parser.add_argument(
"--error-on-duplicate-strains",
action="store_true",
help=
"exit with an error if any duplicate strains are found. By default, duplicates are resolved by preferring most recent accession id or the last record."
)
parser.add_argument("--output", required=True, help="sanitized metadata")
args = parser.parse_args()
# If the input is a tarball, try to find a metadata file inside the archive.
metadata_file = args.metadata
tar_handle = None
if ".tar" in Path(args.metadata).suffixes:
try:
metadata_file, tar_handle = extract_tar_file_contents(
args.metadata, "metadata")
except FileNotFoundError as error:
print(f"ERROR: {error}", file=sys.stderr)
sys.exit(1)
# Read metadata with pandas because Augur's read_metadata utility does not
# support metadata without a "strain" or "name" field.
metadata = pd.read_csv(metadata_file,
sep=None,
engine="python",
skipinitialspace=True,
dtype={
"strain": "string",
"name": "string",
}).fillna("")
parser.add_argument("--sequences", nargs="+", required=True, help="sequences to be sanitized")
parser.add_argument("--strip-prefixes", nargs="+", help="prefixes to strip from strain names in the sequences")
parser.add_argument('--error-on-duplicate-strains', action="store_true", help="exit with an error when the same strain is detected multiple times with different sequences. By default, use the first occurrence of each duplicated sequence.")
parser.add_argument("--output", required=True, help="sanitized sequences")
args = parser.parse_args()
sequence_files = []
tar_handles = []
for sequence_filename in args.sequences:
# If the input is a tarball, try to find a sequence file inside the
# archive.
if ".tar" in Path(sequence_filename).suffixes:
try:
sequence_file, tar_handle = extract_tar_file_contents(
sequence_filename,
"sequences"
)
# The extracted tar file is an io.BufferedReader which provides
# a binary stream. BioPython's SeqIO reader expects a text
# stream. We decode each line of the BufferedReader in a
# generator such that decoding happens on the fly per line
# downstream. For more details see:
# https://docs.python.org/3/library/tarfile.html#tarfile.TarFile.extractfile
sequence_files.append(sequence_file)
tar_handles.append(tar_handle)
except FileNotFoundError as error:
print(f"ERROR: {error}", file=sys.stderr)
sys.exit(1)
else:
sequence_files.append(sequence_filename)
parser.add_argument("--output", required=True, help="sanitized metadata")
args = parser.parse_args()
# Get user-defined metadata id columns to look for.
metadata_id_columns = args.metadata_id_columns
# Get user-defined database id columns to use for deduplication.
database_id_columns = args.database_id_columns
# If the input is a tarball, try to find a metadata file inside the archive.
metadata_file = args.metadata
metadata_is_temporary = False
if ".tar" in Path(args.metadata).suffixes:
try:
temporary_dir, metadata_file = extract_tar_file_contents(
args.metadata, "metadata")
metadata_is_temporary = True
except FileNotFoundError as error:
print(f"ERROR: {error}", file=sys.stderr)
sys.exit(1)
# In the first pass through the metadata, map strain names to database ids.
# We will use this mapping to deduplicate records in the second pass.
# Additionally, this pass checks for missing id columns and the presence of
# any duplicate records, in case the user has requested an error on
# duplicates.
try:
database_ids_by_strain = get_database_ids_by_strain(
metadata_file,
metadata_id_columns,
database_id_columns,