def test_get_mime_type(filename, expected_mime_type):
with open(Path(__file__).parent.parent / 'sample_files' / filename, 'rb') as f:
assert get_mime_type(f) == expected_mime_type
def upload_document(service_id):
no_document_error = jsonify(error='No document upload'), 400
if request.is_json:
if 'document' not in request.json:
return no_document_error
raw_content = b64decode(request.json['document'])
if len(raw_content) > current_app.config['MAX_CONTENT_LENGTH']:
abort(413)
file_data = BytesIO(raw_content)
is_csv = request.json.get('is_csv', False)
else:
if 'document' not in request.files:
return no_document_error
file_data = request.files['document']
is_csv = False
if not isinstance(is_csv, bool):
return jsonify(error='Value for is_csv must be a boolean'), 400
mimetype = get_mime_type(file_data)
if mimetype not in current_app.config['ALLOWED_FILE_TYPES']:
allowed_file_types = ', '.join(
sorted({
f"'.{x}'"
for x in current_app.config['ALLOWED_FILE_TYPES'].values()
}))
return jsonify(
error=
f"Unsupported file type '{mimetype}'. Supported types are: {allowed_file_types}"
), 400
# Our mimetype auto-detection resolves CSV content as text/plain, so we use
# an explicit POST body parameter `is_csv` from the caller to resolve it as text/csv
if is_csv and mimetype == 'text/plain':
mimetype = 'text/csv'
if current_app.config['ANTIVIRUS_ENABLED']:
try:
virus_free = antivirus_client.scan(file_data)
except AntivirusError:
return jsonify(error='Antivirus API error'), 503
if not virus_free:
return jsonify(error="File did not pass the virus scan"), 400
document = document_store.put(service_id, file_data, mimetype=mimetype)
return jsonify(status='ok',
document={
'id':
document['id'],
'direct_file_url':
get_direct_file_url(
service_id=service_id,
document_id=document['id'],
key=document['encryption_key'],
mimetype=mimetype,
),
'url':
get_frontend_download_url(
service_id=service_id,
document_id=document['id'],
key=document['encryption_key'],
),
'mimetype':
mimetype,
}), 201
def test_get_mime_type(filename, expected_mime_type):
file = open(sample_files_path / filename, 'rb')
assert get_mime_type(file) == expected_mime_type
def test_get_mime_type_zip_xml(filename, expected_mime_type, mocker):
# libmagic on PaaS sometimes mistakes docx, xlsx, etc. files as ZIPs
mocker.patch('app.utils.magic.from_buffer', return_value='application/zip')
file = open(sample_files_path / filename, 'rb')
assert get_mime_type(file) == expected_mime_type
def upload_document(service_id):
if 'document' not in request.files:
return jsonify(error='No document upload'), 400
mimetype = get_mime_type(request.files['document'])
if not mime_type_is_allowed(mimetype, service_id):
return jsonify(
error="Unsupported document type '{}'. Supported types are: {}".
format(mimetype, current_app.config['ALLOWED_MIME_TYPES'])), 400
file_content = request.files['document'].read()
filename = request.form.get('filename')
file_extension = None
if filename and '.' in filename:
file_extension = ''.join(pathlib.Path(
filename.lower()).suffixes).lstrip('.')
# Our MIME type auto-detection resolves CSV content as text/plain,
# so we fix that if possible
if (filename or '').lower().endswith('.csv') and mimetype == 'text/plain':
mimetype = 'text/csv'
sending_method = request.form.get('sending_method')
if current_app.config["MLWR_HOST"]:
sid = upload_to_mlwr(file_content)
else:
sid = False
document = document_store.put(service_id,
file_content,
sending_method=sending_method,
mimetype=mimetype)
return jsonify(status='ok',
document={
'id':
document['id'],
'direct_file_url':
get_direct_file_url(
service_id=service_id,
document_id=document['id'],
key=document['encryption_key'],
sending_method=sending_method,
),
'url':
get_frontend_download_url(
service_id=service_id,
document_id=document['id'],
key=document['encryption_key'],
filename=filename,
),
'mlwr_sid':
sid,
'filename':
filename,
'sending_method':
sending_method,
'mime_type':
mimetype,
'file_size':
len(file_content),
'file_extension':
file_extension,
}), 201
