[
https://issues.apache.org/jira/browse/CRUNCH-337?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13890784#comment-13890784
]
Micah Whitacre commented on CRUNCH-337:
---------------------------------------
+1 to the changes. If we end up with too many more variations we might need to
look at changing to more of a builder pattern but for this simple of a use case
it works and makes things easier.
> Make it easier to use multiple input paths
> ------------------------------------------
>
> Key: CRUNCH-337
> URL: https://issues.apache.org/jira/browse/CRUNCH-337
> Project: Crunch
> Issue Type: Improvement
> Components: Core
> Affects Versions: 0.9.0
> Reporter: Matthew Basil
> Assignee: Josh Wills
> Priority: Minor
> Attachments: CRUNCH-337.patch
>
>
> It would be more user-friendly, especially for newbies, to provides methods
> on {{From}} for creating sources from multiple {{Path}}s. I'm currently
> attempting to write my first Crunch Pipeline, which needs to read from
> multiple paths using a custom input format, and I needed to dig into the
> source for {{From.formattedFile}} to see I need to do something like this
> {code}
> PTableType<K, V> tableType = keyType.getFamily().tableOf(keyType, valueType);
> return new FileTableSourceImpl<K, V>(paths, tableType, formatClass);
> {code}
> I don't particularly mind, but other potential new users might be a bit put
> off by having to look at the source on the first line of their first
> pipeline. If it's undesirable to double the number of methods in {{From}} by
> doing this (which is understandable), it might be nice to add some note on
> multiple input paths to the section of the users guide on {{Source}}s.
> Thanks!
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
