case -- trying to select one of the embedded
frames in the following page would fail:
- http://raymondhill.net/ublock/tiles1.html
Lines in AdGuard filter lists have trailing `\r`
characters, and these caused the redirect engine
compile code to reject as invalid the redirect
token.
This is trivially fixed by trimming the raw option
strings before parsing it in the redirect engine.