v0.1.12

Dataset dependencies and TextBlock hashing

• Dataset dependencies fixed
• TextBlock object made hashable

v0.1.11

Fix the following issues with the huggingface Dataset dump:

• languages property (document languages) was missing
• translated and nontranslated of documents were treated as single documents

v0.1.10

add Dataset().filter_by_corpus() method

v0.1.9

Fix: add missing family slug to BackendDocument pydantic model.

v0.1.8

This Release:

---

Updates the method vertically flip coords on the parser output object to raise a custom exception should the method fail.

v0.1.7

This Release:

---

Updates to a parser output method to improve logging.

v0.1.6

Allow new data to be optional in pipeline Update type

v0.1.5

Add support for BaseParserOutput that does not require a fully specced BackendDocument set as document_metadata

v0.1.4

Support pipeline_metadata in ParserInput and ParserOutput objects

v0.1.3

This release:

• Integrates an update to the BlockTypes enum to container another field.