def validate_source(self, source):
"""
Does the given source have both the methods `seek` and `read`?
:raises InvalidDataProviderSource: if not.
"""
source = super(ChunkDataProvider, self).validate_source(source)
if ((not hasattr(source, 'seek')) or (not hasattr(source, 'read'))):
raise exceptions.InvalidDataProviderSource(source)
return source
def validate_source( self, source ):
"""
Is this a valid source for this provider?
:raises InvalidDataProviderSource: if the source is considered invalid.
Meant to be overridden in subclasses.
"""
if not source or not hasattr( source, '__iter__' ):
# that's by no means a thorough check
raise exceptions.InvalidDataProviderSource( source )
return source
def __init__(self, dataset, **kwargs):
"""
:param dataset: the Galaxy dataset whose file will be the source
:type dataset: model.DatasetInstance
:raises exceptions.InvalidDataProviderSource: if not a DatsetInstance
"""
if not isinstance(dataset, galaxy.model.DatasetInstance):
raise exceptions.InvalidDataProviderSource(
"Data provider can only be used with a DatasetInstance")
self.dataset = dataset
# this dataset file is obviously the source
#TODO: this might be a good place to interface with the object_store...
super(DatasetDataProvider, self).__init__(open(dataset.file_name,
'rb'))
def __init__(self,
dataset,
options_string='',
options_dict=None,
regions=None,
**kwargs):
"""
:param options_string: samtools options in string form (flags separated
by spaces)
Optional: defaults to ''
:type options_string: str
:param options_dict: dictionary of samtools options
Optional: defaults to None
:type options_dict: dict or None
:param regions: list of samtools regions strings
Optional: defaults to None
:type regions: list of str or None
"""
#TODO: into validate_source
#TODO: have to import these here due to circular ref in model/datatypes
import galaxy.datatypes.binary
import galaxy.datatypes.tabular
if (not (isinstance(dataset.datatype, galaxy.datatypes.tabular.Sam) or
isinstance(dataset.datatype, galaxy.datatypes.binary.Bam))):
raise exceptions.InvalidDataProviderSource(
'dataset must be a Sam or Bam datatype: %s' %
(str(dataset.datatype)))
self.dataset = dataset
options_dict = options_dict or {}
# ensure regions are strings
regions = [str(r) for r in regions] if regions else []
#TODO: view only for now
#TODO: not properly using overriding super's validate_opts, command here
subcommand = 'view'
#TODO:?? do we need a path to samtools?
subproc_args = self.build_command_list(subcommand, options_string,
options_dict, regions)
#TODO: the composition/inheritance here doesn't make a lot sense
subproc_provider = external.SubprocessDataProvider(*subproc_args)
super(SamtoolsDataProvider, self).__init__(subproc_provider, **kwargs)