Update src/storage.py

src/storage.py

@@ -3,68 +3,51 @@ import json
 import os
 from datetime import datetime
 from pathlib import Path
+
 from huggingface_hub import HfApi, hf_hub_download
 
-# ==============================
-# CONFIG
-# ==============================
 DATASET_REPO = "Usmansafder/engchain-annotations"
-DATA_FILE = "reviews.jsonl"
-
-HF_TOKEN = os.environ.get("HF_TOKEN")  # must be set as Space secret
+HF_TOKEN = os.environ.get("HF_TOKEN")
 
-if HF_TOKEN is None:
+if not HF_TOKEN:
     raise RuntimeError(
-        "HF_TOKEN environment variable not set. "
-        "Add it as a Space secret with write access."
+        "HF_TOKEN not set. Add it in Space Settings → Secrets with write access."
     )
 
 api = HfApi(token=HF_TOKEN)
 
-# ==============================
-# HELPERS
-# ==============================
-def _load_existing_reviews(tmp_path: Path) -> list:
-    if not tmp_path.exists():
-        return []
-    with open(tmp_path, "r", encoding="utf-8") as f:
-        return f.readlines()
+def _safe_slug(s: str) -> str:
+    s = s.strip()
+    s = "".join(c for c in s if c.isalnum() or c in (" ", "_", "-"))
+    s = s.replace(" ", "_")
+    return s or "unknown"
 
-# ==============================
-# MAIN API
-# ==============================
-def save_review(
-    annotator_name,
-    branch,
-    area,
-    template_name,
-    scores,
-    decision,
-    feedback,
-):
+def _download_json_if_exists(path_in_repo: str, local_path: Path) -> list:
     """
-    Appends a single review to a JSONL file stored
-    in a Hugging Face Dataset repository.
+    Returns list of existing reviews; empty list if file doesn't exist yet.
     """
-
-    # Temporary local path
-    local_file = Path("/tmp") / DATA_FILE
-
-    # Download existing dataset file if it exists
     try:
         downloaded = hf_hub_download(
             repo_id=DATASET_REPO,
-            filename=DATA_FILE,
+            filename=path_in_repo,
             repo_type="dataset",
             token=HF_TOKEN,
         )
-        local_file.write_text(
-            Path(downloaded).read_text(encoding="utf-8"),
-            encoding="utf-8",
-        )
+        local_path.write_text(Path(downloaded).read_text(encoding="utf-8"), encoding="utf-8")
+        data = json.loads(local_path.read_text(encoding="utf-8"))
+        return data if isinstance(data, list) else []
     except Exception:
-        # File does not exist yet — first write
-        pass
+        return []
+
+def save_review(annotator_name, branch, area, template_name, scores, decision, feedback):
+    safe_name = _safe_slug(annotator_name)
+    safe_branch = _safe_slug(branch)
+
+    # Store per-reviewer, per-branch
+    path_in_repo = f"reviews/{safe_name}/{safe_branch}.json"
+    local_file = Path("/tmp") / f"{safe_name}_{safe_branch}.json"
+
+    existing = _download_json_if_exists(path_in_repo, local_file)
 
     review_data = {
         "timestamp": datetime.now().isoformat(),
@@ -81,17 +64,16 @@ def save_review(
         "feedback": feedback,
     }
 
-    # Append new record
-    with open(local_file, "a", encoding="utf-8") as f:
-        f.write(json.dumps(review_data) + "\n")
+    existing.append(review_data)
+
+    local_file.write_text(json.dumps(existing, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
 
-    # Upload back to dataset repo
     api.upload_file(
         path_or_fileobj=str(local_file),
-        path_in_repo=DATA_FILE,
+        path_in_repo=path_in_repo,
         repo_id=DATASET_REPO,
         repo_type="dataset",
-        commit_message=f"Add review by {annotator_name}",
+        commit_message=f"Add review: {safe_name} ({safe_branch})",
     )
 
-    return f"hf://datasets/{DATASET_REPO}/{DATA_FILE}"
+    return f"hf://datasets/{DATASET_REPO}/{path_in_repo}"