There were two issues here:
- The comparison was backwards, causing scroller.at_bottom() to always return
true.
- When zoomed in, jsret['px']['y'] can be a float, which means we can be
slightly off when checking the difference - math.ceil() fixes that.
Trying to read from the sql database from another process was flaky.
This adds a debug-dump-history command which is used by the history BDD
tests to validate the history contents.
It outputs history in the old pre-SQL text format, so it might be
useful for those who want to manipulate their history as text.
Returning "next" was no longer possible as the SQL query does not fetch
more items than necessary. This is solved by using a start time, a
limit, and an offset. The offset is needed to prevent fetching duplicate
items if multiple entries have the same timestamp.
Two of the history tests that relied on qute://history were changed to
rely on qute://history/data instead to make them less failure-prone.
Calling sql.init() in version.version() would replace the existing sql
connection and cause a crash when accessed by opening qute://version.
Now version relies on sql already being initted, and app.py inits sql early if
the --version arg is given.
Two history end2end tests are failing because sqlite is not flushing to disk in
time to be read by the test process. My understanding is that sqlite should
take an exclusive lock while writing, so it is difficult to understand why this
is happening. This can be fixed by adding a delay, but that seems flaky.
I'm fixing it by checking qute://history instead of reading the database file.
See:
https://github.com/qutebrowser/qutebrowser/pull/2295#issuecomment-292786138
and the following discussion.
Now that sql is only used for history (not quickmarks/bookmarks) a number of
functions are no longer needed. In addition, primary key support was removed as
we actually need to support multiple entries for the same url with different
access times. The completion model will have to handle this by selecting
something like (url, title, max(atime)).
This also fixes up a number of tests that were broken with the last few
sql-related commits.
Change the logging to report the completion function name and have the end2end
tests check for this.
Remove the tests for realtime completion, as it was decided this is not an
important feature and the code is much simpler without it.
When we open a background tab, it gets a hardcoded size (800x600 or so) because
it doesn't get resized by the layout yet.
By resizing it to the size it'll actually have later, we make sure scrolling to
an anchor in an background tab works, and JS also gets the correct size for
background tabs.
Fixes#1190Fixes#2495
See #1417
In a48ea597d0 we fixed settings in private
QtWebEngine windows.
However, this means we also enable local storage for private windows, which was
disabled in QtWebEngine by default:
4ef5831a39 (diff-44ac7d27348388501944f6a8e2e67d8dR207)
It should be safe to enable it, as we get the same behavior as in Chromium, i.e.
a working local storage which entirely lives in RAM.
This also makes those tests work on QtWebKit-NG, presumably because private
browsing for cookies is implemented there.
It also adds a test to at least check whether local storage is isolated from
non-private tabs. I tried writing a test which ensures nothing lands on the hard
disk, but due to QTBUG-52121 this might not happen at all:
https://bugreports.qt.io/browse/QTBUG-52121
Turns out QWebEngineSettings.globalSettings() only sets things on the default
profile. We now get everything from the default profile settings, but set it on
both the default and the private profile.
Fixes#2638
(cherry picked from commit b11a4388cd10b6ff2fd917fca689ebdc50d581ae)
If we don't wait here, we might end up running the subsequent commands (like
:command-history-prev) on the old window while it's still closing, causing an
exception at least on AppVeyor:
Traceback (most recent call last):
File "C:\projects\qutebrowser\qutebrowser\app.py", line 110, in <lambda>
target_arg=target_arg))
File "C:\projects\qutebrowser\qutebrowser\app.py", line 265, in process_pos_args
win_id = mainwindow.get_window(via_ipc, force_tab=True)
File "C:\projects\qutebrowser\qutebrowser\mainwindow\mainwindow.py", line 89, in get_window
window.setWindowState(window.windowState() & ~Qt.WindowMinimized)
RuntimeError: wrapped C/C++ object of type MainWindow has been deleted
There's actually no good reason to filter javascript links as we might want to
click them (or copy their URL) just like any other link - this fixes#2404.
With that being gone, we don't need FILTERS at all anymore, as we can check for
existence of the href attribute in the CSS selector instead.
Looks like we get this sometimes:
----> Waiting for 'Clicked non-editable element!' in the log
14:02:14.976 DEBUG webview webkittab:find_at_pos:618 Hit test result element is null!
14:02:14.976 DEBUG mouse mouse:_mousepress_insertmode_cb:149 Got None element, scheduling check on mouse release
14:02:14.977 DEBUG mouse webview:mousePressEvent:299 Normal click, setting normal target
14:02:14.978 DEBUG mouse mouse:mouserelease_insertmode_cb:173 Element vanished!
This could happen for any of the attributes, but for tagName this actually
happens in the wild... Since elem.tagName is equal to elem.nodeName we just try
to use this.
Fixes#2569
Before, we just returned the same data for both, but then we'll run into
same-origin restrictions as qute:history and qute:history/data are not the same
host.
Problem 1: Entering a command of `:::save` gives an error.
Problem 2: Entering a command of `:save\n` gives an error.
Both scenarios may seem a bit silly at first, but I encountered both by
copy/pasting a command:
1. Enter `:` in qutebrowser.
2. Copy a full line from a terminal starting with `:`.
3. You will now have both of the above problems.
Solution: Trim all whitespace and `:` of a command. This is also what
Vim does, by the way.
This really tripped me up yesterday, My "Vim default" is to use tabs.
This (where `!···` is a tab) does not work as you'll hope it works:
Scenario: Retrying a failed download when the directory didn't exist (issue 2445)
When I download http://localhost:(port)/data/downloads/download.bin to <path>
And I wait for the error "Download error: No such file or directory: *"
And I make the directory <mkdir>
And I run :download-retry
!···!···And I wait until the download is finished
Then the downloaded file <expected> should exist
Examples:
| path | mkdir | expected |
| asd/zxc/ | asd/zxc | asd/zxc/download.bin |
Unfortunately, pytest-bdd uses the "Python 2 behaviour" of "expand all
tabs to 8 spaces", and doesn't give any errors on strange/inconsistent
whitespace. It can cause very confusing errors.
Using focus() in JS there means that existing text in the field gets selected.
Move the cursor to the end after focusing it to prevent that.
Fixes#2359
Add two extra tests for checking navigation on pages with rel "next" and
"prev" links which are also rel "nofollow" to test for the correct
functionality of navigating pages with rel "next" and "prev" links with
multiple rel attributes.
- We need to clean open tabs to avoid reusing target=_blank child tabs
- We don't check the active tab with target=_blank anymore
- Remove some weird :tab-close
We now use click() or focus() in JS if possible, or manually follow links in a
href attribute.
While this probably introduces some new corner cases, it fixes a handful of
older ones:
- window.open() in JS can now be handled correctly as we don't need hacks in
createWindow anymore.
- Focusing input fields with images now works - fixes#1613, #1879
- Hinting now works better on QtWebEngine with Qt 5.8 - fixes#2273
Also see #70.
Ignore all keys with an empty .text() return value, not just modifier
keys. You can still use unusual things like ß for registers, but
XF86WakeUp is out. Fixes#2125.
I tested everything that I thought was interesting enough to warrant a
test: especially the ability to test multiple parameters deep, as well
as testing :set --cycle and some involved example to make sure
completion actually works and updates in realtime
Chooses the next value from the provided list of values (string-wise
comparison). Technically, the 'option!' syntax for toggling bools is now
redundant, but a translation from 'option!' to '--cycle option false
true' is kept for backwards compatibility.
The '--cycle' flag could also be technically optionally, since the only
thing that depends on it is preserving the error message for specifying
multiple values. (But I think it's best to keep it explicit, as a
principle-of-least-surprise thing)
Note: The business logic of picking the next value and setting it was
moved out to a separate function to avoid tripping pylint's
too-many-branches detector.
Fixes#47
This initially seemed like a nice feature, but it means 0 can't be bound
as a separate key anymore, and 0<Esc> gives weird error messages...
Reverts #1953.
Fixes#2032.
When starting a download due to unsupportedContent being emitted, we
need to use (and later adopt) the page's QNetworkAccessManager.
Since we need the whole adopting logic for that case anyways, let's keep
things as they were and always run downloads in per-tab QNAMs.
This reverts 53e360ec4b and fixes#2134.
When the user pressed esc during an SSL prompt, the message.ask call
returned None, which was handled fine by QtWebKit (which simply used an
'if') but failed with QtWebEngine (which returned the value to Qt).
Fixes#2123.
webelem.text() was only used without use_js=True from webelem.__str__.
Now we instead default to the use_js=True behavior and renamed the
method from text() to value().
The old behavior is instead directly implemented in __str__.
This changes how webelem.value (aka text) handles elements for which
is_content_editable() returns True, but I haven't found any cases where
this makes a difference.
This also fixes getting existing text from elements with QtWebEngine,
which closes#1957.
There were actually two issues here:
- The override_target was reset too early - now
acceptNavigationRequest/createWindow is responsible for resetting it.
- The QTBUG-54419 workaround opened any tabs in the background instead
of preserving their position/focus.
Fixes#2086.
