Added option to apply an old mapping to a new set of codes

This can be helpful when the vocabulary and/or the source data has updated.

Added error reporting during index building

Added error reporting during index building to investigate reports of Usagi hanging during build.

Minor updates, using n-grams has become the standard

This release includes some minor updates:

• Ability to import vocabulary without CPT-4
• Lowered minimum heap size to 1GB so Usagi will run on 32-bit Java
• Using UTF-8 to store and load mapping files, so additional information can contain Chinese characters

Using n-grams for term similarity has now become the standard.

Experimental release: using n-grams

This release is for experimenting with the use of n-grams instead of tokens for matching. This should make the matching more robust against variations in word morphology. For example, 'infantile' will now match (partially) with 'infant', which was not the case before.

Vocabulary, domain, and concept class IDs now come from vocab files

Vocabulary IDs, domain IDs and concept class IDs in the UI are now

automatically derived from standard concepts in the vocabulary (instead
of being hard-coded).

Supporting the first official release of Vocabulary v5.0

This version is designed to work with the first official release of Vocabulary v5.0.

Release for beta testing

This version should be stable, but hasn't been tested extensively yet.

Note: you need the OMOP Vocab V5 CSV files to use Usagi, which at this point in time have not yet been officially released.

Evaluation version

This release is to evaluate whether Usagi is ready for users to try out