- Skip Building MAC tables

Author: pbashyal-nmdp

pyard/data_repository.py

@@ -486,7 +486,9 @@ def generate_short_nulls(db_connection, who_group):
     return shortnulls
 
 
-def generate_mac_codes(db_connection: sqlite3.Connection, refresh_mac: bool):
+def generate_mac_codes(
+    db_connection: sqlite3.Connection, refresh_mac: bool = False, load_mac: bool = True
+):
     """
     MAC files come in 2 different versions:
 
@@ -530,29 +532,31 @@ def generate_mac_codes(db_connection: sqlite3.Connection, refresh_mac: bool):
 
     :param db_connection: Database connection to the sqlite database
     :param refresh_mac: Refresh the database with newer MAC data ?
+    :param load_mac: Should MAC be loaded at all
     :return: None
     """
-    mac_table_name = "mac_codes"
-    if refresh_mac or not db.table_exists(db_connection, mac_table_name):
-        # Load the MAC file to a DataFrame
-        mac_url = "https://hml.nmdp.org/mac/files/numer.v3.zip"
-        df_mac = pd.read_csv(
-            mac_url,
-            sep="\t",
-            compression="zip",
-            skiprows=3,
-            names=["Code", "Alleles"],
-            keep_default_na=False,
-        )
-        # Create a dict from code to alleles
-        mac = df_mac.set_index("Code")["Alleles"].to_dict()
-        # Save the mac dict to db
-        db.save_dict(
-            db_connection,
-            table_name=mac_table_name,
-            dictionary=mac,
-            columns=("code", "alleles"),
-        )
+    if load_mac:
+        mac_table_name = "mac_codes"
+        if refresh_mac or not db.table_exists(db_connection, mac_table_name):
+            # Load the MAC file to a DataFrame
+            mac_url = "https://hml.nmdp.org/mac/files/numer.v3.zip"
+            df_mac = pd.read_csv(
+                mac_url,
+                sep="\t",
+                compression="zip",
+                skiprows=3,
+                names=["Code", "Alleles"],
+                keep_default_na=False,
+            )
+            # Create a dict from code to alleles
+            mac = df_mac.set_index("Code")["Alleles"].to_dict()
+            # Save the mac dict to db
+            db.save_dict(
+                db_connection,
+                table_name=mac_table_name,
+                dictionary=mac,
+                columns=("code", "alleles"),
+            )
 
 
 def to_serological_name(locus_name: str):

pyard/pyard.py

@@ -80,7 +80,11 @@ class ARD(object):
     """
 
     def __init__(
-        self, imgt_version: str = "Latest", data_dir: str = None, config: dict = None
+        self,
+        imgt_version: str = "Latest",
+        data_dir: str = None,
+        load_mac: bool = True,
+        config: dict = None,
     ):
         """
         ARD will load valid alleles, xx codes and MAC mappings for the given
@@ -100,7 +104,7 @@ def __init__(
         self.db_connection = db.create_db_connection(data_dir, imgt_version)
 
         # Load MAC codes
-        dr.generate_mac_codes(self.db_connection, False)
+        dr.generate_mac_codes(self.db_connection, refresh_mac=False, load_mac=load_mac)
         # Load ARS mappings
         self.ars_mappings, p_group = dr.generate_ars_mapping(
             self.db_connection, imgt_version
@@ -393,7 +397,7 @@ def validate(self, glstring):
         except InvalidAlleleError as e:
             raise InvalidTypingError(
                 f"{glstring} is not valid GL String. \n {e.message}", e
-            )
+            ) from None
 
     def is_XX(self, glstring: str, loc_antigen: str = None, code: str = None) -> bool:
         if loc_antigen is None or code is None:
@@ -718,7 +722,7 @@ def refresh_mac_codes(self) -> None:
         Refreshes MAC code for the current IMGT db version.
         :return: None
         """
-        dr.generate_mac_codes(self.db_connection, True)
+        dr.generate_mac_codes(self.db_connection, refresh_mac=True)
 
     def get_db_version(self) -> str:
         """

scripts/pyard-import

@@ -92,6 +92,12 @@ if __name__ == "__main__":
         action="store_true",
         help="reinstall a fresh version of database",
     )
+    parser.add_argument(
+        "--skip-mac",
+        dest="skip_mac",
+        action="store_true",
+        help="Skip creating MAC mapping",
+    )
     args = parser.parse_args()
 
     if args.show_versions:
@@ -118,8 +124,14 @@ if __name__ == "__main__":
             db_fullname.unlink(missing_ok=True)
 
     print(f"Importing IMGT database version: {imgt_version}")
+    if args.skip_mac:
+        load_mac = False
+        print(f"Skipping MAC tables creation")
+    else:
+        load_mac = True
+
     try:
-        ard = pyard.ARD(imgt_version=imgt_version, data_dir=data_dir)
+        ard = pyard.ARD(imgt_version=imgt_version, data_dir=data_dir, load_mac=load_mac)
     except ValueError as e:
         print(f"Error importing version {imgt_version}:", e)
         sys.exit(1)
@@ -142,5 +154,5 @@ if __name__ == "__main__":
     if args.refresh_mac:
         print(f"Updating MACs")
         db_connection = db.create_db_connection(data_dir, imgt_version, ro=False)
-        data_repository.generate_mac_codes(db_connection, True)
+        data_repository.generate_mac_codes(db_connection, refresh_mac=True)
         print(f"Updated MACs for {imgt_version} IMGT database.")

scripts/pyard-status

@@ -84,7 +84,9 @@ if __name__ == "__main__":
             print(f"|{'Table Name':20}|{'Rows':20}|")
             print(f"|{'-' * 41}|")
             for table in (
-                data_repository.ars_mapping_tables + data_repository.code_mapping_tables
+                data_repository.ars_mapping_tables
+                + data_repository.code_mapping_tables
+                + ["mac_codes"]
             ):
                 if db.table_exists(db_connection, table):
                     total_rows = db.count_rows(db_connection, table)