[
https://issues.apache.org/jira/browse/HADOOP-3315?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12629910#action_12629910
]
Hong Tang commented on HADOOP-3315:
-----------------------------------
> If you instead extended ByteArrayOutputStream, BoundedByteArrayOutputStream
> would only need to override the two write() methods, rather than implement
> all of the methods.
[Hong] This class is optimized for TFile key appending, The whole
implementation only consists of the two writes() and needs nothing else from
BufferedInputStream.
> Then it should be named something different. Why support negative values at
> all if you don't use them? Lucene also defines a VInt format that might be
> considered. Personally, I'd prefer Hadoop used a single VInt format and I
> don't think it is worth defining yet another VInt and String format for TFile
> is wise.
[Hong] We can sure rename it to avoid confusion. We do need to represent
negative integers. But they are either very small (such as -1, or -2), or very
large (-1M).
> Then CompressionCodecFactory should be extended, rather than duplicated.
[Hong] This is a first cut implementation. As it stand, it does not duplicate
much code. We will consider refactoring based on your suggestions later.
> New binary file format
> ----------------------
>
> Key: HADOOP-3315
> URL: https://issues.apache.org/jira/browse/HADOOP-3315
> Project: Hadoop Core
> Issue Type: New Feature
> Components: io
> Reporter: Owen O'Malley
> Assignee: Amir Youssefi
> Attachments: HADOOP-3315_TFILE_PREVIEW.patch,
> HADOOP-3315_TFILE_PREVIEW_WITH_LZO_TESTS.patch, TFile Specification Final.pdf
>
>
> SequenceFile's block compression format is too complex and requires 4 codecs
> to compress or decompress. It would be good to have a file format that only
> needs
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
