decomposer: generate files for Build source connectors that retrieve content and source permissions from SCM and CMS systems and expose normalized content-change records to downstream ingestion workflows.

tests/test_connectors.py

@@ -0,0 +1,29 @@
+from source_connectors import GitHubConnector, SharePointConnector, StaticTokenProvider, Cursor
+
+class FakeHTTP:
+    def __init__(self, responses): self.responses, self.urls = responses, []
+    def get(self, url, headers):
+        self.urls.append((url, headers)); return self.responses[url]
+
+def test_github_enumeration_is_scoped_and_authenticated():
+    url = "https://api.github.com/repos/acme/docs/git/trees/main?recursive=1"
+    http = FakeHTTP({url: {"tree": [{"type": "blob", "path": "a.md", "sha": "s", "url": "u"}]}})
+    rows = list(GitHubConnector(StaticTokenProvider("t"), http).enumerate({"owner":"acme","repo":"docs","ref":"main"}))
+    assert rows[0].path == "a.md" and http.urls[0][1]["Authorization"] == "Bearer t"
+
+def test_github_webhook_normalizes_delete():
+    event = GitHubConnector(StaticTokenProvider("t"), FakeHTTP({})).normalize_webhook({"X-GitHub-Delivery":"d","X-GitHub-Event":"push"}, {"ref":"refs/heads/main","after":"sha","repository":{"owner":{"login":"a"},"name":"r"}})
+    assert event.event_id == "d" and event.cursor_hint.value == "sha"
+
+def test_sharepoint_delta_and_acl():
+    delta = "https://graph.microsoft.com/v1.0/sites/s/drives/d/root/delta"
+    acl = "https://graph.microsoft.com/v1.0/sites/s/drives/d/root/permissions"
+    http = FakeHTTP({delta: {"value":[{"id":"1","name":"x","file":{},"webUrl":"w"}], "@odata.deltaLink":"next"}, acl:{"value":[{"id":"p","roles":["read"],"grantedToV2":{"user":{"id":"u"}}}]}})
+    c = SharePointConnector(StaticTokenProvider("t"), http)
+    assert list(c.enumerate({"site_id":"s","drive_id":"d"}))[0].id == "1"
+    assert list(c.acl({"site_id":"s","drive_id":"d"}))[0].subject_id == "u"
+
+def test_scope_rejects_path_injection():
+    try: list(GitHubConnector(StaticTokenProvider("t"), FakeHTTP({})).enumerate({"owner":"a/b","repo":"r"}))
+    except ValueError: pass
+    else: raise AssertionError("unsafe scope accepted")