make dataid of an artifact downlable

koubaa-hmc · Dec 18, 2024 · da890cc · da890cc
1 parent 786240f
commit da890cc
Show file tree

Hide file tree

Showing 4 changed files with 85 additions and 5 deletions.
diff --git a/databusclient/client.py b/databusclient/client.py
@@ -461,6 +461,12 @@ def __handle_databus_collection__(endpoint, uri: str) -> str:
     return requests.get(uri, headers=headers).text
 
 
+def __handle_databus_artefact__(endpoint, uri: str) -> str:
+    headers = {"Accept": "application/ld+json"}
+    resp = requests.get(uri, headers=headers).text
+    return resp
+
+
 def __download_list__(urls: List[str], local_dir: str):
     for url in urls:
         __download_file__(url=url, filename=local_dir+"/"+wsha256(url))
@@ -474,7 +480,7 @@ def download(
     """
     Download datasets to local storage from databus registry
     ------
-    localDir: the local directory
+    local_dir: the local directory
     databus_uris: identifiers to access databus registered datasets
     """
     for databus_uri in databus_uris:
@@ -483,14 +489,15 @@ def download(
             # databus collection
             if "/collections/" in databus_uri:  # TODO "in" is not safe! there could be an artifact named collections, need to check for the correct part position in the URI
                 query = __handle_databus_collection__(endpoint, databus_uri)
-                res = __handle__databus_file_query__(endpoint, query)
+                response = __handle__databus_file_query__(endpoint, query)
             else:
-                print("dataId not supported yet")  # TODO add support for other DatabusIds here (artifact, group, etc.)
+                response = __handle_databus_artefact__(endpoint, databus_uri)
+                # print("dataId not supported yet")  # TODO add support for other DatabusIds here (artifact, group, etc.)
         # query in local file
         elif databus_uri.startswith("file://"):
             print("query in file not supported yet")
         # query as argument
         else:
             print("QUERY {}", databus_uri.replace("\n", " "))
-            res = __handle__databus_file_query__(endpoint, databus_uri)
+            response = __handle__databus_file_query__(endpoint, databus_uri)
             __download_list__(res, local_dir)
diff --git a/databusclient/consume/download.py b/databusclient/consume/download.py
@@ -1,4 +1,5 @@
 # All kind of download functionalities for Databus ###
 
-class Downloder:
+
+class Downloader:
     pass
diff --git a/tests/data/file_0001.json b/tests/data/file_0001.json
@@ -0,0 +1,67 @@
+{
+  "name": "iai_photovoltaic_active_power_data_202303",
+  "title": "Active power generation through a photovoltaic installation in March 2023",
+  "id": "https://openenergy-platform.org/dataedit/view/model_draft/iai_active_power_pv_202303",
+  "description": "Measured active power produced",
+  "publicationDate": "2023-03-13",
+  "context": {
+    "contact": "[email protected]",
+    "fundingAgency": "Helmholtz Metadata Collaboration"
+  },
+  "temporal": {
+    "referenceDate": "2023-02-13",
+    "timeseries": [
+      {
+        "start": "2023-02-13",
+        "end": "2023-03-13",
+        "resolution": "1h",
+        "alignment": "right",
+        "aggregationType": "mean"
+      }
+    ]
+  },
+  "resources": [
+    {
+      "name": "iai_active_power_pv_202303",
+      "schema": {
+        "foreignKeys": [],
+        "fields": [
+          {
+            "name": "id",
+            "description": "unique identifier like generated for the panda data frame",
+            "type": "bigint"
+          },
+          {
+            "name": "datetime",
+            "description": "Date and time in yyyy-MM-dd hh:mm:ss",
+            "type": "timestamp"
+          },
+          {
+            "name": "active_power",
+            "description": "Active power generated by the photovoltaic installation",
+            "type": "float",
+            "unit": "KW"
+          }
+        ]
+      }
+    }
+  ],
+  "metaMetadata": {
+    "metadataVersion": "OEP-1.5.2",
+    "metadataLicense": {
+      "name": "CC0-1.0",
+      "title": "Creative Commons Zero v1.0 Universal",
+      "path": "https://creativecommons.org/publicdomain/zero/1.0/"
+    }
+  },
+  "_comment": {
+    "metadata": "Metadata documentation and explanation (https://github.com/OpenEnergyPlatform/oemetadata)",
+    "dates": "Dates and time must follow the ISO8601 including time zone (YYYY-MM-DD or YYYY-MM-DDThh:mm:ss±hh)",
+    "units": "Use a space between numbers and units (100 m)",
+    "languages": "Languages must follow the IETF (BCP47) format (en-GB, en-US, de-DE)",
+    "licenses": "License name must follow the SPDX License List (https://spdx.org/licenses/)",
+    "review": "Following the OEP Data Review (https://github.com/OpenEnergyPlatform/data-preprocessing/blob/master/data-review/manual/review_manual.md)",
+    "null": "If not applicable use: null",
+    "todo": "If a value is not yet available, use: todo"
+  }
+}
diff --git a/tests/test_download.py b/tests/test_download.py
@@ -9,6 +9,7 @@
 } LIMIT 10
 """
 TEST_COLLECTION = "https://databus.openenergyplatform.org/koubaa-hmc/collections/sandbox"
+TEST_ARTEFACT = "https://databus.openenergyplatform.org/koubaa-hmc/active_photovoltaic/01-2022"
 
 
 def test_with_query():
@@ -17,3 +18,7 @@ def test_with_query():
 
 def test_with_collection():
     cl.download("tmp", DEFAULT_ENDPOINT, [TEST_COLLECTION])
+
+
+def test_with_artefact():
+    cl.download("tmp", DEFAULT_ENDPOINT, [TEST_ARTEFACT])