airawat / cascading.accumulo Goto Github PK

Hey thanks for doing this! I recently modified for my company's use cases, specifically:

• A builder pattern to establish the Accumulo connection
• Support for multiple ranges
• Better iterator support
• Support for reading offline tables (directly reading rfiles)
• Support for bulk ingest (directly writing rfiles)
• Support for scalding

It does break the API in a couple of ways:

• In source mode, the tuple emits a Key and Value
• In write mode I expect a tuple of type String,String,String,ColumnVisibility, Long, Value

This means you may not want to accept a PR, but thought you should be aware of this. The code is at https://www.github.com/darinj/cascading.accumulo. It's still a work in progress I hope to add test cases and documentation in the coming week or so.