def get_workflow_inputs(l, pkd, gi, git, history_name, library):
# LibraryDatasetDatasetAssociation (ldda), LibraryDataset (ld), HistoryDatasetAssociation (hda),
# or HistoryDatasetCollectionAssociation (hdca).
st = get_time_stamp()
hc = HistoryClient(gi)
worklow_inputs_d = {}
for table, filter, dinput_name, dinput_step, dinput_type in l:
pks = pkd[str(table.prefix)]
# will get multiple inputs here because we can multiple galaxyfilelinks per file. They are all the same
# file so we can just get unique
selected_objects = GenericFile.objects.filter(pk__in=pks).distinct()
print('PKS', pks, dinput_type)
print(selected_objects)
if dinput_type == 'data_input':
# can only use the first selection (need to use data collection for multiple files, currently this
# approach doesn't support using 'multiple files' as input as not possible with BioBlend (i think)
s = selected_objects[0]
gid = s.galaxyfilelink_set.filter(
galaxy_library=library)[0].galaxy_id
print(gid)
worklow_inputs_d[dinput_step] = {'id': gid, 'src': 'ld'}
elif dinput_type == 'data_collection_input':
element_identifiers = []
hist = hc.create_history('{}-(data-history-{})-{}'.format(
history_name, dinput_name, st))
for s in selected_objects:
print(s)
gfl = s.galaxyfilelink_set.filter(galaxy_library=library)[0]
if library:
dataset = hc.upload_dataset_from_library(
hist['id'], lib_dataset_id=gfl.galaxy_id)
element_identifiers.append({
'id':
dataset['id'],
'name':
os.path.basename(dataset['file_name']),
'src':
'hda'
})
else:
element_identifiers.append({
'id':
gfl.galaxy_id,
'name':
gfl.genericfile.data_file.name,
'src':
'hda'
})
c_descript = {
'collection_type': 'list',
'element_identifiers': element_identifiers,
'name': dinput_name,
}
dc = hc.create_dataset_collection(hist['id'], c_descript)
worklow_inputs_d[dinput_step] = {'id': dc['id'], 'src': 'hdca'}
return worklow_inputs_d
print("ready files: {}".format(ready))
old_ready = ready
ready = 0
for f in fc.show_folder(folder["id"], contents=True)["folder_contents"]:
if f["state"] == "ok":
ready = ready + 1
sleep(5)
print("...")
print("All {} datasets ready!".format(ready))
# add files to history
history = hc.create_history("{}".format(now_string))
print(history)
# create dataset collection
collection_description = {
'collection_type': 'list',
'element_identifiers': [],
'name': 'manifest collection'
}
for f in files:
element_identifier = {
'id': f["id"],
'name': f["name"],
'src': 'ldda'}
collection_description["element_identifiers"].append(element_identifier)
print(collection_description)
hc.create_dataset_collection(history["id"], collection_description)