From 88110b0aaa0db182d79451f26cee8b3806c748a7 Mon Sep 17 00:00:00 2001
From: Sven Geboers <geboers.sven@gmail.com>
Date: Sun, 29 Mar 2026 23:28:40 +0200
Subject: [PATCH] Fix update_existing_motions: single write connection and
 module-level duckdb import

Use one DuckDB write connection for the entire update loop instead of
opening/closing per row, wrapped in try/finally for proper cleanup.
Move 'import duckdb' to module level with other imports.
---
 scripts/download_past_year.py | 106 ++++++++++++++++++----------------
 1 file changed, 56 insertions(+), 50 deletions(-)

diff --git a/scripts/download_past_year.py b/scripts/download_past_year.py
index 50bf060..a1f3e69 100644
--- a/scripts/download_past_year.py
+++ b/scripts/download_past_year.py
@@ -18,6 +18,8 @@ from datetime import datetime, timedelta
 from typing import Optional, Tuple
 from urllib.parse import urlparse
 
+import duckdb
+
 sys.path.insert(0, ".")  # run from project root
 
 from api_client import TweedeKamerAPI
@@ -71,8 +73,6 @@ def update_existing_motions(
     Returns:
         (updated_count, skipped_count) tuple.
     """
-    import duckdb
-
     # Read motions with missing body_text
     conn_read = duckdb.connect(db_path, read_only=True)
     rows = conn_read.execute(
@@ -89,56 +89,62 @@ def update_existing_motions(
     updated = 0
     skipped = 0
 
-    for row in rows:
-        motion_id, url, title, description = row
-
-        besluit_id = extract_besluit_id(url or "")
-        if not besluit_id:
-            print(f"  Skipping motion {motion_id}: cannot extract besluit_id from URL")
-            skipped += 1
-            continue
-
-        print(f"  Fetching details for motion {motion_id} (besluit_id={besluit_id})...")
-        details = api._get_motion_details(besluit_id)
-
-        if not details or not details.get("body_text"):
-            print(f"  Skipping motion {motion_id}: no body_text returned")
-            skipped += 1
-            continue
-
-        # Build update: always set body_text; also update title/description if
-        # they were placeholder values (e.g. "Motion abc12345" or "No description available")
-        new_body = details["body_text"]
-        new_title = title
-        new_desc = description
-
-        if title and (title.startswith("Motion ") or title.startswith("Besluit ")):
-            new_title = details.get("title") or title
-
-        if description in (
-            None,
-            "",
-            "No description available",
-            "Geen beschrijving beschikbaar",
-        ):
-            new_desc = details.get("description") or description
-
-        conn_write = duckdb.connect(db_path, read_only=False)
-        conn_write.execute(
-            """
-            UPDATE motions
-            SET body_text = ?, title = ?, description = ?
-            WHERE id = ?
-            """,
-            (new_body, new_title, new_desc, motion_id),
-        )
-        conn_write.close()
+    conn_write = duckdb.connect(db_path, read_only=False)
+    try:
+        for row in rows:
+            motion_id, url, title, description = row
+
+            besluit_id = extract_besluit_id(url or "")
+            if not besluit_id:
+                print(
+                    f"  Skipping motion {motion_id}: cannot extract besluit_id from URL"
+                )
+                skipped += 1
+                continue
+
+            print(
+                f"  Fetching details for motion {motion_id} (besluit_id={besluit_id})..."
+            )
+            details = api._get_motion_details(besluit_id)
+
+            if not details or not details.get("body_text"):
+                print(f"  Skipping motion {motion_id}: no body_text returned")
+                skipped += 1
+                continue
+
+            # Build update: always set body_text; also update title/description if
+            # they were placeholder values (e.g. "Motion abc12345" or "No description available")
+            new_body = details["body_text"]
+            new_title = title
+            new_desc = description
+
+            if title and (title.startswith("Motion ") or title.startswith("Besluit ")):
+                new_title = details.get("title") or title
+
+            if description in (
+                None,
+                "",
+                "No description available",
+                "Geen beschrijving beschikbaar",
+            ):
+                new_desc = details.get("description") or description
+
+            conn_write.execute(
+                """
+                UPDATE motions
+                SET body_text = ?, title = ?, description = ?
+                WHERE id = ?
+                """,
+                (new_body, new_title, new_desc, motion_id),
+            )
 
-        updated += 1
-        print(f"  Updated motion {motion_id}")
+            updated += 1
+            print(f"  Updated motion {motion_id}")
 
-        if delay > 0 and updated + skipped < len(rows):
-            time.sleep(delay)
+            if delay > 0 and updated + skipped < len(rows):
+                time.sleep(delay)
+    finally:
+        conn_write.close()
 
     return updated, skipped