Parser

Frinj shares the dataset for units, prefixes etc with Frink.

It uses Frink's units.txt file for building up it's own database. The units.txt file is written in the Frink programming language, so Frinj contains a parser for (a subset) of that language. It can parse enough to slurp the file, with a few minor edits (search for the tag <<Frinj edit>> in the file for details.

The database has to be built up incrementally as the data file's definition of units and prefixes depends on previous defined ones.

This process (of parsing Frink's units.txt) is rather slow, so Frinj also comes with a pre-parsed version of it's database.