We're still getting sporadic test failures with BrowserStack. So per one
discussion online, attempting to run without concurrency to see if it
improves stability.
I note that the video from tests where we "fail to capture" shows a
browser with an error: "Safari Can't Connect to the Server" "Safari
can't open the page "localhost:9879/?id=75185158" because Safari can't
connect to the server "localhost".
1971215883
Fixes#7227.
We want to use keyman.com for now even for alpha builds, as
keyman-staging.com is going to be used for transition to Docker-based
hosting, and it will be problematic to try and share that process with
alpha builds.
Adds an `index` command as suggested by #7186.
> Future enhancement suggestion: include ... index() for block indices
> within these repeated blocks. Would be very useful for string tables
> in the example ref.
This makes it easier to write a binary file which contains offsets and
size values. Using this for unit tests for the LDML keyboard compiler.
Example kmldmlc basic.txt is updated and included.
Future enhancement suggestion: include count() for repeated blocks and
index() for block indices within these repeated blocks. Would be very
useful for string tables in the example ref.
Part of #7045.
Indicates that the .kmx file has additional header structure immediately
following the COMP_KEYBOARD header. Used by KMXPlus file format for
supporting LDML keyboards. Requires VERSION_160 or higher in keyboard
file.
Fixing up a number of minor bits and pieces identified during demo:
* invalid action or :target would not be picked up if paired with a
valid one (e.g.: build:foo should not have worked)
* colors defined in both shellHelperFunctions and build-utils were
problematic; able to remove them from shellHelperFunctions because
all references included both scripts
* builder_has_action and builder_report now allow both `action:target`
and `action :target`
* keyman-version/build.sh did not have working test action...
builder_init was unnecessarily and inconsistently duplicating
builder_describe and builder_parse functionality. Removing it -- all
scripts should use builder_describe and builder_parse instead, which
gives us more flexibility and consistency into the future.
