Upload folder using huggingface_hub

src/gradio_app.py

@@ -47,16 +47,109 @@ BASIN_NAMES = {0: "La Eure", 1: "La Risle"}
 BASIN_FILE_NAMES = {0: "eure", 1: "risle"}
 N_CLICK_TARGETS = 300
 
-
-def parse_data_root() -> Path:
+def find_data_root() -> Path:
+    """
+    Automatically locate the dataset directory.
+
+    Priority:
+      1. Explicit --data-root argument
+      2. DATA_ROOT environment variable
+      3. Common locations relative to app.py
+      4. Recursive search for a directory containing
+         station_elevations.csv
+    """
     parser = argparse.ArgumentParser()
-    parser.add_argument("--data-root", type=Path, default=Path("datasets"))
+    parser.add_argument("--data-root", type=Path, default=None)
     args, _ = parser.parse_known_args()
-    return args.data_root
+
+    # 1. Explicit CLI argument
+    if args.data_root is not None:
+        root = args.data_root.expanduser().resolve()
+        if root.exists():
+            return root
+
+    # 2. Environment variable
+    import os
+
+    env_root = os.environ.get("DATA_ROOT")
+    if env_root:
+        root = Path(env_root).expanduser().resolve()
+        if root.exists():
+            return root
+
+    # app.py is in src/
+    app_dir = Path(__file__).resolve().parent
+    project_root = app_dir.parent
+
+    # 3. Common locations
+    candidates = [
+        Path.cwd() / "datasets",
+        Path.cwd() / "data",
+        project_root / "datasets",
+        project_root / "data",
+        app_dir / "datasets",
+        app_dir / "data",
+    ]
+
+    for root in candidates:
+        if (root / "station_elevations.csv").exists():
+            return root.resolve()
+
+    # 4. Search recursively from likely roots
+    search_roots = [
+        Path.cwd(),
+        project_root,
+        app_dir,
+    ]
+
+    seen = set()
+
+    for search_root in search_roots:
+        if not search_root.exists():
+            continue
+
+        try:
+            for station_file in search_root.rglob("station_elevations.csv"):
+                root = station_file.parent.resolve()
+
+                if root in seen:
+                    continue
+
+                seen.add(root)
+
+                # Make sure this actually looks like our dataset
+                if (
+                    (root / "station_elevations.csv").exists()
+                    and (
+                        (root / "centerlines").exists()
+                        or (root / "reach_graph").exists()
+                        or (root / "hydrometric").exists()
+                        or (root / "ades").exists()
+                    )
+                ):
+                    return root
+        except (PermissionError, OSError):
+            continue
+
+    raise FileNotFoundError(
+        "Could not automatically locate the dataset directory. "
+        "Expected a directory containing 'station_elevations.csv'. "
+        "Use --data-root PATH or set DATA_ROOT."
+    )
+
+
+DATA_ROOT = find_data_root()
+print(f"[River Network Explorer] DATA_ROOT = {DATA_ROOT}")
+
+# def parse_data_root() -> Path:
+#     parser = argparse.ArgumentParser()
+#     parser.add_argument("--data-root", type=Path, default=Path("datasets"))
+#     args, _ = parser.parse_known_args()
+#     return args.data_root
 
 
-DATA_ROOT = parse_data_root()
 
+# DATA_ROOT = parse_data_root()
 
 # --- Data Loaders (Cached via simple dictionaries/memoization) ---
 _cache = {}