Add very basic GithubAlgorithmRepository

Author: soxofaan

src/esa_apex_toolbox/algorithms.py

@@ -3,7 +3,7 @@
 import dataclasses
 import json
 from pathlib import Path
-from typing import Optional, Union
+from typing import List, Optional, Union
 
 import requests
 
@@ -102,3 +102,38 @@ def from_ogc_api_record(cls, src: Union[dict, str, Path]) -> Algorithm:
             description=properties.get("description"),
             udp_link=udp_link,
         )
+
+
+class GithubAlgorithmRepository:
+    """
+    GitHub based algorithm repository.
+    """
+
+    # TODO: caching
+
+    def __init__(self, owner: str, repo: str, folder: str = "", branch: str = "main"):
+        self.owner = owner
+        self.repo = repo
+        self.folder = folder
+        self.branch = branch
+        self._session = requests.Session()
+
+    def _list_files(self):
+        url = f"https://api.github.com/repos/{self.owner}/{self.repo}/contents/{self.folder}".strip("/")
+        resp = self._session.get(url, headers={"Accept": "application/vnd.github.object+json"})
+        resp.raise_for_status()
+        listing = resp.json()
+        assert listing["type"] == "dir"
+        for item in listing["entries"]:
+            if item["type"] == "file":
+                yield item
+
+    def list_algorithms(self) -> List[str]:
+        # TODO: method to list names vs method to list parsed Algorithm objects?
+        return [item["name"] for item in self._list_files()]
+
+    def get_algorithm(self, name: str) -> Algorithm:
+        # TODO: get url from listing from API request, instead of hardcoding this raw url?
+        url = f"https://raw.githubusercontent.com/{self.owner}/{self.repo}/{self.branch}/{self.folder}/{name}"
+        # TODO: how to make sure GitHub URL is requested with additional headers?
+        return Algorithm.from_ogc_api_record(url)

tests/test_algorithms.py

@@ -3,7 +3,12 @@
 
 import pytest
 
-from esa_apex_toolbox.algorithms import Algorithm, InvalidMetadataError, UdpLink
+from esa_apex_toolbox.algorithms import (
+    Algorithm,
+    GithubAlgorithmRepository,
+    InvalidMetadataError,
+    UdpLink,
+)
 
 DATA_ROOT = Path(__file__).parent / "data"
 
@@ -174,3 +179,32 @@ def test_from_ogc_api_record_url(self, requests_mock):
             href="https://esa-apex.test/udp/algorithm01.json",
             title="UDP One",
         )
+
+
+class TestGithubAlgorithmRepository:
+    @pytest.fixture
+    def repo(self) -> GithubAlgorithmRepository:
+        # TODO: avoid depending on an actual GitHub repository. Mock it instead?
+        #       Or run this as an integration test?
+        return GithubAlgorithmRepository(
+            owner="ESA-APEx",
+            repo="apex_algorithms",
+            folder="algorithm_catalog",
+        )
+
+    def test_list_algorithms(self, repo):
+        assert repo.list_algorithms() == [
+            "worldcereal.json",
+        ]
+
+    def test_get_algorithm(self, repo):
+        algorithm = repo.get_algorithm("worldcereal.json")
+        assert algorithm == Algorithm(
+            id="worldcereal_maize",
+            title="ESA worldcereal global maize detector",
+            description="A maize detection algorithm.",
+            udp_link=UdpLink(
+                href="https://github.com/ESA-APEX/apex_algorithms/blob/main/openeo_udp/worldcereal_inference.json",
+                title="openEO UDP",
+            ),
+        )