[
https://issues.apache.org/jira/browse/HADOOP-13207?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15311270#comment-15311270
]
Aaron Fabbri commented on HADOOP-13207:
---------------------------------------
Took a quick scan of the attached patch.
Love the documentation, thanks.
{{+value is an instance of the `LocatedFileStatus` subclass of a
`FileSystatus`,}}
Typo in last word (FileStatus)
{quote}
+Callers MUST assume that if the iterator is not used immediately then
+the iteration operation itself MAY fail.
{quote}
Not sure if this is well-defined. "Immediately" has little meaning on a
non-realtime system. Maybe specify that, after some time, the iteration may
fail, and thus clients should handle this with a retry?
{{+is defined as a generator of a `LocatedFileS`tatus` instance `ls`}}
Typo.
{quote}
+The ordering in which the elements of `resultset` are returned in the iterator
+is undefined.
+
+It follows, that the set of paths in
+
{quote}
Unfinished sentence?
{{+of data which must be collected in a single}}
Ditto.
{{+Get the block size for a s}}
Here as well.
Test code looks good to me.
> Specify FileSystem listStatus and listFiles
> -------------------------------------------
>
> Key: HADOOP-13207
> URL: https://issues.apache.org/jira/browse/HADOOP-13207
> Project: Hadoop Common
> Issue Type: Improvement
> Components: documentation, fs
> Affects Versions: 2.8.0
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Attachments: HADOOP-13207-branch-2-001.patch,
> HADOOP-13207-branch-2-002.patch
>
>
> The many `listStatus`, `listLocatedStatus` and `listFiles` operations have
> not been completely covered in the FS specification. There's lots of implicit
> use of {{listStatus()}} path, but no coverage or tests of the others.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
