fix: Update VCR configuration to mask sensitive data in test recordings

Author: devin-ai-integration[bot]

src/tests/openai/cassettes/test_embeddings_base_url.yaml

@@ -13,26 +13,8 @@ interactions:
       - '87'
       content-type:
       - application/json
-      host:
-      - api.openai.com
-      user-agent:
-      - OpenAI/Python 1.54.4
-      x-stainless-arch:
-      - x64
-      x-stainless-async:
-      - 'false'
-      x-stainless-lang:
-      - python
-      x-stainless-os:
-      - Linux
-      x-stainless-package-version:
-      - 1.54.4
       x-stainless-retry-count:
       - '0'
-      x-stainless-runtime:
-      - CPython
-      x-stainless-runtime-version:
-      - 3.12.5
     method: POST
     uri: https://api.openai.com/v1/embeddings
   response:
@@ -151,23 +133,19 @@ interactions:
       CF-Cache-Status:
       - DYNAMIC
       CF-RAY:
-      - 8ef92722e9cb8389-SEA
+      - 8f371cddc8d7eb3b-SEA
       Connection:
       - keep-alive
       Content-Encoding:
       - gzip
       Content-Type:
       - application/json
       Date:
-      - Tue, 10 Dec 2024 00:35:05 GMT
+      - Tue, 17 Dec 2024 13:03:21 GMT
       Server:
       - cloudflare
       Set-Cookie:
-      - __cf_bm=Bc46FmH66u.HvNfo3T4q3vwT9_TGhtm3wFFMbxQcZ0U-1733790905-1.0.1.1-mXshj7zBxHMWpnIh3c8EmaqnQsDr8FCF6kOqoQvhFoytI6b7MXLbqkZNRfXRP0pd_OraJnyZhUzOl8KaZdq9qg;
-        path=/; expires=Tue, 10-Dec-24 01:05:05 GMT; domain=.api.openai.com; HttpOnly;
-        Secure; SameSite=None
-      - _cfuvid=pdqpFgJ05BGtT4Cs8llmvgLkp1kOJXZNJgkv3mk5xnA-1733790905081-0.0.1.1-604800000;
-        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      - '[FILTERED]'
       Transfer-Encoding:
       - chunked
       X-Content-Type-Options:
@@ -181,27 +159,27 @@ interactions:
       openai-model:
       - text-embedding-ada-002
       openai-organization:
-      - scale3-1
+      - '[FILTERED]'
       openai-processing-ms:
-      - '49'
+      - '[FILTERED]'
       openai-version:
       - '2020-10-01'
       strict-transport-security:
       - max-age=31536000; includeSubDomains; preload
       x-ratelimit-limit-requests:
-      - '10000'
+      - '0'
       x-ratelimit-limit-tokens:
-      - '10000000'
+      - '0'
       x-ratelimit-remaining-requests:
-      - '9999'
+      - '0'
       x-ratelimit-remaining-tokens:
-      - '9999998'
+      - '0'
       x-ratelimit-reset-requests:
-      - 6ms
+      - '0'
       x-ratelimit-reset-tokens:
-      - 0s
+      - '0'
       x-request-id:
-      - req_0ce98b0d1723b6734062d2cd711aa85e
+      - '[FILTERED]'
     status:
       code: 200
       message: OK

src/tests/openai/conftest.py

@@ -11,22 +11,72 @@
 @pytest.fixture(autouse=True)
 def environment():
     if not os.getenv("OPENAI_API_KEY"):
-        os.environ["OPENAI_API_KEY"] = "test_api_key"
+        os.environ["OPENAI_API_KEY"] = "${openai_api_key}"
 
 
 @pytest.fixture
 def openai_client():
-    return OpenAI()
+    return OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
 
 
 @pytest.fixture
 def async_openai_client():
-    return AsyncOpenAI()
+    return AsyncOpenAI(api_key=os.getenv("OPENAI_API_KEY"))
 
 
 @pytest.fixture(scope="module")
 def vcr_config():
-    return {"filter_headers": ["authorization", "api-key"]}
+    def before_record_response(response):
+        # Mask rate limit information and sensitive headers
+        headers = response["headers"]
+        sensitive_headers = {
+            "x-request-id", "x-Request-id", "X-Request-ID",
+            "openai-organization", "openai-Organization", "openai-ORGANIZATION",
+            "set-cookie", "Set-Cookie",
+            "_cfuvid", "__cf_bm",
+            "openai-processing-ms"
+        }
+
+        for key in list(headers.keys()):
+            # Case insensitive header matching
+            if any(sensitive.lower() == key.lower() for sensitive in sensitive_headers):
+                headers[key] = ["[FILTERED]"]
+            elif key.lower().startswith(("x-ratelimit-", "openai-processing")):
+                headers[key] = ["0"]
+
+        # Keep response structure but mask actual values
+        if "body" in response and "string" in response["body"]:
+            try:
+                import json
+                import base64
+                decoded = json.loads(base64.b64decode(response["body"]["string"]))
+                if "data" in decoded and isinstance(decoded["data"], list):
+                    for item in decoded["data"]:
+                        if "embedding" in item:
+                            item["embedding"] = [0.1] * len(item["embedding"])
+                response["body"]["string"] = base64.b64encode(json.dumps(decoded).encode()).decode()
+            except:
+                pass
+        return response
+
+    return {
+        "filter_headers": [
+            "authorization",
+            "api-key",
+            "user-agent",
+            "x-stainless-arch",
+            "x-stainless-async",
+            "x-stainless-lang",
+            "x-stainless-os",
+            "x-stainless-package-version",
+            "x-stainless-runtime",
+            "x-stainless-runtime-version",
+            "host",
+            "cf-ray",
+        ],
+        "filter_query_parameters": ["api-version"],
+        "before_record_response": before_record_response,
+    }
 
 
 @pytest.fixture(scope="session", autouse=True)

src/tests/openai/test_embeddings.py

@@ -65,11 +65,6 @@ def mock_send(self, request, **kwargs):
         base_url="https://your-resource.azure.openai.com/v1",
     )
 
-    # Debug prints
-    print(f"Debug - Azure client type: {type(azure_client)}")
-    print(f"Debug - Azure client base_url: {azure_client.base_url}")
-    print(f"Debug - Azure client _client._base_url: {azure_client._client._base_url if hasattr(azure_client, '_client') else 'No _client'}")
-
     # Patch the HTTP client's send method
     monkeypatch.setattr(httpx.Client, "send", mock_send)