Tidies up silent vs quiet vs verbose by adding logLevel as an option.
* Notes that we need to consolidate all the various CompilerOptions
interfaces before they get too much further out of hand.
* Cleans up a couple of other minor TODO items.
In order to match existing behaviour with the legacy KeymanWeb compiler,
we should use quoted numbers for row ids. This behaviour can probably be
changed over to the new model in the future, but for now this prevents
unit tests from causing us grief.
We still have cjs modules that rely on the ldml-keyboard-xml-reader, and
this meant that an exported const was being calculated with an invalid
meta url, which crashed the cjs module require() call.
The url module is a node module. We need to move responsibility for
resolving the path of the LDML XML <import> statements out of
common/web/types, and into the ultimate consumer, so it's now surfaced
as an option, along with a helper constant that reports the
import.meta.url-relative base path of the standard imports that are
compiled into common/web/types.
This hopefully means we can use this module in both browser and node
contexts without trouble.
Fixes#8885.
Adds some optimization and error checking to the loadFile callback in
kmc-kmn. Also handles case of zero-byte bitmap file so we won't crash on
it.
kmcmplib no longer has any filesystem access, so it cannot verify if a
referenced filename in a source file has the same case as the actual
filename on disk (a risk when moving projects between platforms). So
I opted to move this to the `loadFile` callback in kmc, which is the
only place where filesystem is actually accessed, and added
corresponding unit test.
Small additional fixes here:
1. Move from `Buffer` to `Uint8Array` in all kmc-* modules, so that we
remove that barrier to running on web.
2. Use `callbacks.loadFile` instead of `callbacks.fs.readFileSync`, so
that we can be sure to run the filename consistency check.
3. Fixed kps parser silently swallowing xml errors on load.
4. Added silent mode to NodeCompilerCallbacks so we could cleanly test
the new filename consistency hint.
5. Noted a location where we still have NodeJS deps in kmc-ldml.
Fixes#8889.
No code changes, just moves WASM interfaces into
CompilerInterfacesWasm.cpp, and CompileKeyboardHandle is renamed to
CompileKeyboardBuffer and moved into its own source file.
Fixes#8886.
.kvks compiler now returns helpful errors for xml parse failures and
schema validation errors, and for invalid virtual key codes. Introduces
extra infrastructure for reporting and unit testing messages to kmc-kmn.
This also fixes unhandled xml load exceptions and simplifies the error
reporting coming out of the kvks loader, on the basis that only one type
of error was being reported anyway.
* Moves filesystem access out of kmcmplib into kmc-kmn
* Adds filesystem access callback to kmcmplib unit tests
* Cleans up callback interface through wasm
* Adds unit tests for various file load scenarios
* Removes nodefs dependency from kmcmplib wasm build, and removes
corresponding path mappings which were previously required for wasm
builds; note that these are still present for the unit tests for
kmcmplib.
Relates to #8493.
* Removes kmcmplib calls from kmcmpdll (now that we have kmc)
* Removes old kmcmp_CompileKeyboardFile and
kmcmp_CompileKeyboardFileToBuffer functions in preference for a much
cleaner kmcmp_CompileKeyboard function
* Removes json validation helper from kmcmplib (we'll use js-native
json schema validation instead)
This change means that we no longer need to keep compfile.h consistent
between kmcmplib and kmcmpdll. This will simplify upcoming refactoring
of kmcmplib.
kmcmplib no longer writes files, but it does still read them. The next
refactor will move file load responsibility into the caller.
This moves the remainder of the WASM interfaces in kmcmplib to using
emscripten bind. It is a little bit of a step backwards at present for
parseUnicodeSet, because I've changed the output buffer type to an int
for the purposes of simplifying the binding just now. But that can be
improved later, and at least we are consistent with the binding methods.
Next step is to move the filesystem access out of kmcmplib.
