problem with "Input dataset" workflow control feature and custom non-subclass datatypes
Hi, There seems to be a weird bug with the "Input dataset" workflow control feature, hard to explain clearly but I'll try my best. If you define a custom datatype that is a simple subclass of an existing galaxy datatype, e.g.: <datatype extension="myext" type="galaxy.datatypes.data:Text" subclass="True" display_in_upload="true"/> And if this datatype will be the input to a workflow where you want to use the multiple input files feature you must put into your workflow editor an "Input dataset" box at the beginning and connect it. If you define a custom datatype that's it's own custom class, e.g.: <datatype extension="myext" type="galaxy.datatypes.data:MyExt" display_in_upload="true"/> with a simple class in lib/galaxy/datatypes/data.py e.g.: class MyExt( Data ): file_ext = "myext" And if this datatype will be the input data to a workflow if you have an "Input dataset" box at the beginning for some reason the drop-down menu (or multi-select) won't not have files of this type from your history it just ignores them. Now what is strange is if I edit the workflow and remove the beginning "Input dataset" box and start the workflow with just the first tool which has this custom datatype as an input parameter then when I try to run the workflow everything shows up properly :-/ Hope I explained this ok, seems like something is broken with the "Input dataset" workflow control feature. best, Leandro
Leandro, Thanks for reporting this, I'm able to reproduce it and will let you know when I have a fix. -Dannon On Jan 20, 2012, at 11:07 AM, Leandro Hermida wrote:
Hi,
There seems to be a weird bug with the "Input dataset" workflow control feature, hard to explain clearly but I'll try my best.
If you define a custom datatype that is a simple subclass of an existing galaxy datatype, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:Text" subclass="True" display_in_upload="true"/>
And if this datatype will be the input to a workflow where you want to use the multiple input files feature you must put into your workflow editor an "Input dataset" box at the beginning and connect it.
If you define a custom datatype that's it's own custom class, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:MyExt" display_in_upload="true"/>
with a simple class in lib/galaxy/datatypes/data.py e.g.:
class MyExt( Data ): file_ext = "myext"
And if this datatype will be the input data to a workflow if you have an "Input dataset" box at the beginning for some reason the drop-down menu (or multi-select) won't not have files of this type from your history it just ignores them. Now what is strange is if I edit the workflow and remove the beginning "Input dataset" box and start the workflow with just the first tool which has this custom datatype as an input parameter then when I try to run the workflow everything shows up properly :-/
Hope I explained this ok, seems like something is broken with the "Input dataset" workflow control feature.
best, Leandro ___________________________________________________________ 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:
Hi Dannon, Thanks I'm glad you could reproduce it! One thing I was thinking about since you are working in that code area is whether you could make it that the Input Dataset workflow control feature is not required. When you use this feature it cannot filter out the dataset history for a particular datatype, although when don't use it and start a workflow with a tool that has an input data parameter it is able to filter out for the desired datatype. It used to be in old galaxy-dist versions that if you didn't start with the Input Dataset control feature then you couldn't use the multiple input files workflow parallelization feature but now it doesn't seem to be the case which is very good. thanks, Leandro On Wed, Jan 25, 2012 at 2:55 PM, Dannon Baker <dannonbaker@me.com> wrote:
Leandro,
Thanks for reporting this, I'm able to reproduce it and will let you know when I have a fix.
-Dannon
On Jan 20, 2012, at 11:07 AM, Leandro Hermida wrote:
Hi,
There seems to be a weird bug with the "Input dataset" workflow control feature, hard to explain clearly but I'll try my best.
If you define a custom datatype that is a simple subclass of an existing galaxy datatype, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:Text" subclass="True" display_in_upload="true"/>
And if this datatype will be the input to a workflow where you want to use the multiple input files feature you must put into your workflow editor an "Input dataset" box at the beginning and connect it.
If you define a custom datatype that's it's own custom class, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:MyExt" display_in_upload="true"/>
with a simple class in lib/galaxy/datatypes/data.py e.g.:
class MyExt( Data ): file_ext = "myext"
And if this datatype will be the input data to a workflow if you have an "Input dataset" box at the beginning for some reason the drop-down menu (or multi-select) won't not have files of this type from your history it just ignores them. Now what is strange is if I edit the workflow and remove the beginning "Input dataset" box and start the workflow with just the first tool which has this custom datatype as an input parameter then when I try to run the workflow everything shows up properly :-/
Hope I explained this ok, seems like something is broken with the "Input dataset" workflow control feature.
best, Leandro ___________________________________________________________ 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:
Input Dataset datatype filtering actually used to work, this is a regression that happened in changeset 6329:ab90893a7cf5. I should have a fix for it today and you'll be back in business. -Dannon On Jan 25, 2012, at 10:53 AM, Leandro Hermida wrote:
Hi Dannon,
Thanks I'm glad you could reproduce it! One thing I was thinking about since you are working in that code area is whether you could make it that the Input Dataset workflow control feature is not required. When you use this feature it cannot filter out the dataset history for a particular datatype, although when don't use it and start a workflow with a tool that has an input data parameter it is able to filter out for the desired datatype. It used to be in old galaxy-dist versions that if you didn't start with the Input Dataset control feature then you couldn't use the multiple input files workflow parallelization feature but now it doesn't seem to be the case which is very good.
thanks, Leandro
On Wed, Jan 25, 2012 at 2:55 PM, Dannon Baker <dannonbaker@me.com> wrote: Leandro,
Thanks for reporting this, I'm able to reproduce it and will let you know when I have a fix.
-Dannon
On Jan 20, 2012, at 11:07 AM, Leandro Hermida wrote:
Hi,
There seems to be a weird bug with the "Input dataset" workflow control feature, hard to explain clearly but I'll try my best.
If you define a custom datatype that is a simple subclass of an existing galaxy datatype, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:Text" subclass="True" display_in_upload="true"/>
And if this datatype will be the input to a workflow where you want to use the multiple input files feature you must put into your workflow editor an "Input dataset" box at the beginning and connect it.
If you define a custom datatype that's it's own custom class, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:MyExt" display_in_upload="true"/>
with a simple class in lib/galaxy/datatypes/data.py e.g.:
class MyExt( Data ): file_ext = "myext"
And if this datatype will be the input data to a workflow if you have an "Input dataset" box at the beginning for some reason the drop-down menu (or multi-select) won't not have files of this type from your history it just ignores them. Now what is strange is if I edit the workflow and remove the beginning "Input dataset" box and start the workflow with just the first tool which has this custom datatype as an input parameter then when I try to run the workflow everything shows up properly :-/
Hope I explained this ok, seems like something is broken with the "Input dataset" workflow control feature.
best, Leandro ___________________________________________________________ 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:
Ok, should be fixed in 6620:c60760713d26. Let me know if you still have any issues. -Dannon On Jan 25, 2012, at 10:53 AM, Leandro Hermida wrote:
Hi Dannon,
Thanks I'm glad you could reproduce it! One thing I was thinking about since you are working in that code area is whether you could make it that the Input Dataset workflow control feature is not required. When you use this feature it cannot filter out the dataset history for a particular datatype, although when don't use it and start a workflow with a tool that has an input data parameter it is able to filter out for the desired datatype. It used to be in old galaxy-dist versions that if you didn't start with the Input Dataset control feature then you couldn't use the multiple input files workflow parallelization feature but now it doesn't seem to be the case which is very good.
thanks, Leandro
On Wed, Jan 25, 2012 at 2:55 PM, Dannon Baker <dannonbaker@me.com> wrote: Leandro,
Thanks for reporting this, I'm able to reproduce it and will let you know when I have a fix.
-Dannon
On Jan 20, 2012, at 11:07 AM, Leandro Hermida wrote:
Hi,
There seems to be a weird bug with the "Input dataset" workflow control feature, hard to explain clearly but I'll try my best.
If you define a custom datatype that is a simple subclass of an existing galaxy datatype, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:Text" subclass="True" display_in_upload="true"/>
And if this datatype will be the input to a workflow where you want to use the multiple input files feature you must put into your workflow editor an "Input dataset" box at the beginning and connect it.
If you define a custom datatype that's it's own custom class, e.g.:
<datatype extension="myext" type="galaxy.datatypes.data:MyExt" display_in_upload="true"/>
with a simple class in lib/galaxy/datatypes/data.py e.g.:
class MyExt( Data ): file_ext = "myext"
And if this datatype will be the input data to a workflow if you have an "Input dataset" box at the beginning for some reason the drop-down menu (or multi-select) won't not have files of this type from your history it just ignores them. Now what is strange is if I edit the workflow and remove the beginning "Input dataset" box and start the workflow with just the first tool which has this custom datatype as an input parameter then when I try to run the workflow everything shows up properly :-/
Hope I explained this ok, seems like something is broken with the "Input dataset" workflow control feature.
best, Leandro ___________________________________________________________ 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:
participants (2)
-
Dannon Baker
-
Leandro Hermida