Merge pull request #4 from aradhyacp/fix/db-disc-sync-issue

Fix/db disc sync issue

aperisolve/app.py

@@ -112,21 +112,34 @@ def upload_image() -> tuple[Response, int]:
     submission_hash = hashlib.md5(submission_data).hexdigest()
     submission_path = RESULT_FOLDER / img_hash / submission_hash
 
-    # If submission already exists, return submission hash
+    # Check if hash is present on the DB
-    if submission_path.exists():
     submission: Submission = Submission.query.filter_by(
         hash=submission_hash
     ).first()
+
+    # Only return if BOTH the file exists AND the DB entry exists
+    if submission_path.exists() and submission is not None:
         return jsonify({"submission_hash": submission.hash}), 200
     
-    # If image is new, create a new Image entry
+
+    # Self-Healing Logic
     new_img_path = RESULT_FOLDER / img_hash / image_name
-    if not new_img_path.parent.exists():
+
+    # Try to find image in DB by hash
+    sub_img = Image.query.filter_by(
+        hash=img_hash
+    ).first()
+
+    # If DB entry is missing (even if file exists on disk), create it!
+    if sub_img is None:
+        if not new_img_path.parent.exists() or not new_img_path.exists():
             new_img_path.parent.mkdir(parents=True, exist_ok=True)
-        with open(new_img_path, "wb") as f:  # Write file to disk
+            with open(new_img_path, "wb") as f:
                 f.write(image_data)
 
-        new_img = Image(
+
+        # Recreate the missing Database Entry
+        sub_img = Image(
             file=str(new_img_path),
             hash=img_hash,
             size=len(image_data),
@@ -134,15 +147,19 @@ def upload_image() -> tuple[Response, int]:
             first_submission_date=datetime.now(timezone.utc),
             last_submission_date=datetime.now(timezone.utc),
         )
-        db.session.add(new_img)  # pylint: disable=no-member
+        db.session.add(sub_img)
-        db.session.commit()  # pylint: disable=no-member
+        db.session.commit()
 
-    sub_img = Image.query.filter_by(hash=img_hash).first()  # type: ignore
+    # Increment upload count for the image
     sub_img.upload_count += 1
-    db.session.commit()  # pylint: disable=no-member
+    db.session.commit()
 
     # Create new Submission entry
     submission_path.mkdir(parents=True, exist_ok=True)
+
+    # Only create a new submission if it does not already exist
+    # New submission for new image
+    if not submission:
         submission = Submission(
             filename=image.filename,
             password=password,
@@ -150,17 +167,19 @@ def upload_image() -> tuple[Response, int]:
             hash=submission_hash,
             status="pending",
             date=time.time(),
-        image_hash=sub_img.hash,  # type: ignore
+            image_hash=sub_img.hash,
         )
-    db.session.add(submission)  # pylint: disable=no-member
+        db.session.add(submission)
+        db.session.commit()
 
-    db.session.commit()  # pylint: disable=no-member
     # Start the analysis jobs
     queue.enqueue("aperisolve.workers.analyze_image", submission.hash, job_timeout=300)
 
+    # Return submission identifier
     return jsonify({"submission_hash": submission.hash}), 200
 
 
+
 @app.route("/status/<hash_val>", methods=["GET"])
 def get_status(hash_val: str) -> Response:
     """Get the processing status of a submission."""