Materialize triplets as real files in working dir

.gitignore

@@ -2,3 +2,4 @@ __pycache__/
 *.pyc
 .gradio/
 flagged/
+triplets_local/

app.py

@@ -13,6 +13,7 @@ import json
 import os
 import random
 import re
+import shutil
 import uuid
 from datetime import datetime, timezone
 from pathlib import Path
@@ -33,19 +34,29 @@ EMAIL_RE = re.compile(r"^[^@\s]+@[^@\s]+\.[^@\s]+$")
 
 
 def load_triplets() -> tuple[Path, list[dict]]:
-    local = snapshot_download(
-        repo_id=TRIPLETS_REPO,
-        repo_type="dataset",
-        token=HF_TOKEN,
+    """Download the triplets dataset, then materialize as real files under the
+    working directory so Gradio will serve them. The HF cache uses symlinks
+    into a `blobs/` sibling, which Gradio's symlink-resolving allowlist check
+    rejects even when the snapshot dir is whitelisted."""
+    snapshot = Path(
+        snapshot_download(
+            repo_id=TRIPLETS_REPO,
+            repo_type="dataset",
+            token=HF_TOKEN,
+        )
     )
-    root = Path(local)
+    local_root = Path(__file__).resolve().parent / "triplets_local"
+    if local_root.exists():
+        shutil.rmtree(local_root)
+    # follow_symlinks=True (default) copies file contents from the blobs sibling.
+    shutil.copytree(snapshot, local_root)
     rows: list[dict] = []
-    with open(root / "metadata.jsonl") as f:
+    with open(local_root / "metadata.jsonl") as f:
         for line in f:
             line = line.strip()
             if line:
                 rows.append(json.loads(line))
-    return root, rows
+    return local_root, rows
 
 
 def email_slug(email: str) -> str:
@@ -77,10 +88,6 @@ TRIPLETS_ROOT, TRIPLETS = load_triplets()
 TRIPLET_BY_ID = {r["triplet_id"]: r for r in TRIPLETS}
 print(f"loaded {len(TRIPLETS)} triplets from {TRIPLETS_ROOT}")
 
-# Triplets live in the HF snapshot cache (outside the gradio working dir),
-# so Gradio refuses to serve them by default. Whitelist the snapshot dir.
-gr.set_static_paths([str(TRIPLETS_ROOT)])
-
 
 # ---------- session helpers ---------------------------------------------------
 
@@ -344,4 +351,4 @@ with gr.Blocks(title="Denoising A/B Judging", theme=gr.themes.Soft()) as demo:
 
 
 if __name__ == "__main__":
-    demo.launch(allowed_paths=[str(TRIPLETS_ROOT)])
+    demo.launch()