Ngram based sentence generator + training engine service implemented in Elixir
Make sure you have Elixir and Docker Compose installed. Then run docker-compose up, mix deps.get, and mix run --no-halt
POST /index_text
{ "bucket": "myBucket", "text": "The quick brown fox jumps over the lazy dog" } Formless will accept either application/json or application/x-www-form-urlencoded encodings in the form body. Response body should be ok if the text was successfully indexed.
GET /query_random?from_bucket=bucket1&to_bucket=bucket2
Response body will be a sentence that was generated between the two specified buckets. Will return an empty body if the specified buckets don't exist, or if there's no intersection between the two indexes.