You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@giraph.apache.org by "Jakob Homan (JIRA)" <ji...@apache.org> on 2012/07/14 01:58:34 UTC
[jira] [Created] (GIRAPH-253) Create printf/pretty-printer style
output format
Jakob Homan created GIRAPH-253:
----------------------------------
Summary: Create printf/pretty-printer style output format
Key: GIRAPH-253
URL: https://issues.apache.org/jira/browse/GIRAPH-253
Project: Giraph
Issue Type: Bug
Reporter: Jakob Homan
It would be nice to have an outputformat where the user could specify (from the command-line in bin/giraph) the output of each vertex. Something like:
{noformat}--pretty-print="Hello, my name is %id and my value is %value. I have %edges.count edges"{noformat}
This would result in text files, each line of which would correspond to a vertex.
This would make it very easy for people to spit out different output formats without having to actually write any code...
Bonus points: only print the line if some predicate holds true --pretty-print-if="edges.count>10"
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (GIRAPH-253) Create printf/pretty-printer style
output format
Posted by "Jakob Homan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/GIRAPH-253?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13414188#comment-13414188 ]
Jakob Homan commented on GIRAPH-253:
------------------------------------
alternatively, it may just be easier to turn this into a streaming-type of approach, where one could specify a shell script to pass the vertex to and that returns the line to be output.
> Create printf/pretty-printer style output format
> ------------------------------------------------
>
> Key: GIRAPH-253
> URL: https://issues.apache.org/jira/browse/GIRAPH-253
> Project: Giraph
> Issue Type: Bug
> Reporter: Jakob Homan
>
> It would be nice to have an outputformat where the user could specify (from the command-line in bin/giraph) the output of each vertex. Something like:
> {noformat}--pretty-print="Hello, my name is %id and my value is %value. I have %edges.count edges"{noformat}
> This would result in text files, each line of which would correspond to a vertex.
> This would make it very easy for people to spit out different output formats without having to actually write any code...
> Bonus points: only print the line if some predicate holds true --pretty-print-if="edges.count>10"
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (GIRAPH-253) Create printf/pretty-printer style
output format
Posted by "Eli Reisman (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/GIRAPH-253?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13414540#comment-13414540 ]
Eli Reisman commented on GIRAPH-253:
------------------------------------
What about a PrintfOutputFormat which is specified at the command line like all the others, using the -of switch, and expecting to pick up a -D option from the command line whenever it is used that could be specified in GiraphJob with the other Configuration constants? then you could do
bin/giraph ... -Dgiraph.printf="my vertex id is %ID and value is %VALUE" ... -of PrintfOutputFormat ...
this way if you get lazy and forget to remove the -D from your command line (or want to go from testing output to real output and back and forth between runs quickly) the -D part only gets picked up in the application if you specified -of as PrintfOutputFormat otherwise is ignored harmlessly.
> Create printf/pretty-printer style output format
> ------------------------------------------------
>
> Key: GIRAPH-253
> URL: https://issues.apache.org/jira/browse/GIRAPH-253
> Project: Giraph
> Issue Type: Bug
> Reporter: Jakob Homan
>
> It would be nice to have an outputformat where the user could specify (from the command-line in bin/giraph) the output of each vertex. Something like:
> {noformat}--pretty-print="Hello, my name is %id and my value is %value. I have %edges.count edges"{noformat}
> This would result in text files, each line of which would correspond to a vertex.
> This would make it very easy for people to spit out different output formats without having to actually write any code...
> Bonus points: only print the line if some predicate holds true --pretty-print-if="edges.count>10"
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira