From 3c5032ee71dcb89d9e2bcf99472890bb44437a24 Mon Sep 17 00:00:00 2001
From: Five Boroughs <40718516+FiveBoroughs@users.noreply.github.com>
Date: Sat, 14 Feb 2026 21:46:18 +0100
Subject: [PATCH] Batch ProgramData query and invalidate index on EPG refresh

- Replace per-entry ProgramData DB query with single batch fetch
- Pass pre-loaded EPGData objects to find_current_program_for_tvg_id
  to avoid redundant select_related re-fetch
- Delete Redis programme index before XMLTV download so stale byte
  offsets are never used during the refresh window
---
 apps/epg/api_views.py | 22 +++++++++++++++-------
 apps/epg/tasks.py     | 25 ++++++++++++++++++-------
 2 files changed, 33 insertions(+), 14 deletions(-)

diff --git a/apps/epg/api_views.py b/apps/epg/api_views.py
index 4fa6b478..68ef4b34 100644
--- a/apps/epg/api_views.py
+++ b/apps/epg/api_views.py
@@ -765,7 +765,7 @@ class CurrentProgramsAPIView(APIView):
                 )
 
             try:
-                epg_data_ids = [int(id) for id in epg_data_ids]
+                epg_data_ids = [int(eid) for eid in epg_data_ids]
             except (ValueError, TypeError):
                 return Response(
                     {"error": "epg_data_ids must contain valid integers"},
@@ -777,12 +777,20 @@ class CurrentProgramsAPIView(APIView):
 
             epg_data_entries = EPGData.objects.select_related('epg_source').filter(id__in=epg_data_ids)
 
+            # Batch-fetch current programs for all requested EPG entries in one query
+            db_programs = ProgramData.objects.filter(
+                epg__in=epg_data_entries, start_time__lte=now, end_time__gt=now
+            ).select_related('epg')
+            # Map epg_data id -> first matching program
+            programs_by_epg = {}
+            for prog in db_programs:
+                if prog.epg_id not in programs_by_epg:
+                    programs_by_epg[prog.epg_id] = prog
+
             current_programs = []
             for epg_data in epg_data_entries:
-                # Check DB first (already-parsed channels)
-                program = ProgramData.objects.filter(
-                    epg=epg_data, start_time__lte=now, end_time__gt=now
-                ).first()
+                # Check batch-fetched DB results first
+                program = programs_by_epg.get(epg_data.id)
 
                 if program:
                     program_data = ProgramDataSerializer(program).data
@@ -794,8 +802,8 @@ class CurrentProgramsAPIView(APIView):
                 if epg_data.epg_source and epg_data.epg_source.source_type == 'dummy':
                     continue
 
-                # Fall back to byte-offset index lookup
-                result = find_current_program_for_tvg_id(epg_data.id)
+                # Fall back to byte-offset index lookup, pass the object to avoid re-fetch
+                result = find_current_program_for_tvg_id(epg_data)
 
                 if result == "timeout":
                     current_programs.append({
diff --git a/apps/epg/tasks.py b/apps/epg/tasks.py
index 05291d4a..6991c14b 100644
--- a/apps/epg/tasks.py
+++ b/apps/epg/tasks.py
@@ -296,6 +296,14 @@ def refresh_epg_data(source_id):
         # Continue with the normal processing...
         logger.info(f"Processing EPGSource: {source.name} (type: {source.source_type})")
         if source.source_type == 'xmltv':
+            # Invalidate the byte-offset index before downloading the new file
+            # so stale offsets are never used during the refresh window.
+            from core.utils import RedisClient
+            try:
+                RedisClient.get_client().delete(f'epg_programme_index_{source.id}')
+            except Exception:
+                pass
+
             fetch_success = fetch_xmltv(source)
             if not fetch_success:
                 logger.error(f"Failed to fetch XMLTV for source {source.name}")
@@ -2499,19 +2507,22 @@ def build_programme_index_task(source_id):
     build_programme_index(source_id)
 
 
-def find_current_program_for_tvg_id(epg_id):
+def find_current_program_for_tvg_id(epg_or_id):
     """
-    Look up the currently-airing program for an EPGData id using the
-    byte-offset index. Falls back to sequential scan if no index exists.
+    Look up the currently-airing program for an EPGData instance (or id) using
+    the byte-offset index. Falls back to sequential scan if no index exists.
 
     Returns dict, None, or "timeout".
     """
     from core.utils import RedisClient
 
-    try:
-        epg = EPGData.objects.select_related('epg_source').get(id=epg_id)
-    except EPGData.DoesNotExist:
-        return None
+    if isinstance(epg_or_id, EPGData):
+        epg = epg_or_id
+    else:
+        try:
+            epg = EPGData.objects.select_related('epg_source').get(id=epg_or_id)
+        except EPGData.DoesNotExist:
+            return None
 
     source = epg.epg_source
     if not source or source.source_type == 'dummy':