/corpusbuilder

Corpus Build OCR platform

Primary LanguageCSSGNU Affero General Public License v3.0AGPL-3.0

CorpusBuilder

Dependencies

The development has been conducted in the environment as follows:

  • Ruby 2.4.1p111
  • Postgres 9.6.8 (TODO: check the lowest viable version given app usage of the database)

External tools and libraries

All above libraries can potentially be installed from the system's packages with most likely exception of both tesseract and leptonica.

Ruby based dependencies are handled as usual via the Gemfile. Some gems may require some external libraries, all of which should be easy to get as system packages.

Development time only

  • Node (v8.6.0)
  • Yarn (1.2.1)

All JavaScript related dependencies should be easy to get with just yarn.

Environment variables

The app expects vertain environment variables set in order to operate properly:

CORPUSBUILDER_EXCEPTION_RECIPIENTS=iamanadmin@corpusbuilder.org
CORPUSBUILDER_HOST=http://corpusbuilder.org
CORPUSBUILDER_PORT=1234

Integration

The app isn't meant to be used directly. It's purpose is to be both the database of corpuses and the tools to work on them — all to be consumed by an external application.

This means that to use it, one needs to integrate it with some other, existing app.

For the Ruby-world, a helper integration gem has been provided: https://github.com/berkmancenter/corpusbuilder-ruby-client

A more complete platform integration has been made with the SHARIAsource plaftorm. The SHARIAsource platform is open-source software and can be found on github: https://github.com/berkmancenter/SHARIAsource

Contributors

CorpusBuilder was built with the collaborations of SHARIAsource and OpenITI.

License

CorpusBuilder is licensed under the GNU AGPL 3.0 License.

Copyright

2019 President and Fellows of Harvard College.