Hi John,

Thanks.  To elaborate a little more: we have a LabKey Server (web front end for a database) that manages raw files and metadata.  The idea is to make a galaxy tool where a user could do thing like query for all genomes from males patients, age x-y, etc.  I can use LabKey's APIs to return a list of those files.  However, the information at this point is text (basically a list of filepaths).  

It would be nice to automatically create datasets for those files.  Some of the time there's a simple 1:1 between file and the user-facing dataset (like images); however, for genomes we really want to make a paired collection.  BioBlend makes it relatively easy to go from filepaths to datasets; however, the authentication issue is what wasnt clear.

If there's a more standard path to go from list of files -> galaxy dataset I'm all ears.  

-Ben

On Wed, Apr 29, 2015 at 12:13 PM, John Chilton <jmchilton@gmail.com> wrote:
Unofficial way of doing this and the workaround of using configfiles
can be found in this thread:
http://dev.list.galaxyproject.org/Simple-standard-for-API-use-of-a-global-user-key-that-all-loaded-tools-can-draw-upon-td4665659.html.
There is a Trello card outlining platform work that should be done to
support this better but we have not made progress on that.

I would be interested in your use case. One can dynamically discover
datasets with various db keys and build collections composted of
dynamically discovered datasets and keys - but there is no way to use
the tool xml to dynamically discover a variable number of collections
- but it should be supported - as should building a list of paired or
lists of lists where the outer list describes the key and the inner
the sample.

-John



On Wed, Apr 29, 2015 at 2:58 PM, Ben Bimber <bbimber@gmail.com> wrote:
> Hello,
>
> I am new to galaxy. I"m trying to write a data input tool wrapper.  It will
> call a script that does a query, produces a list of files, and then the plan
> is to use the BioBlend API to create datasets/data collections in galaxy.
> In other words I am making the dataset(s) based on the contents of this
> file, rather than making a dataset from that file.  In my case this file has
> a pair of columns representing FASTQ files, and I want to create one paired
> dataset collection in galaxy for each genome.
>
> In theory using bioblend to create datasets is easy.  However, when I call
> this as a galaxy tool, I have not found a clean way to pass the credentials
> to bioblend.  BioBlend needs to know the serverURL and either the user's API
> key or username/password.  I think if I poke around $__app__ or $__user__
> there's a good change I will find a property with the information I need;
> however, this has the obvious problem of writing the API key to the log.
>
> Is there another way to approach this problem?
>
> Thanks in advance for any help,
> Ben
>
>
> ___________________________________________________________
> Please keep all replies on the list by using "reply all"
> in your mail client.  To manage your subscriptions to this
> and other Galaxy lists, please use the interface at:
>   https://lists.galaxyproject.org/
>
> To search Galaxy mailing lists use the unified search at:
>   http://galaxyproject.org/search/mailinglists/