[jira] [Commented] (MAPREDUCE-6461) An implementation of RecordReader/FileInputFormat for generic binary data

["Ted Malaska (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MAPREDUCE-6461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14711377#comment-14711377 ] 

Ted Malaska commented on MAPREDUCE-6461:
----------------------------------------

I made this a long time ago and I know it works, because I've seen it work at a number of companies.  

Let me know if you want me to add this to the MapReduce project.

https://github.com/tmalaska/CopybookInputFormat

:)

> An implementation of RecordReader/FileInputFormat for generic binary data
> -------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-6461
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-6461
>             Project: Hadoop Map/Reduce
>          Issue Type: Wish
>          Components: mrv2
>            Reporter: Dustin Cote
>            Priority: Minor
>
> It would be nice to have a RecordReader implementation designed for binary data for ETL processing (especially for mainframe encodings like EBCDIC).  Currently, we have: http://hadoop.apache.org/docs/current/api/org/apache/hadoop/mapred/FixedLengthInputFormat.html
> This is great for fix length fields, but it would be even better if we could delimit records by a byte string dynamically.  



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)