[
https://issues.apache.org/jira/browse/HADOOP-10047?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13801909#comment-13801909
]
Colin Patrick McCabe commented on HADOOP-10047:
-----------------------------------------------
I don't think extended ZlibDecompressor is the way forward (and neither do you,
judging from your comment that "this is a violation of everything I believe
about java" :)
It sounds like the next step is to create an interface
{{org.apache.hadoop.io.compress.DirectDecompressor}}, completely separate from
the {{Decompressor}} interface, and implement
{{org.apache.hadoop.io.compress.ZlibDirectDecompressor}}. It should be pretty
straightforward-- just moving what you have now, right?
> Allow Compressor/Decompressor APIs to expose a Direct ByteBuffer API
> --------------------------------------------------------------------
>
> Key: HADOOP-10047
> URL: https://issues.apache.org/jira/browse/HADOOP-10047
> Project: Hadoop Common
> Issue Type: New Feature
> Components: io
> Reporter: Gopal V
> Assignee: Gopal V
> Labels: compression
> Attachments: DirectCompressor.html, DirectDecompressor.html,
> HADOOP-10047-WIP.patch, HADOOP-10047-with-tests.patch
>
>
> With the Zero-Copy reads in HDFS (HDFS-5260), it becomes important to perform
> all I/O operations without copying data into byte[] buffers or other buffers
> which wrap over them.
> This is a proposal for adding new DirectCompressor and DirectDecompressor
> interfaces to the io.compress, to indicate codecs which want to surface the
> direct buffer layer upwards.
> The implementation may or may not copy the buffers passed in, but should work
> with direct heap/mmap buffers and cannot assume .array() availability.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
