In the example they provide it to the sh tool:

<command>

example_tool.sh $input1 $output1 $output1.id $__new_file_path__

</command>



...so then the 3rd and 4th variables given to example_tool.sh
contain the <id> and the output directory path. Save all the
extra output files to that directory with the exact format:
primary_<id>_somename_visible_filetype (eg
primary_546_output2_visible_bed)



Stef



----- Original message -----

From: Martin Christiansen <[1]martinchristianse...@hotmail.com>

To: Stef van Lieshout <[2]stefvanliesh...@fastmail.fm>,
"[3]galaxy-dev@lists.bx.psu.edu"
<[4]galaxy-dev@lists.bx.psu.edu>

Subject: RE: [galaxy-dev] Multiple output datasets

Date: Wed, 27 Aug 2014 14:52:03 +0200



I went back to not using the options in the .ini file, but it
still doesn't work.
How to you specify $__new_file_path__? Or do you just enter the
path in the xml command line?

Martin
  __________________________________________________________

From: stefvanliesh...@fastmail.fm
To: martinchristianse...@hotmail.com;
galaxy-dev@lists.bx.psu.edu
Subject: Re: [galaxy-dev] Multiple output datasets
Date: Wed, 27 Aug 2014 14:17:13 +0200

I do not use "new_file_path" and "collect_outputs_from" in
universe_wsgi.ini

In my tool code I directly save to files to:
$__new_file_path__/primary_<id>_somename_visible_filetype

As they say on the page:
To use this method at least one output dataset (output1 in our
example) must be specified - this dataset will act as the
reference for our additional datasets. The new additional
datasets must be placed in the directory provided by
$__new_file_path__ and named in the form: "%s_%s_%s_%s_%s_%s" %
( 'primary', output1.id, name, 'visible', file_type,
dbkey(build) ) Where the '_dbkey' portion of the name is
optional.

Stef


----- Original message -----
From: Martin Christiansen <[5]martinchristianse...@hotmail.com>
To: Stef van Lieshout <[6]stefvanliesh...@fastmail.fm>,
"[7]galaxy-dev@lists.bx.psu.edu"
<[8]galaxy-dev@lists.bx.psu.edu>
Subject: RE: [galaxy-dev] Multiple output datasets
Date: Wed, 27 Aug 2014 14:01:03 +0200

Hi Stef,

I always restart galaxy and have done the other parts correct
as well, but with no luck.

In my universe_wsgi.ini file I have:

new_file_path = my/file/path
collect_outputs_from = new_file_path, job_working_directory

Does this look familiar?

Martin
  __________________________________________________________

From: stefvanliesh...@fastmail.fm
To: galaxy-dev@lists.bx.psu.edu
Date: Wed, 27 Aug 2014 12:22:35 +0200
Subject: Re: [galaxy-dev] Multiple output datasets

It does works for me if I do exactly as described on that page
without using the universe_wsgi.ini option:

- I have the "force_history_refresh"
- I follow the exact file-name structure
(primary_<id>_somename_visible_filetype)
- I save the files in the dir from passed param
$__new_file_path__

Can you check if that works? Maybe the change in
universe_wsgi.ini isn't picked up correctly... Did you restart
galaxy?

Stef


----- Original message -----
From: Martin Christiansen <[9]martinchristianse...@hotmail.com>
To: "[10]galaxy-dev@lists.bx.psu.edu"
<[11]galaxy-dev@lists.bx.psu.edu>
Subject: [galaxy-dev] Multiple output datasets
Date: Wed, 27 Aug 2014 12:04:46 +0200

Hi all,

I'm implementing a tool for our local galaxy instance which has
multiple output datasets.
The number of datasets are determined by the input so I've been
following the instructions at the bottom of this page:
[12]https://wiki.galaxyproject.org/Admin/Tools/Multiple%20Outpu
t%20Files.

As I understand, the $__new_file_path__ option is changed in
theuniverse_wsgi.ini file.
Regardless, the extra files created from the tool does not show
in the galaxy history.

Any ideas?

Best,
Martin Christiansen
___________________________________________________________
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:
[13]http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:
[14]http://galaxyproject.org/search/mailinglists/

___________________________________________________________
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:
http://lists.bx.psu.edu/ To search Galaxy mailing lists use the
unified search at:
http://galaxyproject.org/search/mailinglists/

References

1. mailto:martinchristianse...@hotmail.com
2. mailto:stefvanliesh...@fastmail.fm
3. mailto:galaxy-dev@lists.bx.psu.edu
4. mailto:galaxy-dev@lists.bx.psu.edu
5. mailto:martinchristianse...@hotmail.com
6. mailto:stefvanliesh...@fastmail.fm
7. mailto:galaxy-dev@lists.bx.psu.edu
8. mailto:galaxy-dev@lists.bx.psu.edu
9. mailto:martinchristianse...@hotmail.com
  10. mailto:galaxy-dev@lists.bx.psu.edu
  11. mailto:galaxy-dev@lists.bx.psu.edu
  12. https://wiki.galaxyproject.org/Admin/Tools/Multiple%20Output%20Files.
  13. http://lists.bx.psu.edu/
  14. http://galaxyproject.org/search/mailinglists/
___________________________________________________________
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:
  http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:
  http://galaxyproject.org/search/mailinglists/

Reply via email to