I'm glad to announce the first release of an implementation of map-reduce framework in Lua, using luamongo and temporal disk files as its communication protocol between processes. The implementation stills very preliminar, but is working nice with large datasets. For example, word-count example, in a 2M sentences file could be executed obtaining the same performance as a naive shell-script word-count implementation, even with only one worker. Suggestions and contributions will be welcome. Documentation stills being inexistent, I hope to write more details whenever I can.
Cheers,
