From eec2d076e429dfdbfe61dca37c993b70ebe47ee9 Mon Sep 17 00:00:00 2001
From: Jeena <hello@jeena.net>
Date: Wed, 11 Mar 2026 23:32:15 +0000
Subject: [PATCH] ingest: Accept RoomMessageImage events regardless of body
 content

WhatsApp bridge images arrive as RoomMessageImage events with an empty
body field, so the previous .jpg/.jpeg extension check silently rejected
all of them. Accept all RoomMessageImage events and fall back to
"image.jpg" as filename when body is empty. File content is still
validated via magic bytes before upload.
---
 ingest.py | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/ingest.py b/ingest.py
index 29585ab..fd9346c 100644
--- a/ingest.py
+++ b/ingest.py
@@ -206,7 +206,7 @@ def is_supported_file(event) -> bool:
     if isinstance(event, RoomMessageFile):
         return (event.body or "").lower().endswith(".pdf")
     if isinstance(event, RoomMessageImage):
-        return (event.body or "").lower().endswith((".jpg", ".jpeg"))
+        return True  # validate magic bytes later; body may be empty (e.g. WhatsApp bridge)
     if isinstance(event, BadEvent):
         return _bad_event_encrypted_file_info(event) is not None
     return False
@@ -219,7 +219,8 @@ def extract_event_fields(event) -> tuple[str, str, str, Optional[str]]:
         filename = content.get("body", "unknown")
         file_info = content.get("file", {})
         return event.event_id, filename, file_info["url"], json.dumps(file_info)
-    return event.event_id, event.body, event.url, None
+    filename = event.body or "image.jpg"
+    return event.event_id, filename, event.url, None
 
 
 def content_type_for(filename: str) -> str:
@@ -417,6 +418,8 @@ async def catchup_history(
     log.info("Starting historical catchup...")
     token = start_token
     total = 0
+    batches = 0
+    events_seen = 0
 
     while True:
         response = await matrix_client.room_messages(
@@ -430,6 +433,9 @@ async def catchup_history(
             log.error("room_messages error: %s", response)
             break
 
+        batches += 1
+        events_seen += len(response.chunk)
+
         for event in response.chunk:
             if is_supported_file(event):
                 total += 1
@@ -443,7 +449,10 @@ async def catchup_history(
             break
         token = response.end
 
-    log.info("Historical catchup complete — processed %d file event(s).", total)
+    log.info(
+        "Historical catchup complete — processed %d file event(s) across %d batches (%d total events).",
+        total, batches, events_seen,
+    )
 
 
 # ---------------------------------------------------------------------------