[
https://issues.apache.org/jira/browse/BLUR-304?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13845012#comment-13845012
]
Aaron McCurry commented on BLUR-304:
------------------------------------
All this sounds good, where would you suggest adding QueryWriter code? As a
standalone chunk of code? Something built into the servers? Something that
would be turned on and off? Or on all the time? Perhaps turned on as a part
of a trace?
The rest of this all sounds like a great idea and it very doable.
Aaron
> Add a Run or Execute command to the shell
> -----------------------------------------
>
> Key: BLUR-304
> URL: https://issues.apache.org/jira/browse/BLUR-304
> Project: Apache Blur
> Issue Type: New Feature
> Reporter: Tim Williams
> Priority: Minor
>
> It'd be helpful to be able to execute a query from a file in the shell. For
> example:
> run mytable /path/to/query.json
> I think we could also provide a simple utility that does the writing, to make
> it easy to write to a logger... e.g.
> QueryWriter.toJsonString(BlurQuery q) {
> ...
> TProtocol p = new TJSONProtocol(new TIOStreamTransport(...));
> q.write(p);
> ...
> }
> Then the shell can just inflate the object and run the query, e.g.
> BIS in = new BufferedInputStream(new FIS(query.json));
> TProtocol p = new TJSONProtocol(new TIOStreamTransport(in));
> TBase<BlurQuery, BlurQuery._Fields> tbase = new BlurQuery();
> tbase.read(p);
> It'd be helpful for debugging large queries with unique characteristics (e.g.
> loads of facets).
--
This message was sent by Atlassian JIRA
(v6.1.4#6159)
