Loading documents by tags

BiblioAlly use cases

Loading documents by tags

All documents in the BiblioAlly Catalog are tagged with one or more labels. Those tags allow to understand the state of a given document and drive the process building the selected corpus.

The code for loading documents from the BiblioAlly Catalog is below. Documents can be loaded by the method document_by() (loads a single document) or documents_by() (loads a collection of documents).

Those two methods accept specifying tags that the document must have (tagged_as parameter), tags that the document must not have (untagged_as parameter) or specific attribute values.

non_duplicates = catalog.documents_by(untagged_as=domain.TAG_DUPLICATE)
print(f'found {len(non_duplicates)} non-duplicate documents')

only_imported = catalog.documents_by(tagged_as=domain.TAG_IMPORTED)
print(f'found {len(only_imported)} documents tagged as IMPORTED')

selected_and_pre_selected = catalog.documents_by(tagged_as=[domain.TAG_SELECTED, domain.TAG_PRE_SELECTED])
print(f'found {len(selected_and_pre_selected)} documents SELECTED and PRE-SELECTED')

Rejected documents are tagged as TAG_REJECTED. The rejection reason is also a tag, so documents rejected by a specific reason can be loaded by the same pattern, just specifying the reason.

only_offtopic = catalog.documents_by(tagged_as='Off-topic')
print(f'found {len(only_offtopic)} documents tagged as "Off-topic"')