The CLDR repository is quite large, and takes a long time to clone.
NVDA only requires ~14MB of data from CLDR, and it is changed infrequently.
Rather than pull the whole CLDR repository into NVDA, this repo us used to create
the source/locale/*/cldr.dic
files, from CLDR data.
- Python 3.11-32
- Intention: Match the NVDA python version, reduce tooling complexity for developers.
- Ensure the output directory is empty or non-existent.
py -3.7-32
build.py
See out\
directory.
locale
directory contains all the generated cldr dictionaries.cldrLocaleDicts.zip
is thelocale
directory zipped.
The contents of the cldrLocaleDicts.zip
should be extracted to the NVDA source/locale/
directory.
Example path after extraction: source\locale\en\cldr.dic
.
Default log level in INFO
.
Change the level in build.py
to DEBUG
for higher verbosity.