refactor: split leaderboard read path into its own module

Mirrors the shape adyen/DABstep uses on their reference Space (UI
assembly in app.py, business logic in a sibling module). app.py drops
from 224 lines to 158 by extracting the env-driven repo constants,
the LOCAL_RESULTS_PATH + LEADERBOARD_COLS schema, the Hub/local row
loaders, and load_leaderboard() into leaderboard.py. The submit
handler stays in app.py for now; it'll move to a third module
(submit.py) when Step 6 (E) lands the real validation + async eval.

No behaviour change: load_leaderboard is imported back into app.py
and wired into the same gr.Dataframe / refresh button pair as before.
Smoke-tested locally (cadgenbench-space env): app module imports
cleanly, load_leaderboard() returns a dataframe with the expected
columns.

app.py

@@ -1,97 +1,20 @@
-"""CADGenBench Leaderboard Space.
+"""CADGenBench Leaderboard Space - Gradio UI assembly.
 
-Step 3 prototype: a hand-crafted ``results.jsonl`` drives the leaderboard
-table, and the Submit tab is a UI-only stub. The read path (Step 5) will
-swap the JSONL for ``datasets.load_dataset(HF_SUBMISSIONS_REPO, 'results')``
-and the write path (Step 6) will run ``cadgenbench evaluate`` and push a
-result row back to the submissions dataset via ``HfApi``.
+Read path lives in :mod:`leaderboard`. The submit handler is a UI-only
+stub here; the real validation + async eval lands in :mod:`submit` as
+part of Step 6 (E).
 """
-
 from __future__ import annotations
 
-import json
-import os
 from pathlib import Path
 
 import gradio as gr
-import pandas as pd
-from huggingface_hub import hf_hub_download
 
-HF_ORG = os.getenv("HF_ORG", "michaelr27")
-HF_SUBMISSIONS_REPO = os.getenv(
-    "HF_SUBMISSIONS_REPO", f"{HF_ORG}/cadgenbench-submissions"
+from leaderboard import (
+    HF_DATA_REPO,
+    HF_SUBMISSIONS_REPO,
+    load_leaderboard,
 )
-HF_DATA_REPO = os.getenv("HF_DATA_REPO", f"{HF_ORG}/cadgenbench-data")
-
-LOCAL_RESULTS_PATH = Path(__file__).parent / "results.jsonl"
-
-LEADERBOARD_COLS = [
-    "submission_name",
-    "submitter_name",
-    "aggregate_score",
-    "validity_rate",
-    "submitted_at",
-    "cadgenbench_version",
-]
-
-
-def _load_rows_from_hub() -> list[dict] | None:
-    """Pull results.jsonl from the submissions dataset.
-
-    Returns None on any failure so callers can fall back to the local file.
-    """
-    try:
-        path = hf_hub_download(
-            repo_id=HF_SUBMISSIONS_REPO,
-            filename="results.jsonl",
-            repo_type="dataset",
-            force_download=True,
-        )
-        return [
-            json.loads(line)
-            for line in Path(path).read_text().splitlines()
-            if line.strip()
-        ]
-    except Exception as e:  # noqa: BLE001 — any failure should fall back
-        print(f"[load_leaderboard] Hub fetch failed ({type(e).__name__}: {e})")
-        return None
-
-
-def _load_rows_from_local() -> list[dict]:
-    if not LOCAL_RESULTS_PATH.exists():
-        return []
-    return [
-        json.loads(line)
-        for line in LOCAL_RESULTS_PATH.read_text().splitlines()
-        if line.strip()
-    ]
-
-
-def _fmt_pct(x: float | None) -> str:
-    """Render a 0-1 fraction as 'NN%' (or 'NN.N%' for non-whole values)."""
-    if x is None:
-        return ""
-    pct = float(x) * 100
-    return f"{pct:.0f}%" if pct == int(pct) else f"{pct:.1f}%"
-
-
-def load_leaderboard() -> pd.DataFrame:
-    rows = _load_rows_from_hub()
-    if rows is None:
-        print("[load_leaderboard] falling back to local results.jsonl")
-        rows = _load_rows_from_local()
-    if not rows:
-        return pd.DataFrame(columns=LEADERBOARD_COLS)
-    df = pd.DataFrame(rows)
-    cols = [c for c in LEADERBOARD_COLS if c in df.columns]
-    df = (
-        df[cols]
-        .sort_values("aggregate_score", ascending=False, na_position="last")
-        .reset_index(drop=True)
-    )
-    if "validity_rate" in df.columns:
-        df["validity_rate"] = df["validity_rate"].map(_fmt_pct)
-    return df
 
 
 def handle_submit(

leaderboard.py

@@ -0,0 +1,91 @@
+"""Leaderboard read path.
+
+Loads `results.jsonl` from the submissions dataset on the Hub (or falls
+back to the local mirror on any Hub error) and shapes the rows into the
+dataframe shown on the Leaderboard tab. Module-level constants describe
+the env-var-driven repo identities that the submit path also consumes.
+"""
+from __future__ import annotations
+
+import json
+import os
+from pathlib import Path
+
+import pandas as pd
+from huggingface_hub import hf_hub_download
+
+HF_ORG = os.getenv("HF_ORG", "michaelr27")
+HF_SUBMISSIONS_REPO = os.getenv(
+    "HF_SUBMISSIONS_REPO", f"{HF_ORG}/cadgenbench-submissions"
+)
+HF_DATA_REPO = os.getenv("HF_DATA_REPO", f"{HF_ORG}/cadgenbench-data")
+
+LOCAL_RESULTS_PATH = Path(__file__).parent / "results.jsonl"
+
+LEADERBOARD_COLS = [
+    "submission_name",
+    "submitter_name",
+    "aggregate_score",
+    "validity_rate",
+    "submitted_at",
+    "cadgenbench_version",
+]
+
+
+def _load_rows_from_hub() -> list[dict] | None:
+    """Pull results.jsonl from the submissions dataset.
+
+    Returns None on any failure so callers can fall back to the local file.
+    """
+    try:
+        path = hf_hub_download(
+            repo_id=HF_SUBMISSIONS_REPO,
+            filename="results.jsonl",
+            repo_type="dataset",
+            force_download=True,
+        )
+        return [
+            json.loads(line)
+            for line in Path(path).read_text().splitlines()
+            if line.strip()
+        ]
+    except Exception as e:  # noqa: BLE001 - any failure should fall back
+        print(f"[load_leaderboard] Hub fetch failed ({type(e).__name__}: {e})")
+        return None
+
+
+def _load_rows_from_local() -> list[dict]:
+    if not LOCAL_RESULTS_PATH.exists():
+        return []
+    return [
+        json.loads(line)
+        for line in LOCAL_RESULTS_PATH.read_text().splitlines()
+        if line.strip()
+    ]
+
+
+def _fmt_pct(x: float | None) -> str:
+    """Render a 0-1 fraction as 'NN%' (or 'NN.N%' for non-whole values)."""
+    if x is None:
+        return ""
+    pct = float(x) * 100
+    return f"{pct:.0f}%" if pct == int(pct) else f"{pct:.1f}%"
+
+
+def load_leaderboard() -> pd.DataFrame:
+    rows = _load_rows_from_hub()
+    if rows is None:
+        print("[load_leaderboard] falling back to local results.jsonl")
+        rows = _load_rows_from_local()
+    if not rows:
+        return pd.DataFrame(columns=LEADERBOARD_COLS)
+    df = pd.DataFrame(rows)
+    cols = [c for c in LEADERBOARD_COLS if c in df.columns]
+    df = (
+        df[cols]
+        .sort_values("aggregate_score", ascending=False, na_position="last")
+        .reset_index(drop=True)
+    )
+    if "validity_rate" in df.columns:
+        df["validity_rate"] = df["validity_rate"].map(_fmt_pct)
+    return df