Based on the assertion failures seen in
https://github.com/atom/atom/issues/15158#issue-247808059, it seems that
the flash for class 'c' sometimes ends before the flash for class 'd'
happens. Prior to this change, we only flashed class 'c' for 100ms, and
perhaps that isn't always enough time.
In this commit, we increase the flash duration from 100ms to 1000ms,
greatly increasing the likelihood that we're allowing enough time for
the flash on class 'd' to take place before the flash for class 'c'
ends. We also extract the scenario into its own test, so that 1) we can
more clearly explain the scenario that these assertions are testing and
2) future intermittent test failures will be easier to isolate.
As shown in #15122, this test sometimes fails in CI with the following
error:
TextEditorComponent
rendering
it blinks cursors when the editor is focused and the cursors are not moving
Expected '0' to be '1'.
at it (C:\projects\atom\spec\text-editor-component-spec.js:414:49)
Expected '0' to be '1'.
at it (C:\projects\atom\spec\text-editor-component-spec.js:415:49)
I *think* this might be a case of overspecification in the test's
assertions. Prior to this commit, the test expected the blinking cursor
to *start* in the visible state, and then transition to the invisible
state. When we see the failure above, I suspect that the cursor has
already transitioned from the visible state to the invisible state by
the time the assertion runs.
Since the test aims to verify that the cursor blinks, it seems like we
should focus on the blinking, and not worry about the *initial* state of
the cursor. This commit removes the assertions that verify the initial
state of the cursor, and instead asserts that the cursor toggles between
the visible and the invisible state.
Since this was newly introduced, let's only set it up when clock mocking
is specifically requested to minimize breakage. We really need to
deprecate the global spec helper someday. It's bad news.
/cc @hansonw
Previously, when `setModel` was called, we forgot to update the pointer
to the component in the newly supplied editor. This was causing the
element to not update in response to model updates but only as a result
of focus or visibility changes. We suspect this regressed during the
rewrite of the editor rendering layer.
With this commit we will now correctly swap the element's underlying
editor by updating the component pointer on the newly supplied editor.
Also, if the element was already attached to another editor, we will
null out the component reference on it, because one instance of
`TextEditorElement` can only represent one instance of `TextEditor`.
This commit uses a new private API on Workspace that emits an event
every time the window title gets updated (e.g. as a result of an active
pane item changing, the project paths changing, etc.).
This fixes a bug that left the custom title bar with a stale
document.title under some circumstances.
Signed-off-by: Jason Rudolph <jasonrudolph@github.com>
Before rendering block decorations, we read their heights by putting
them into a special div called `blockDecorationMeasurementsArea`.
Previously, this div was not explicitly sized, which was causing
decorations to wrap while being measured but not when actually rendering
them.
This commit fixes this inconsistency by explicitly styling the
measurement area so that it has the same width as the component scroll
width.
By the time that the animation frame is delivered, the requested
autoscroll
position could not exist anymore. This could cause the editor component
to measure a non-existent line and, as a result, throw an exception.
With this commit we will always ignore measurements for screen lines
that do not exist.
