incubator-crunch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gabriel Reid (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CRUNCH-97) Add helpers for parsing PCollection<String> instances
Date Sat, 24 Nov 2012 09:50:58 GMT

    [ https://issues.apache.org/jira/browse/CRUNCH-97?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13503364#comment-13503364
] 

Gabriel Reid commented on CRUNCH-97:
------------------------------------

I just (finally) took a look at the new ReviewBoard patch of this (https://reviews.apache.org/r/8151/),
and I'm even more enthusiastic about this now -- it almost seems amazing that we didn't have
something like this already, as it's going to be incredibly useful.

Just a few small comments on it (I'm putting my comments here instead of on ReviewBoard, not
sure which is best)
 - The group name used for the error counters in Parse.java is "test". It would be good to
have a better default ("parse"?), and/or add it as a parameter to Parse#parse and Parse#parseTable
- It might be helpful to add debug logging to the parsing error condition in AbstractSimpleExtractor
and AbstractCompositeExtractor. As it is right now, if there's a field that isn't parsing
properly somewhere in an input using the built-in numerical parsing, there's no way of finding
where it's going wrong.
- It might be good to add a bit of javadoc to the fieldErrors field and getter in ExtractorStats,
as it wasn't immediately clear to me exactly how this was to be filled in (on the other hand,
in most cases this won't be used directly anyhow, so it's probably not that important).
                
> Add helpers for parsing PCollection<String> instances
> -----------------------------------------------------
>
>                 Key: CRUNCH-97
>                 URL: https://issues.apache.org/jira/browse/CRUNCH-97
>             Project: Crunch
>          Issue Type: Improvement
>          Components: Core
>            Reporter: Josh Wills
>            Assignee: Josh Wills
>             Fix For: 0.5.0
>
>         Attachments: CRUNCH-97.patch, CRUNCH-97-take2.patch
>
>
> We should make it a bit easier to parse delimited text files into specific data types
(e.g., ints, floats, etc.) or combinations of types-- e.g., pairs of strings and ints, a Tuple3
of booleans, etc.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message