Merge pull request #13 from VForWaTer/param-rework

Param rework

Dockerfile

@@ -54,6 +54,9 @@ RUN arch=$(uname -m | sed s/x86_64/amd64/) && \
     chmod +x ./duckdb && \
     mv ./duckdb /duck/duckdb
 
+# pre-install the spatial extension into duckdb as it will be used
+RUN /duck/duckdb -c "INSTALL spatial;"
+
 # go to the source directory of this tool
 WORKDIR /src
 CMD ["python", "run.py"]

README.md

@@ -59,7 +59,6 @@ tables are listed accordingly and license information is extracted and presented
 | keep_data_files | If set to `false`, the data files clipped to the spatial and temporal scale will not be kept. |
 | precision | The precision for aggregations along the temporal scale of the datasets. |
 | resolution | The resolution of the output data. This parameter is only relevant for areal datasets. |
-| connection | The connection URI to the MetaCatalog instance. |
 
 
 ## Development and local run

src/run.py

@@ -20,6 +20,9 @@
 from clip import reference_area_to_file
 from version import __version__
 
+# always load .env files
+load_dotenv()
+
 # parse parameters
 kwargs = get_parameter()
 
@@ -30,20 +33,8 @@
 if toolname != 'vforwater_loader':
     raise AttributeError(f"[{dt.now().isocalendar()}] Either no TOOL_RUN environment variable available, or '{toolname}' is not valid.\n")
 
-
 # use the pydantic model to handle the input parameters
-params = load_params(**kwargs)
-
-# check if a connection was given and if it is a valid path
-# this is handled extra to keep the pydantic model simpler
-if 'connection' in kwargs:
-    if Path(kwargs['connection']).exists():
-        load_dotenv(dotenv_path=kwargs['connection'])
-    else:
-        # it is interpreted as a connection uri
-        connection = kwargs['connection']
-else:
-    load_dotenv()
+params = load_params(**kwargs)   
 
 # check if a connection evironment variable is given
 if 'VFW_POSTGRES_URI' in os.environ:

src/tool.yml

@@ -1,6 +1,6 @@
 tools:
   vforwater_loader:
-    title: Loader for datasets stored in a metacatalog instance 
+    title: Dataset Loader
     description: |
       This tool will use `metacatalog` to load datasets stored in a metacatalog instance, like V-FOR-WaTer.
       The requested datasources will be made available in the output directory of the tool. Areal datasets
@@ -76,12 +76,3 @@ tools:
           If the dataset is areal and the parameter is set, the dataset will be resampled to the given resolution.
           Note: Right now, the aggregations can only write parquet files. For larger (espeically spatio-temporal) datasets, these
           aggreations can be large. A future version will write netCDF or Zarr for these cases.
-      connection:
-        type: string
-        optional: true
-        description: | 
-          THIS OPTION IS DEPRECATED. PLEASE SET THE ENVIRONMENT VARIABLE METACATALOG_URI INSTEAD.
-          To connect to the V-FOR-WaTer database, you need to provide a connection URI. These URIs should be saved to an evnironment variable
-          called `VFW_POSTGRES_URI`. Alternatively, you can pass in a file path to an `.env` file that contains the connection URI. 
-          Passing the string directly as this parameter is technically also possible, but not recommended.
-