def tag_using_seperator(self, tag: str, separator: str,
value_position: int):
"""Tag an experiment raw data using file name and separator
Parameters
----------
tag
The name (or key) of the tag to add to the data
separator
The character used as a separator in the filename (ex: _)
value_position
Position of the value to extract with respect to the separators
"""
self.set_tag(tag, False)
_rawdataset = RawDataSet(self.metadata.rawdataset)
for i in range(_rawdataset.size()):
_rawdata = _rawdataset.get(i)
basename = os.path.splitext(os.path.basename(
_rawdata.metadata.uri))[0]
splited_name = basename.split(separator)
value = ''
if len(splited_name) > value_position:
value = splited_name[value_position]
_rawdata.set_tag(tag, value)
def set_tag(self, tag: str, add_to_data: bool = True):
"""Add a tag key to the experiment
It add the tag key (if not already exists) to the
experiment metadata and also add a tag key to all
the images
Parameters
----------
tag
Tag key to be added
add_to_data
if True add an empty tag to all the data in
the RawDataSet
"""
# add the tag to the experiment
if tag not in self.metadata.tags:
self.metadata.tags.append(tag)
self.write()
if add_to_data:
raw_dataset = RawDataSet(self.metadata.rawdataset)
for i in range(int(raw_dataset.size())):
raw_data = raw_dataset.get(i)
if tag not in raw_data.metadata.tags:
raw_data.set_tag(tag, '')
def test_write_rawdataset(self):
raw_data = RawDataSet(self.tst_dataset_file)
raw_data.metadata = create_dataset()
raw_data.write()
self.assertTrue(
filecmp.cmp(self.tst_dataset_file,
self.ref_dataset_file,
shallow=False))
def tag_from_name(self, tag: str, values: list):
"""Tag an experiment raw data using file name
Parameters
----------
tag
The name (or key) of the tag to add to the data
values
List of possible values for the tag to find in the filename
"""
self.set_tag(tag, False)
_rawdataset = RawDataSet(self.metadata.rawdataset)
for i in range(_rawdataset.size()):
_rawdata = _rawdataset.get(i)
for value in values:
if value in _rawdata.metadata.name:
_rawdata.set_tag(tag, value)
break
def get_data(
self, dataset_name: str, query: str, origin_output_name: str = ''
) -> list:
"""query a specific dataset of an experiment
In this version only AND queries are supported
(ex: tag1=value1 AND tag2=value2)
and performed on the data set named dataset
Parameters
----------
dataset_name
Name of the dataset to query
query
String query with the key=value format.
origin_output_name
Name of the ouput origin (ex: -o) in the case of ProcessedDataset
search
Returns
-------
list
List of selected data (md.json files urls are returned)
"""
# search the dataset
raw_dataset = RawDataSet(self.metadata.rawdataset)
if raw_dataset.metadata.name == dataset_name:
return raw_dataset.get_data(query)
else:
for i in range(len(self.metadata.processeddatasets)):
processeddataset = ProcessedDataSet(
self.metadata.processeddatasets[i])
if processeddataset.metadata.name == dataset_name:
return processeddataset.get_data(query, origin_output_name)
raise MetadataQueryError('Query dataset ', dataset_name, ' not found')
def get_dataset(self, name: str):
"""Get the metadata of a dataset
Returns a RawDataset or a ProcessedDataSet
Parameters
----------
name
Name of the dataset
"""
if name == 'data':
return RawDataSet(self.metadata.rawdataset)
else:
for dataset_name in self.metadata.processeddatasets:
pdataset = ProcessedDataSet(dataset_name)
if pdataset.metadata.name == name:
return pdataset
return None
def test_read_rawdataset(self):
raw_dataset_read = RawDataSet(self.ref_dataset_file)
raw_data_ref_metadata = create_dataset()
self.assertEqual(raw_dataset_read.metadata.serialize(),
raw_data_ref_metadata.serialize())