def generate_export_file(project, tasks, output_format, get_args):
# prepare for saving
now = datetime.now()
data = json.dumps(tasks, ensure_ascii=False)
md5 = hashlib.md5(json.dumps(data).encode('utf-8')).hexdigest()
name = 'project-' + str(project.title) + '-at-' + now.strftime(
'%Y-%m-%d-%H-%M') + f'-{md5[0:8]}'
input_json = DataExport.save_export_files(project, now, get_args, data,
md5, name)
converter = Converter(config=project.get_parsed_config(),
project_dir=None,
upload_dir=os.path.join(settings.MEDIA_ROOT,
settings.UPLOAD_DIR))
with get_temp_dir() as tmp_dir:
converter.convert(input_json, tmp_dir, output_format, is_dir=False)
files = get_all_files_from_dir(tmp_dir)
# if only one file is exported - no need to create archive
if len(files) == 1:
output_file = files[0]
ext = os.path.splitext(output_file)[-1]
content_type = f'application/{ext}'
out = read_bytes_stream(output_file)
filename = name + os.path.splitext(output_file)[-1]
return out, content_type, filename
# otherwise pack output directory into archive
shutil.make_archive(tmp_dir, 'zip', tmp_dir)
out = read_bytes_stream(os.path.abspath(tmp_dir + '.zip'))
content_type = 'application/zip'
filename = name + '.zip'
return out, content_type, filename
def convert_file(self, to):
with get_temp_dir() as tmp_dir:
converter = Converter(
config=self.project.get_parsed_config(),
project_dir=None,
upload_dir=tmp_dir,
# download_resources=download_resources,
)
input_name = pathlib.Path(self.file.name).name
input_file_path = pathlib.Path(tmp_dir) / input_name
with open(input_file_path, 'wb') as out_file:
out_file.write(self.file.open().read())
converter.convert(input_file_path, tmp_dir, to, is_dir=False)
files = get_all_files_from_dir(tmp_dir)
output_file = [
file_name for file_name in files
if pathlib.Path(file_name).name != input_name
][0]
out = read_bytes_stream(output_file)
filename = pathlib.Path(input_name).stem + pathlib.Path(
output_file).suffix
return File(
out,
name=filename,
)
def get_export_formats(project):
converter = Converter(config=project.get_parsed_config(),
project_dir=None)
formats = []
supported_formats = set(converter.supported_formats)
for format, format_info in converter.all_formats().items():
format_info['name'] = format.name
if format.name not in supported_formats:
format_info['disabled'] = True
formats.append(format_info)
return sorted(formats, key=lambda f: f.get('disabled', False))
def reload(self):
self.tasks = None
self.derived_input_schema = []
self.derived_output_schema = {
'from_name_to_name_type': set(),
'labels': defaultdict(set)
}
self._init()
self.label_config_full = config_comments_free(
open(self.config['label_config']).read())
self.label_config_line = config_line_stripped(self.label_config_full)
if self.analytics is None:
self.analytics = Analytics(
self.label_config_line,
self.config.get('collect_analytics', True), self.name)
else:
self.analytics.update_info(
self.label_config_line,
self.config.get('collect_analytics', True), self.name)
# configure project
self.project_obj = ProjectObj(label_config=self.label_config_line,
label_config_full=self.label_config_full)
# configure machine learning backend
if self.ml_backend is None:
ml_backend_params = self.config.get('ml_backend')
if ml_backend_params:
ml_backend = MLBackend.from_params(ml_backend_params)
self.project_obj.connect(ml_backend)
self.converter = Converter(self.label_config_full)
def export_projects():
"""Exports labelled data for all export enabled projects using Label Studio converter"""
# Get all projects
projects = Project.objects.all()
for project in projects:
if project.status == Project.Status.ACTIVE and project.export_format != Project.ExportFormat.NONE:
logger.info(f"Exporting project {project.name}")
output_paths = []
project_annotators = ProjectAnnotators.objects.filter(
project=project)
for project_annotator in project_annotators:
annotator = project_annotator.annotator
annotator_dir = LABELIT_DIRS[
'projects'] / annotator.username / project.name
label_config_file = annotator_dir / 'config.xml'
if label_config_file.exists():
c = Converter(str(label_config_file))
completions_dir = annotator_dir / 'completions/'
output_path = LABELIT_DIRS[
'exports'] / project.name / annotator.username
logger.debug(
f"Exporting completions for annotator {annotator.username}, project {project.name}"
)
if project.export_format == Project.ExportFormat.JSON:
c.convert_to_json(completions_dir, output_path)
elif project.export_format == Project.ExportFormat.CSV:
c.convert_to_csv(completions_dir, output_path, sep=',')
elif project.export_format == Project.ExportFormat.TSV:
c.convert_to_csv(completions_dir,
output_path,
sep='\t')
elif project.export_format == Project.ExportFormat.CONLL:
c.convert_to_conll2003(completions_dir, output_path)
else:
logger.debug(
f"Export format {project.export_format} not supported for project {project.name}"
)
continue
output_paths.append(output_path)
def convert_file(self, to_format):
with get_temp_dir() as tmp_dir:
OUT = 'out'
out_dir = pathlib.Path(tmp_dir) / OUT
out_dir.mkdir(mode=0o700, parents=True, exist_ok=True)
converter = Converter(
config=self.project.get_parsed_config(),
project_dir=None,
upload_dir=out_dir,
download_resources=False,
)
input_name = pathlib.Path(self.file.name).name
input_file_path = pathlib.Path(tmp_dir) / input_name
with open(input_file_path, 'wb') as file_:
file_.write(self.file.open().read())
converter.convert(input_file_path, out_dir, to_format, is_dir=False)
files = get_all_files_from_dir(out_dir)
dirs = get_all_dirs_from_dir(out_dir)
if len(files) == 0 and len(dirs) == 0:
return None
elif len(files) == 1 and len(dirs) == 0:
output_file = files[0]
filename = pathlib.Path(input_name).stem + pathlib.Path(output_file).suffix
else:
shutil.make_archive(out_dir, 'zip', out_dir)
output_file = pathlib.Path(tmp_dir) / (str(out_dir.stem) + '.zip')
filename = pathlib.Path(input_name).stem + '.zip'
out = read_bytes_stream(output_file)
return File(
out,
name=filename,
)
def load_converter(self):
self.converter = Converter(self.parsed_label_config)
def load_converter(self):
self.converter = Converter(self.label_config_full)
def load_converter(self):
self.converter = Converter(self.parsed_label_config, project_dir=self.path)
