Fix geoparquet issues

Author: m-mohr

vecorel_cli/cli/path_url.py

@@ -1,10 +1,12 @@
-import click
 import pathlib
 
-from ..vecorel.util import is_url, get_fs
+import click
 from yarl import URL
 
-IGNORE_FILES = ["collection.json", "catalog.json"] # likely STAC
+from ..vecorel.util import get_fs, is_url
+
+IGNORE_FILES = ["collection.json", "catalog.json"]  # likely STAC
+
 
 class PathOrURL(click.ParamType):
     name = "path_or_url"
@@ -39,7 +41,9 @@ def convert(self, value, param, ctx):
             if fs.exists(value):
                 return URL(value)
             else:
-                self.fail(f"URL '{value}' is does not exist or is currently unavailable.", param, ctx)
+                self.fail(
+                    f"URL '{value}' is does not exist or is currently unavailable.", param, ctx
+                )
 
         # Otherwise, validate as a local path
         filepath = self.path_type.convert(value, param, ctx)
@@ -57,7 +61,7 @@ def convert(self, value, param, ctx):
             self.fail(
                 f"File '{filepath}' must have one of the following extensions: {', '.join(self.extensions)}",
                 param,
-                ctx
+                ctx,
             )
         return filepath
 

vecorel_cli/conversion/base.py

@@ -134,7 +134,7 @@ def download_files(self, uris, cache_folder=None, **kwargs):
                             zip_file.extractall(zip_folder)
                     except NotImplementedError as e:
                         if str(e) != "That compression method is not supported":
-                            raise
+                            raise e
                         import zipfile_deflate64
 
                         with zipfile_deflate64.ZipFile(cache_file, "r") as zip_file:

vecorel_cli/convert.py

@@ -75,7 +75,7 @@ def __init__(self, dataset: str):
         try:
             self.converter = converters.load(self.dataset)
         except (ImportError, NameError, OSError, RuntimeError, SyntaxError) as e:
-            raise Exception(f"Converter for '{self.dataset}' not available or faulty: {e}")
+            raise Exception(f"Converter for '{self.dataset}' not available or faulty: {e}") from e
 
     @runnable
     def convert(

vecorel_cli/converters.py

@@ -86,8 +86,8 @@ def get_class(self, name):
         base_class = self.get_base_class()
         try:
             module = self.get_module(name)
-        except ModuleNotFoundError:
-            raise ValueError(f"Converter '{name}' not found")
+        except ModuleNotFoundError as e:
+            raise ValueError(f"Converter '{name}' not found") from e
 
         try:
             clazz = next(

vecorel_cli/encoding/geoparquet.py

@@ -1,14 +1,14 @@
 import json
 from pathlib import Path
 from typing import Optional, Union
-from yarl import URL
 
 import pyarrow as pa
 import pyarrow.parquet as pq
 from geopandas import GeoDataFrame
 from geopandas.io.arrow import _arrow_to_geopandas
 from pyarrow import NativeFile
 from pyarrow.fs import FSSpecHandler, PyFileSystem
+from yarl import URL
 
 from ..const import GEOPARQUET_DEFAULT_VERSION, GEOPARQUET_VERSIONS
 from ..encoding.geojson import VecorelJSONEncoder
@@ -181,17 +181,18 @@ def write(
             dtype = schema.get("type")
 
             # Convert the data types in the GeoDataFrame
-            gp_type = get_geopandas_dtype(dtype, required, schema)
-            if gp_type is None:
-                self.warning(f"{column}: No type conversion available for {dtype}")
-            else:
-                try:
-                    if callable(gp_type):
-                        data[column] = gp_type(data[column])
-                    else:
-                        data[column] = data[column].astype(gp_type, copy=False)
-                except Exception as e:
-                    self.warning(f"{column}: Can't convert to {dtype}: {e}")
+            if dtype is not None:
+                gp_type = get_geopandas_dtype(dtype, required, schema)
+                if gp_type is None:
+                    self.warning(f"{column}: No type conversion available for {dtype}")
+                else:
+                    try:
+                        if callable(gp_type):
+                            data[column] = gp_type(data[column])
+                        else:
+                            data[column] = data[column].astype(gp_type, copy=False)
+                    except Exception as e:
+                        self.warning(f"{column}: Can't convert to {dtype}: {e}")
 
             # Create the Parquet schema
             field = None

vecorel_cli/validation/geoparquet.py

@@ -18,7 +18,7 @@ def _validate(self, num: Optional[int] = None, schema_map: SchemaMapping = {}):
 
         # Load data
         try:
-            data = self.encoding.read(num=num, hydrate=True)
+            data = self.encoding.read(num=num, schema_map=schema_map, hydrate=True)
         except Exception as e:
             return self.error(e)
 

vecorel_cli/vecorel/schemas.py

@@ -51,7 +51,7 @@ def resolve_schema_uris(
                 if validator:
                     validator.error(message)
                 else:
-                    raise ValueError(message)
+                    raise ValueError(message) from e
 
         return loaded