Add credential options to polytope source (#363)

Author: sandorkertesz

docs/guide/sources.rst

@@ -695,19 +695,23 @@ mars
 polytope
 --------
 
-.. py:function:: from_source("polytope", collection, *args, stream=True,  batch_size=1, group_by=None, **kwargs)
+.. py:function:: from_source("polytope", collection, *args, address=None, user_email=None, user_key=None, stream=True, batch_size=1, group_by=None, **kwargs)
   :noindex:
 
   The ``polytope`` source accesses the `Polytope web services <https://polytope-client.readthedocs.io/en/latest/>`_ , using the polytope-client_ package.
 
   :param str collection: the name of the polytope collection
   :param tuple *args: specify the request as a dict
+  :param str address: specify the address of the polytope service
+  :param str user_email: specify the user email credential. Must be used together with ``user_key``. This is an alternative to using the ``POLYTOPE_USER_EMAIL`` environment variable. *Added in version 0.7.0*
+  :param str user_key: specify the user key credential. Must be used together with ``user_email``. This is an alternative to using the ``POLYTOPE_USER_KEY`` environment variable. *Added in version 0.7.0*
   :param bool stream: when it is ``True`` the data is read as a stream. Otherwise the data is retrieved into a file and stored in the :ref:`cache <caching>`. Stream-based access only works for :ref:`grib` and CoverageJson data.
   :param int batch_size: used when ``stream=True`` and ``group_by`` is unset. It defines how many GRIB messages are consumed from the stream and kept in memory at a time. ``batch_size=0`` means all the data is read straight to memory. For details see :ref:`stream source <data-sources-stream>`.
   :param group_by: used when ``stream=True`` and can specify one or more metadata keys to control how GRIB messages are read from the stream. For details see :ref:`stream source <data-sources-stream>`.
   :type group_by: str, list of str
   :param dict **kwargs: other keyword arguments specifying the request
 
+
   The following example retrieves GRIB data from the "ecmwf-mars" polytope collection:
 
   .. code-block:: python

docs/release_notes/version_0.7_updates.rst

@@ -19,7 +19,9 @@ New features
 - added support for Lambert Conformal projection when using :meth:`Field.projection() <data.core.fieldlist.Field.projection>`
 - changed the default of the ``bits_per_value`` option to None in :meth:`NumpyFieldList.save() <data.sources.numpy_list.NumpyFieldList.save>`. None means the original ``bits_per_value`` in the GRIB header is kept when the data is written to disk.
 - added the ``model`` option to the :ref:`data-sources-eod` source
+- added the ``user_email`` and ``user_key`` options to the :ref:`data-sources-polytope` source
 - added the ``prompt`` optional argument to certain retrievals to control whether the prompt is to use. When enabled, the prompt asks the user to provide credentials when none seems to be specified. See e.g. the :ref:`data-sources-cds` or :ref:`data-sources-mars` sources.
+- added the
 - allowed using :func:`save` without specifying a file name. In this case an attempt is made to generate the filename automatically, when it fails an exception is thrown.
 - added the :ref:`data-sources-sample` source to access data used in tests and examples
 - :func:`from_source` now fails when trying to load an empty file

earthkit/data/sources/polytope.py

@@ -59,7 +59,13 @@ class Polytope(Source):
     >>> src.to_xarray()  # if datacube
     """
 
-    def __init__(self, dataset, request, address=None, stream=True, **kwargs) -> None:
+    def __init__(
+        self,
+        dataset,
+        request,
+        stream=True,
+        **kwargs,
+    ) -> None:
         super().__init__()
         assert isinstance(dataset, str)
 
@@ -70,14 +76,30 @@ def __init__(self, dataset, request, address=None, stream=True, **kwargs) -> Non
 
         self.stream = stream
 
-        self.request = dict(dataset=dataset, request=request)
-
-        credentials = PolytopeWebKeyPrompt().check(load=True)
-
+        # Polytope client configuration options
         client_kwargs = {}
-        if address is not None:
-            client_kwargs = {"address": address}
-        self.client = polytope.api.Client(**credentials, **client_kwargs)
+        for k in [
+            "config_path",
+            "address",
+            "port",
+            "username",
+            "key_path",
+            "quiet",
+            "verbose",
+            "log_file",
+            "log_level",
+            "user_key",
+            "user_email",
+            "password",
+            "insecure",
+            "skip_tls",
+            "cli",
+        ]:
+            if k in kwargs:
+                client_kwargs[k] = kwargs.pop(k, None)
+
+        self.request = dict(dataset=dataset, request=request)
+        self.client = polytope.api.Client(**client_kwargs)
 
     def __repr__(self) -> str:
         return f"{self.__class__.__name__}({self.request['dataset']}, {self.request['request']})"

environment.yml

@@ -25,7 +25,7 @@ dependencies:
   - multiurl
   - git+https://github.com/ecmwf/pyfdb
   - ecmwf-opendata>=0.1.2
-  - polytope-client>=0.7.2
+  - polytope-client>=0.7.4
   - earthkit-meteo>=0.0.1
   - eccovjson>=0.0.5
   - earthkit-geo

setup.cfg

@@ -46,7 +46,7 @@ include = earthkit.*
 mars = ecmwf-api-client>=1.6.1
 cds = cdsapi
 fdb = pyfdb
-polytope = polytope-client>=0.7.2
+polytope = polytope-client>=0.7.4
 wekeo = hda
 ecmwf-opendata = ecmwf-opendata>=0.3.3
 odb = pyodc

tests/downstream-ci-requirements.txt

@@ -4,7 +4,7 @@ cfgrib>=0.9.10.1
 eccodes>=1.7.0
 ecmwf-api-client>=1.6.1
 ecmwf-opendata>=0.3.3
-polytope-client>=0.7.2
+polytope-client>=0.7.4
 dask
 entrypoints
 filelock

tests/environment-unit-tests.yml

@@ -27,7 +27,7 @@ dependencies:
   - multiurl
   - git+https://github.com/ecmwf/pyfdb
   - ecmwf-opendata>=0.3.3
-  - polytope-client>=0.7.2
+  - polytope-client>=0.7.4
   - earthkit-meteo>=0.0.1
   - git+https://github.com/ecmwf/earthkit-data-demo-source
   - eccovjson>=0.0.5