fix session id for observations

db/migrations/versions/20250604_162113_observation_session_as_string.py

@@ -0,0 +1,39 @@
+"""observation session as string
+
+Revision ID: 58439dd3088b
+Revises: 77cdbfc882e2
+Create Date: 2025-06-04 16:21:13.610668
+
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = "58439dd3088b"
+down_revision: Union[str, None] = "77cdbfc882e2"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.alter_column(
+        "agent_observation",
+        "session_id",
+        existing_type=sa.UUID(),
+        type_=sa.Text(),
+        existing_nullable=True,
+    )
+
+
+def downgrade() -> None:
+    op.alter_column(
+        "agent_observation",
+        "session_id",
+        existing_type=sa.Text(),
+        type_=sa.UUID(),
+        existing_nullable=True,
+    )

src/memory/common/db/models/source_item.py

@@ -81,7 +81,7 @@ def handle_duplicate_sha256(session, flush_context, instances):
 
 
 def clean_filename(filename: str) -> str:
-    return re.sub(r"[^a-zA-Z0-9_]", "_", filename).strip("_")
+    return re.sub(r"[^a-zA-Z0-9_]", "_", filename).strip("_")[:30]
 
 
 def image_filenames(chunk_id: str, images: list[Image.Image]) -> list[str]:

src/memory/common/db/models/source_items.py

@@ -564,9 +564,7 @@ class AgentObservation(SourceItem):
     id = Column(
         BigInteger, ForeignKey("source_item.id", ondelete="CASCADE"), primary_key=True
     )
-    session_id = Column(
+    session_id = Column(Text, nullable=True)
-        UUID(as_uuid=True)
-    )  # Groups observations from same conversation
     observation_type = Column(
         Text, nullable=False
     )  # belief, preference, pattern, contradiction, behavior

src/memory/common/settings.py

@@ -99,7 +99,7 @@ QDRANT_TIMEOUT = int(os.getenv("QDRANT_TIMEOUT", "60"))
 # Worker settings
 # Intervals are in seconds
 EMAIL_SYNC_INTERVAL = int(os.getenv("EMAIL_SYNC_INTERVAL", 60 * 60))
-COMIC_SYNC_INTERVAL = int(os.getenv("COMIC_SYNC_INTERVAL", 60 * 60))
+COMIC_SYNC_INTERVAL = int(os.getenv("COMIC_SYNC_INTERVAL", 60 * 60 * 24))
 ARTICLE_FEED_SYNC_INTERVAL = int(os.getenv("ARTICLE_FEED_SYNC_INTERVAL", 30 * 60))
 CLEAN_COLLECTION_INTERVAL = int(os.getenv("CLEAN_COLLECTION_INTERVAL", 24 * 60 * 60))
 CHUNK_REINGEST_INTERVAL = int(os.getenv("CHUNK_REINGEST_INTERVAL", 60 * 60))

src/memory/workers/ingest.py

@@ -5,6 +5,8 @@ from memory.common.celery_app import (
     app,
     CLEAN_ALL_COLLECTIONS,
     REINGEST_MISSING_CHUNKS,
+    SYNC_ALL_COMICS,
+    SYNC_ALL_ARTICLE_FEEDS,
 )
 
 logger = logging.getLogger(__name__)
@@ -24,11 +26,11 @@ app.conf.beat_schedule = {
         "schedule": settings.EMAIL_SYNC_INTERVAL,
     },
     "sync-all-comics": {
-        "task": "memory.workers.tasks.comic.sync_all_comics",
+        "task": SYNC_ALL_COMICS,
         "schedule": settings.COMIC_SYNC_INTERVAL,
     },
     "sync-all-article-feeds": {
-        "task": "memory.workers.tasks.blogs.sync_all_article_feeds",
+        "task": SYNC_ALL_ARTICLE_FEEDS,
         "schedule": settings.ARTICLE_FEED_SYNC_INTERVAL,
     },
 }

src/memory/workers/tasks/comic.py

@@ -91,7 +91,9 @@ def sync_comic(
     file_type = image_url.split(".")[-1]
     mime_type = f"image/{file_type}"
     filename = (
-        settings.COMIC_STORAGE_DIR / clean_filename(author) / f"{title}.{file_type}"
+        settings.COMIC_STORAGE_DIR
+        / clean_filename(author)
+        / f"{clean_filename(title)}.{file_type}"
     )
 
     filename.parent.mkdir(parents=True, exist_ok=True)