The back story:
Currently Keyman for Android reports errors sent to the console via
`console.error()` into Sentry but sentry-manager itself does not. This
means that Keyman for iPhone and iPad and other users of sentry-manager
do not report these errors. Many of these errors are important.
What's worse is that Keyman for Android's error reporting here (via the
`sendKMWError()` function in Keyman Engine for Android) does not capture
stack traces, and so many of the errors we get do not have enough
information to resolve them.
Furthermore, by having `sendKMWError()` in Keyman for Android, we
capture exceptions and other program errors twice -- once on the web
side, and once on the Java side -- this adds noise to our error
reporting. Sentry also tends to lump many unrelated `sendKMWError()`
events together, so tracking resolution to the errors is painful.
The fix:
This adds a patch to sentry-manager to capture `console.error()` and
`console.warning()` events and report them through Sentry's normal error
reporting, and disables the `sendKMWError()` report (although we leave
the breadcrumb in place for when there are later, related Java errors).
There is a Sentry integration called CaptureConsole, but it does not
support capturing stack traces until v6.14
(https://github.com/getsentry/sentry-javascript/pull/4034). Updating
Sentry to 6.14 or newer is a bigger job (due to ES6 baseline req. etc.)
Note that Keyman for iOS currently has some other stubs in place
overriding the `console.*` functions. These should be removed for
release builds, so that we can use this pattern instead. I will try and
tackle this in a follow-up PR (I will build it on my mac so I can test
it).
A sample error report captured with this mechanism (no sourcemaps here
because this is a -local build):
https://sentry.io/organizations/keyman/issues/3401287467/events/c40fd2cebb7743cc8dfe72e0dd34bf65/?project=5983524
I am proposing we back-port this to 15.0-stable as we are missing a lot
of data in many of our error reports on Android.
Fixes#6788.
Fixes KEYMAN-WEB-22.
Fixes KEYMAN-DEVELOPER-BQ.
Fixes KEYMAN-ANDROID-160.
Fixes a crash introduced in #6473 which was trying to prevent modifier
keys from triggering "fat finger" alternate lookups, due to incomplete
nullish coalescing.
Related to this, from what I can see, a secondary side-effect of the fix
in #6473 was that some key events may have never had their ruleBehaviors
finalized if they matched the `isOnlyLayerSwitchKey` heuristic, leading
to potential issues with edge case 'deadkey+layer switch' keys or keys
that set store values, for example. So this fix also makes the
`isOnlyLayerSwitchKey` test more targeted.
Identified this when adding a Caps Lock layer to sil_euro_latin and
testing the layer switching. At time of fix, the error had been raised
in KeymanWeb, Keyman Developer and Keyman for Android, but not yet
Keyman for iPhone and iPad.
Fixes a race condition when compiling multiple packages, where the
temp path may be re-used before the compiler has finished with it.
To take advantage of this fix, I wanted to make sure I was using the
right kmcomp.exe when building unit tests for Keyman Core. This
uncovered a couple of other things:
* Updates Keyman Core build infrastructure to use the current kmcomp.exe
rather than a system-installed one.
* Corrects a bug that was identified by the 15.0 version of kmcomp in
one unit test (duplicate `&NAME` stores -- the meson test script
prepends a `&NAME` store to each source file during build).
@keymanapp-test-bot skip
As KEYMAN-VERSION changed from being a type to a class, this broke the
bundling and referencing strategies in Developer/Server and
Developer/kmlmc. This updates the bundling process and fixes the
references in both projects.
Note that keyman-version no longer depends on gosh. Given we are
currently manually calling the build for this in all projects (or should
be!), it's better that we keep the build of this explicit for now.
If we want to restore the postinstall/postci steps in the future, the
gosh dependency would need to be manually removed from package.json in
order for the bundling builds to work for Developer/Server and
Developer/kmlmc.
There were a few temporary type patches I had made earlier in the
refactor chain in order to get a build to pass. After completing the
refactor, I was able to return them to their more proper types.
