I tried out for a couple of files, and the metdump alias is not dumping out
any metadata for the respective ProductIds.

I also tried dumping out metadata for other files for which I've not
specified metout-configs, and they seem to be empty too.
The fmquery works fine and it dumps out all the ProductIds.

Thanks,

On Thu, Sep 15, 2011 at 9:08 PM, Mattmann, Chris A (388J) <
[email protected]> wrote:

> Hey Sheryl,
>
> Are you sure that RecordID is being recorded for the particular products
> that you are trying to look up?
>
> One way to find this out:
>
> 1. get the product IDs back for the products you are interested in. One way
> to do this would be
> to remove RecordID from your query as you mention below and then add
> CAS.ProductId to the
> SELECT list, and to change FORMAT='$FileLocation...' to
> FORMAT='$CAS.ProductId'.
>
> 2. Then, for each of the product IDs, check out OODT-306 [1] and use the
> metdump alias to dump
> out the metadata for each product ID to the current working directory.
> Then, inspect that .met file.
> Is RecordID recorded?
>
> Let's start there.
>
> Cheers,
> Chris
>
> [1] http://issues.apache.org/jira/browse/OODT-306
>
> On Sep 15, 2011, at 8:53 PM, Sheryl John wrote:
>
> > Hi,
> >
> > I am not clear about the metadata in .met files generated by the
> met-config.xml files. Particularly, where do they fall in the product-type
> and element mappings?
> >
> > Because in my query, SQL(FORMAT='$FileLocation/$Filename'){ SELECT
> FileLocation,Filename,ISMTable,RecordID FROM ISMRawData WHERE ISMTable =
> 'Chartevents'  AND RecordID = '[PID]' ,  I'm selecting elements that are
> mapped to the product-type, ISMRawData and elements that are mapped to the
> GenericFile product-type.
> >
> >  Except RecordID, which was defined as a key in the metout-config file
> and I'm assuming that the parsing error is because the RecordID is not
> mapped to any product-type. This query works and pulls all the files I need
> when I remove RecordID from the query. Again, since I'm not clear about
> metadata mapping , it's highly likely that I'm wrong about this.
> >
> > Thanks,
> > On Thu, Sep 15, 2011 at 11:17 AM, holenoter <[email protected]> wrote:
> > hey Sheryl,
> >
> > I think there is a misunderstanding on how the met-config.xml files are
> used... the metadata in these files are not added to your workflow context
> metadata... they are only used to create the *.met files for the data file
> for filemgr ingest... about your query I'm not quite clear on how you are
> wanting go use it... it kinda seems like you want to dynamically replace out
> RecordIO... in which case you're gonna need it to look more like RecordIO =
> '[PID]'
> >
> > -brian
> >
> >
> > On Sep 14, 2011, at 09:18 PM, Sheryl John <[email protected]> wrote:
> >
> >> Hi Chris,
> >>
> >> Thanks for pointing that out. I was working on a different query and
> similar to the one I described. Though I've used only single quotes now, I'm
> still getting the same parsing error.
> >>
> >> But, I think what I intended to do with context metadata ( for my
> workflow) doesn't make sense for a single file product and rather, would
> apply for a group of files.
> >> I guess I have revise my workflow logic and metadata use before getting
> back to this query.
> >>
> >> Thanks!
> >>
> >>
> >>
> >> On Wed, Sep 14, 2011 at 8:42 PM, Mattmann, Chris A (388J) <
> [email protected]> wrote:
> >> Hi Sheryl,
> >>
> >> I notice you're using double quotes on the RecordID = "PID", yet you are
> using single quotes on ISMTable = 'Chartevents'.
> >>
> >> Try using single quotes on both and see if that fixes it.
> >>
> >> Thanks!
> >>
> >> Cheers,
> >> Chris
> >>
> >>
> >> On Sep 14, 2011, at 7:03 PM, Sheryl John wrote:
> >>
> >> > Hi,
> >> >
> >> > I have defined some key-val pairs for a file( say Output.csv) in a
> metout-config.xml for my PGETask Workflow. However, after executing the
> workflow, the met-config.xml is not creating a Output.csv.cas file.
> >> >
> >> > I want to be able to use the above keys/metadata later on in an
> SQL-like query from the pgeconfig file. For example, if I've defined
> 'RecordID' as a key in the metout-config.xml, I would want to use this
> metadata in the following query:
> >> >
> >> > SQL(FORMAT='$FileLocation/$Filename'){ SELECT
> FileLocation,Filename,ISMTable,RecordID FROM ISMRawData WHERE ISMTable =
> 'Chartevents'  AND RecordID = "PID"}  The others keys included in the query
> above are elements and product-types that were defined during ingestion in
> the File Manager.
> >> >
> >> > At present, the task fails to parse the above query when I run the
> workflow. Is this because metout-config is not creating the Output.cas file?
>  And, what is the best way to specify metadata files for a group of files or
> for a folder?
> >> >
> >> > Thanks,
> >> > Sheryl
> >>
> >>
> >> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >> Chris Mattmann, Ph.D.
> >> Senior Computer Scientist
> >> NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
> >> Office: 171-266B, Mailstop: 171-246
> >> Email: chris.a.mattmann@nasagov
> >>
> >> WWW:   http://sunset.usc.edu/~mattmann/
> >> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >> Adjunct Assistant Professor, Computer Science Department
> >> University of Southern California, Los Angeles, CA 90089 USA
> >> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >>
> >>
> >>
> >>
> >> --
> >> -Sheryl
> >
> >
> >
> > --
> > -Sheryl
>
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Chris Mattmann, Ph.D.
> Senior Computer Scientist
> NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
> Office: 171-266B, Mailstop: 171-246
> Email: [email protected]
> WWW:   http://sunset.usc.edu/~mattmann/
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Adjunct Assistant Professor, Computer Science Department
> University of Southern California, Los Angeles, CA 90089 USA
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
>


-- 
-Sheryl

Reply via email to