From 096ab99a2d961e864f340c39252b4c8eecc72191 Mon Sep 17 00:00:00 2001
From: Frederick Muriuki Muriithi
Date: Thu, 14 Dec 2023 19:22:58 +0300
Subject: samples: Create external script and fix some bugs.

---
 qc_app/samples.py | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)

(limited to 'qc_app/samples.py')

diff --git a/qc_app/samples.py b/qc_app/samples.py
index dee08e5..88a0fde 100644
--- a/qc_app/samples.py
+++ b/qc_app/samples.py
@@ -159,6 +159,10 @@ def cross_reference_samples(conn: mdb.Connection,
                             strain_names: Iterator[str]):
     """Link samples to their population."""
     with conn.cursor(cursorclass=DictCursor) as cursor:
+        cursor.execute(
+            "SELECT MAX(OrderId) AS loid FROM StrainXRef WHERE InbredSetId=%s",
+            (population_id,))
+        last_order_id = cursor.fetchone()["loid"]
         while True:
             batch = take(strain_names, 5000)
             if len(batch) == 0:
@@ -171,10 +175,13 @@ def cross_reference_samples(conn: mdb.Connection,
                 f"({params_str}) AND sx.StrainId IS NULL",
                 (species_id,) + tuple(batch))
             strain_ids = (sid["Id"] for sid in cursor.fetchall())
-            cursor.execute(
-                "SELECT MAX(OrderId) AS loid FROM StrainXRef WHERE InbredSetId=%s",
-                (population_id,))
-            last_order_id = cursor.fetchone()["loid"]
+            params = tuple({
+                "pop_id": population_id,
+                "strain_id": strain_id,
+                "order_id": last_order_id + (order_id * 10),
+                "mapping": "N",
+                "pedigree": None
+            } for order_id, strain_id in enumerate(strain_ids, start=1))
             cursor.executemany(
                 "INSERT INTO StrainXRef( "
                 "  InbredSetId, StrainId, OrderId, Used_for_mapping, PedigreeStatus"
@@ -183,14 +190,8 @@ def cross_reference_samples(conn: mdb.Connection,
                 "  %(pop_id)s, %(strain_id)s, %(order_id)s, %(mapping)s, "
                 "  %(pedigree)s"
                 ")",
-                tuple({
-                    "pop_id": population_id,
-                    "strain_id": strain_id,
-                    "order_id": order_id,
-                    "mapping": "N",
-                    "pedigree": None
-                } for order_id, strain_id in
-                      enumerate(strain_ids, start=(last_order_id+10))))
+                params)
+            last_order_id += (len(params) * 10)
 
 @samples.route("/upload/samples", methods=["POST"])
 def upload_samples():
-- 
cgit v1.2.3