Text Fabric
(edit) sourceAbout
Processing library for handling texts and annotations. Think of the Hebrew Bible with 4 MB of text and 33 million annotations.
It also defines a minimalistic model for such data, realised in a text-based
file format, .tf.
Tech
This is a Python package, with a lot of attention to performance. It contains quite a few loops that will frequently iterate millions of times, so we have carefully used the parts of Python that perform well.
Docs
The model, file format and API are documented in its wiki.
Status
The library is in full productive use at the ETCBC for their flagship dataset Biblia Hebraica Stuttgartensia Amstelodamensis. See also the pipeline which transports data at the institute to its website shebanq.