Currently three languages not supported by Kolibri:
Skipping lang_title= Hadiyya TODO(ivan): add to le-utils so we can support this
Skipping lang_title= Sidamo TODO(ivan): add to le-utils so we can support this
Skipping lang_title= Wolaytta TODO(ivan): add to le-utils so we can support this
Will need to add to le-utils by end of week in order to import full archive.
- add option to also create separate channels for each Language
- look into making script more generic so it works for any OPDS source, e.g. using https://github.com/internetarchive/bookserver/blob/master/bookserver/catalog/Entry.py#L68
virtualenv -p python3 venv
source venv/bin/activate
pip install -r requirements.txt
./sushichef.py --reset -v --token=<YOURTOKENHERE>
When running the chef, pass lang=<code> where code is one of the following:
'af', 'am', 'bn', 'en', 'hi', 'id', 'km', 'mr', 'nr', 'ne-NP', 'nso',
'sot', 'ss', 'swa', 'tsn', 'ts', 've', 'xho', 'zul'
This will produce a channel with a single language.
Some of the PDFs and EPUB files get corrupted during download.
Consider the PDF file https://books.staging.digitallibrary.io/pdf/ben/af7ad01d-7180-4cea-bda0-747c098e7818.pdf linked to from the Referring page https://opds.staging.digitallibrary.io/ben/root.xml?page-size=100
This PDF downloads correctly with curl or when the browser is used (Firefox/Chrome tested)
but if downloading using wget produces a different file:
curl -v https://books.staging.digitallibrary.io/pdf/ben/af7ad01d-7180-4cea-bda0-747c098e7818.pdf > curl_saved.pdf
wget https://books.staging.digitallibrary.io/pdf/ben/af7ad01d-7180-4cea-bda0-747c098e7818.pdf -O wget_saved.pdf
md5 *pdf
MD5 (curl_saved.pdf) = 17a85a232cd132a48845cd37fd71ca8d
MD5 (wget_saved.pdf) = 39de452fcb41a097910363ce8009d264
Not sure what's goign on; might want to followup with digitallibrary.io dev team if issue continues.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent