hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chris Nauroth (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-12311) Implement stream-based Filesystem API
Date Sun, 09 Aug 2015 21:55:45 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-12311?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14679371#comment-14679371
] 

Chris Nauroth commented on HADOOP-12311:
----------------------------------------

The user-facing entry point for this would be a class of our own implementing {{java.nio.file.FileSystem}},
which is what HDFS-8250 tracks.  Whether that class is implemented in terms of our {{FileSystem}}
or {{FileContext}} is a valid question though, which we should explore during implementation.
 Thanks, Allen!

> Implement stream-based Filesystem API
> -------------------------------------
>
>                 Key: HADOOP-12311
>                 URL: https://issues.apache.org/jira/browse/HADOOP-12311
>             Project: Hadoop Common
>          Issue Type: New Feature
>          Components: fs
>            Reporter: Victor Malov 
>            Priority: Minor
>
> After looking at Filesystem API, I couldn't find Stream-based API, which will be work
well with Java lambda-functions and be able chain calls.
> As Hadoop 3.0 is going to support JDK 8, I propose implement general stream-based Filesystem
API similar to as implemented in Java SE 8:
> static Stream<String> 	lines(Path path, Charset cs)
> This probably will looks similar to this:
> try (Stream<Path> stream = Files.list(Paths.get(""))) { 
>     String joined = stream
>                            . map(String::valueOf)
>                            .filter(path -> !path.startsWith("."))
>                            .sorted()
>                            .collect(Collectors.joining("; "));
>     System.out.println("List: " + joined);
> }



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

Mime
View raw message