cleanup

Author: avsm

geotessera/registry.py

@@ -1149,31 +1149,20 @@ def fetch_landmask(
             # Use existing local file
             return str(local_path)
 
-        # Query hash from landmasks DataFrame for verification if lon/lat provided
+        # Query hash from landmasks index for verification if lon/lat provided
         file_hash = None
         if (
             self.verify_hashes
             and self._landmasks_df is not None
             and lon is not None
             and lat is not None
         ):
-            # Use pre-computed integer grid indices for robust comparison
             lon_i = coord_to_grid_int(lon)
             lat_i = coord_to_grid_int(lat)
-
-            # Use dictionary index for O(1) lookup - more reliable than DataFrame filtering
             key = (int(lon_i), int(lat_i))
-            if hasattr(self, "_landmask_index") and key in self._landmask_index:
+            if key in self._landmask_index:
                 idx = self._landmask_index[key]
                 file_hash = self._landmasks_df.iloc[idx]["hash"]
-            else:
-                # Fallback to DataFrame filtering
-                matches = self._landmasks_df[
-                    (self._landmasks_df["lon_i"] == lon_i)
-                    & (self._landmasks_df["lat_i"] == lat_i)
-                ]
-                if len(matches) > 0:
-                    file_hash = matches.iloc[0]["hash"]
 
         # Download to embeddings_dir
         url = f"{TESSERA_BASE_URL}/{self.version}/{LANDMASKS_DIR_NAME}/{filename}"
@@ -1260,29 +1249,18 @@ def get_tile_file_size(self, year: int, lon: float, lat: float) -> int:
                 "Please update your registry to include file size metadata."
             )
 
-        # Use pre-computed integer grid indices for robust comparison
+        # Use dictionary index for O(1) lookup
         lon_i = coord_to_grid_int(lon)
         lat_i = coord_to_grid_int(lat)
-
-        # Use dictionary index for O(1) lookup - more reliable than DataFrame filtering
         key = (int(year), int(lon_i), int(lat_i))
-        if hasattr(self, "_tile_index") and key in self._tile_index:
-            idx = self._tile_index[key]
-            return int(self._registry_gdf.iloc[idx]["file_size"])
-
-        # Fallback to DataFrame filtering for compatibility
-        matches = self._registry_gdf[
-            (self._registry_gdf["year"] == year)
-            & (self._registry_gdf["lon_i"] == lon_i)
-            & (self._registry_gdf["lat_i"] == lat_i)
-        ]
 
-        if len(matches) == 0:
+        if key not in self._tile_index:
             raise ValueError(
                 f"Tile not found in registry: year={year}, lon={lon:.2f}, lat={lat:.2f}"
             )
 
-        return int(matches.iloc[0]["file_size"])
+        idx = self._tile_index[key]
+        return int(self._registry_gdf.iloc[idx]["file_size"])
 
     def get_scales_file_size(self, year: int, lon: float, lat: float) -> int:
         """Get the file size of a scales file from the registry.
@@ -1304,29 +1282,18 @@ def get_scales_file_size(self, year: int, lon: float, lat: float) -> int:
                 "Please update your registry to include scales file size metadata."
             )
 
-        # Use pre-computed integer grid indices for robust comparison
+        # Use dictionary index for O(1) lookup
         lon_i = coord_to_grid_int(lon)
         lat_i = coord_to_grid_int(lat)
-
-        # Use dictionary index for O(1) lookup - more reliable than DataFrame filtering
         key = (int(year), int(lon_i), int(lat_i))
-        if hasattr(self, "_tile_index") and key in self._tile_index:
-            idx = self._tile_index[key]
-            return int(self._registry_gdf.iloc[idx]["scales_size"])
-
-        # Fallback to DataFrame filtering for compatibility
-        matches = self._registry_gdf[
-            (self._registry_gdf["year"] == year)
-            & (self._registry_gdf["lon_i"] == lon_i)
-            & (self._registry_gdf["lat_i"] == lat_i)
-        ]
 
-        if len(matches) == 0:
+        if key not in self._tile_index:
             raise ValueError(
                 f"Tile not found in registry: year={year}, lon={lon:.2f}, lat={lat:.2f}"
             )
 
-        return int(matches.iloc[0]["scales_size"])
+        idx = self._tile_index[key]
+        return int(self._registry_gdf.iloc[idx]["scales_size"])
 
     def get_landmask_file_size(self, lon: float, lat: float) -> int:
         """Get the file size of a landmask tile from the registry.
@@ -1353,28 +1320,18 @@ def get_landmask_file_size(self, lon: float, lat: float) -> int:
                 "Please update your landmasks registry to include file size metadata."
             )
 
-        # Use pre-computed integer grid indices for robust comparison
+        # Use dictionary index for O(1) lookup
         lon_i = coord_to_grid_int(lon)
         lat_i = coord_to_grid_int(lat)
-
-        # Use dictionary index for O(1) lookup - more reliable than DataFrame filtering
         key = (int(lon_i), int(lat_i))
-        if hasattr(self, "_landmask_index") and key in self._landmask_index:
-            idx = self._landmask_index[key]
-            return int(self._landmasks_df.iloc[idx]["file_size"])
-
-        # Fallback to DataFrame filtering for compatibility
-        matches = self._landmasks_df[
-            (self._landmasks_df["lon_i"] == lon_i)
-            & (self._landmasks_df["lat_i"] == lat_i)
-        ]
 
-        if len(matches) == 0:
+        if key not in self._landmask_index:
             raise ValueError(
                 f"Landmask not found in registry: lon={lon:.2f}, lat={lat:.2f}"
             )
 
-        return int(matches.iloc[0]["file_size"])
+        idx = self._landmask_index[key]
+        return int(self._landmasks_df.iloc[idx]["file_size"])
 
     def calculate_download_requirements(
         self,