Sketch to support private data-proxy dataset

This change is merely adding the ability to recognize and
process non-public dataset data-proxy URLs.

However, it is not enough to support such datasets, because
the underlying `fairgraph` query to get a dataset's file listing
returns no results.

The query is essentially this

```py
batch = omcore.File.list(
    self.client,
    file_repository=dvr,
    size=chunk_size,
    from_index=cur_index)
```

and for the dataset referenced in
#58 it returns an empty
list with

- a properly authenticated `client`
- `dvr`: `FileRepository(name='buckets/d-07ab1665-73b0-40c5-800e-557bc319109d', iri=IRI(https://data-proxy.ebrains.eu/api/v1/buckets/d-07ab1665-73b0-40c5-800e-557bc319109d)...`
- `chunk_size`: 10000
- `cur_index`: 0

With the same requesting account, I can browser-visit
https://data-proxy.ebrains.eu/datasets/07ab1665-73b0-40c5-800e-557bc319109d
and see a file listing.

datalad_ebrains/fairgraph_query.py

@@ -186,9 +186,14 @@ def get_file_records(self, ds, kg_dsver):
         # EBRAINS uses different file repositories that need slightly
         # different handling
         dvr_url_p = urlparse(dvr.iri.value)
+        # public data-proxy datasets
         if dvr_url_p.netloc == 'data-proxy.ebrains.eu' \
                 and dvr_url_p.path.startswith('/api/v1/public/buckets/'):
-            get_fname = _get_fname_dataproxy_v1_bucket
+            get_fname = _get_fname_dataproxy_v1_bucket_public
+        # private data-proxy datasets (e.g. human data gateway)
+        elif dvr_url_p.netloc == 'data-proxy.ebrains.eu' \
+                and dvr_url_p.path.startswith('/api/v1/buckets/'):
+            get_fname = _get_fname_dataproxy_v1_bucket_private
         elif dvr_url_p.netloc == 'object.cscs.ch' \
                 and dvr_url_p.query.startswith('prefix='):
             # get the repos base url by removing the query string
@@ -281,7 +286,7 @@ def get_agent_info(self, kg_dsver):
         }
 
 
-def _get_fname_dataproxy_v1_bucket(f):
+def _get_fname_dataproxy_v1_bucket_public(f):
     f_url_p = urlparse(f.iri.value)
     assert f_url_p.netloc == 'data-proxy.ebrains.eu'
     assert f_url_p.path.startswith('/api/v1/public/buckets/')
@@ -290,6 +295,15 @@ def _get_fname_dataproxy_v1_bucket(f):
     return Path(*path.parts[6:])
 
 
+def _get_fname_dataproxy_v1_bucket_private(f):
+    f_url_p = urlparse(f.iri.value)
+    assert f_url_p.netloc == 'data-proxy.ebrains.eu'
+    assert f_url_p.path.startswith('/api/v1/buckets/')
+    path = PurePosixPath(f_url_p.path)
+    # take everything past the bucket_id and turn into a Platform native path
+    return Path(*path.parts[5:])
+
+
 def _get_fname_cscs_repo(baseurl, prefix, f):
     f_url = f.iri.value
     # we presently have no better way to determine a relative file path