maint(linux): fix source tarball used for packaging 🏗️
The source tarball we upload to downloads.keyman.com contains all files necessary to run `${KEYMAN_ROOT}/build.sh`, i.e. it contains all projects that can be build on Linux. However, that is not the command used when building Debian packages. There we limit the build to the `linux` subfolder. Therefore we only need a subset of the files. Using the full tarball causes some lintian errors.
This change creates and uploads a second tarball on release builds that contains only the files needed for Debian packaging. The file is named `keyman_$VERSION.pkg.tar.xz`.
The source tarball we upload to downloads.keyman.com contains all files
necessary to run `${KEYMAN_ROOT}/build.sh`, i.e. it contains all projects
that can be build on Linux. However, that is not the command used when
building Debian packages. There we limit the build to the `linux`
subfolder. Therefore we only need a subset of the files. Using the full
tarball causes some lintian errors.
This change creates and uploads a second tarball on release builds that
contains only the files needed for Debian packaging. The file is named
`keyman_$VERSION.pkg.tar.xz`.
Build-bot: skip
Test-bot: skip
maint(linux): make emscripten setup on build agents idempotent
The automatic installation of emscripten on build agents did not work if `EMSCRIPTEN_BASE` was set. This change improves things for a corrupt installation by using `EMSCRIPTEN_BASE` if already set, and by skipping clone if it already exists.
The automatic installation of emscripten on build agents did not work
if `EMSCRIPTEN_BASE` was set. This change improves things for a corrupt
installation by using `EMSCRIPTEN_BASE` if already set, and by skipping
clone if it already exists.
Build-bot: skip build:developer
Test-bot: skip
fix(linux): fix context after typing Bksp with Wayland
This fixes a problem with the context after typing Backspace when using Wayland. Wayland uses double-buffering for the text, so we always have to commit after making changes.
This is only a partial fix, so we can't really test this.
Part-of: #15676
`boxArrays` had some optional chaining that was either not needed,
and/or was used in an inappropriate context, for example within the
iterable argument to a `for` statement, which would have caused an "not
iterable" error if it ever actually resolved to `undefined`.
This should have no material impact on the code, but helps to clarify
code readability.
Test-bot: skip
fix(linux): fix memory problem
This fixes a problem identified by devin.ai: Because of the operator precedence the previous code caused `memmove` to read `sizeof(commit_queue_item) - 1` bytes past the end of `commit_queue`. With this change `memmove` now reads the intended `MAX_QUEUE_SIZE - 1` `commit_queue_items`.
This fixes a problem identified by devin.ai: Because of the operator
precedence the previous code caused `memmove` to read
`sizeof(commit_queue_item) - 1` bytes past the end of `commit_queue`.
With this change `memmove` now reads the intended `MAX_QUEUE_SIZE - 1`
`commit_queue_items`.
Test-bot: skip
This fixes a problem with the context after typing Backspace when
using Wayland. Wayland uses double-buffering for the text, so we always
have to commit after making changes.
Fixes: #15676
Update the lastkey, lastscankey and lasttransition in aaitp.
This ensures that when the pfEaten is FALSE and the get_message_hook
does not recieve the key press the "last" value is correct.
