v0.5

This release is a significant step forward with all proof-of-concept apps now commented and documented, and working reliably with the data model. The C++ TestGrammar application has a choice of pdfium or PDFix SDKs. The XML and TSV file sets are now all consistent and updated. There is also a Jupyter Notebook and Python3 script to make a monolithic TSV file that works with pandas.

Prior to predicate parsing and switch to pdfium.

This tag represents the Arlington PDF model prior to the significant rework to fully parse predicates and switch to pdfium as the preferred PDF SDK. In this release PDF string objects use brackets "(hello)" which is ambiguous for parsing with expressions, as well as changes to a few predicates to simplify and disambiguate parsing.