Fix tests

Author: pamelafox

app/backend/app.py

@@ -362,7 +362,7 @@ async def upload(auth_claims: dict[str, Any]):
     adls_manager = AdlsBlobManager(current_app.config[CONFIG_USER_BLOB_CONTAINER_CLIENT])
     file_url = await adls_manager.upload_blob(file, file.filename, user_oid)
     ingester: UploadUserFileStrategy = current_app.config[CONFIG_INGESTER]
-    await ingester.add_file(File(content=file, url=file_url), user_oid=user_oid)
+    await ingester.add_file(File(content=file, url=file_url, acls={"oids": [user_oid]}), user_oid=user_oid)
     return jsonify({"message": "File uploaded successfully"}), 200
 
 

tests/test_chatapproach.py

@@ -46,8 +46,6 @@ def chat_approach():
         query_language="en-us",
         query_speller="lexicon",
         prompt_manager=PromptyManager(),
-        blob_container_client=None,
-        datalake_client=None,
     )
 
 
@@ -199,8 +197,6 @@ async def test_search_results_filtering_by_scores(
         query_language="en-us",
         query_speller="lexicon",
         prompt_manager=PromptyManager(),
-        blob_container_client=None,
-        datalake_client=None,
     )
 
     monkeypatch.setattr(SearchClient, "search", mock_search)
@@ -244,8 +240,6 @@ async def test_search_results_query_rewriting(monkeypatch):
         query_language="en-us",
         query_speller="lexicon",
         prompt_manager=PromptyManager(),
-        blob_container_client=None,
-        datalake_client=None,
     )
 
     query_rewrites = None
@@ -293,8 +287,6 @@ async def test_agent_retrieval_results(monkeypatch):
         query_language="en-us",
         query_speller="lexicon",
         prompt_manager=PromptyManager(),
-        blob_container_client=None,
-        datalake_client=None,
     )
 
     agent_client = KnowledgeAgentRetrievalClient(endpoint="", agent_name="", credential=AzureKeyCredential(""))

tests/test_pdfparser.py

@@ -117,7 +117,7 @@ async def test_process_figure_without_bounding_regions():
     media_describer = MagicMock()
 
     result = await DocumentAnalysisParser.process_figure(doc, figure, media_describer)
-    expected_html = "<figure><figcaption></figcaption></figure>"
+    expected_html = "<figure><figcaption>1 </figcaption></figure>"
 
     assert isinstance(result, ImageOnPage)
     assert result.description == expected_html
@@ -128,7 +128,7 @@ async def test_process_figure_with_bounding_regions(monkeypatch, caplog):
     doc = MagicMock()
     figure = DocumentFigure(
         id="1",
-        caption=DocumentCaption(content="Figure 1"),
+        caption=DocumentCaption(content="Logo"),
         bounding_regions=[
             BoundingRegion(page_number=1, polygon=[1.4703, 2.8371, 5.5409, 2.8415, 5.5381, 6.6022, 1.4681, 6.5978]),
             BoundingRegion(page_number=2, polygon=[1.4703, 2.8371, 5.5409, 2.8415, 5.5381, 6.6022, 1.4681, 6.5978]),
@@ -151,7 +151,7 @@ def mock_crop_image_from_pdf_page(doc, page_number, bounding_box):
 
     with caplog.at_level(logging.WARNING):
         result = await DocumentAnalysisParser.process_figure(doc, figure, media_describer)
-        expected_html = "<figure><figcaption>Figure 1<br>Described Image</figcaption></figure>"
+        expected_html = "<figure><figcaption>1 Logo<br>Described Image</figcaption></figure>"
 
         assert isinstance(result, ImageOnPage)
         assert result.description == expected_html
@@ -325,7 +325,7 @@ async def mock_describe_image(self, image_bytes):
     assert pages[0].offset == 0
     assert (
         pages[0].text
-        == "# Simple Figure\n\nThis text is before the figure and NOT part of it.\n\n\n<figure><figcaption>Figure 1<br>Pie chart</figcaption></figure>\n\n\nThis is text after the figure that's not part of it."
+        == "# Simple Figure\n\nThis text is before the figure and NOT part of it.\n\n\n<figure><figcaption>1.1 Figure 1<br>Pie chart</figcaption></figure>\n\n\nThis is text after the figure that's not part of it."
     )
 
 

tests/test_upload.py

@@ -53,7 +53,6 @@ def mock_directory_get_file_client(self, *args, **kwargs):
 
     async def mock_upload_file(self, *args, **kwargs):
         assert kwargs.get("overwrite") is True
-        assert kwargs.get("metadata") == {"UploadedBy": "OID_X"}
         return None
 
     monkeypatch.setattr(DataLakeFileClient, "upload_data", mock_upload_file)
@@ -216,13 +215,20 @@ async def mock_search(self, *args, **kwargs):
     monkeypatch.setattr(SearchClient, "search", mock_search)
 
     deleted_documents = []
+    deleted_directories = []
 
     async def mock_delete_documents(self, documents):
         deleted_documents.extend(documents)
         return documents
 
     monkeypatch.setattr(SearchClient, "delete_documents", mock_delete_documents)
 
+    async def mock_delete_directory(self):
+        deleted_directories.append("mock_directory_url")
+        return None
+
+    monkeypatch.setattr(DataLakeDirectoryClient, "delete_directory", mock_delete_directory)
+
     response = await auth_client.post(
         "/delete_uploaded", headers={"Authorization": "Bearer test"}, json={"filename": "a's doc.txt"}
     )
@@ -231,3 +237,4 @@ async def mock_delete_documents(self, documents):
     assert searched_filters[0] == "sourcefile eq 'a''s doc.txt'"
     assert len(deleted_documents) == 1, "It should have only deleted the document solely owned by OID_X"
     assert deleted_documents[0]["id"] == "file-a_txt-7465737420646F63756D656E742E706466"
+    assert len(deleted_directories) == 1, "It should have deleted the directory for the file"

todo.txt

@@ -29,8 +29,3 @@ To decide:
 
 Later:
 Agentic: Incompatible since it doesnt retrieve images. We would need to do a follow-up search query to get each document, like filter: id eq 'x' or id eq 'y' or....
-
-
-"id":
-"feb5e192afb6_aHR0cHM6Ly9zdHh4azRxenEzdGFoaWMyLmJsb2IuY29yZS53aW5kb3dzLm5ldC9jb250ZW50L05vcnRod2luZF9IZWFsdGhfUGx1c19CZW5lZml0c19EZXRhaWxzLnBkZg2_pages_65",
-"aHR0cHM6Ly9zdHh4azRxenEzdGFoaWMyLmJsb2IuY29yZS53aW5kb3dzLm5ldC9jb250ZW50L05vcnRod2luZF9IZWFsdGhfUGx1c19CZW5lZml0c19EZXRhaWxzLnBkZg2",