ci: fix tests by adding dynamic retry

Author: wochinge

pyproject.toml

@@ -34,6 +34,7 @@ dev = [
     "langgraph>=1,<2",
     "autoevals>=0.0.130,<0.1",
     "opentelemetry-instrumentation-threading>=0.59b0,<1",
+    "tenacity>=9.1.4",
 ]
 docs = [
     "pdoc>=15.0.4,<16",

tests/test_core_sdk.py

@@ -5,6 +5,7 @@
 from time import sleep
 
 import pytest
+from tenacity import Retrying, stop_after_delay, wait_fixed
 
 from langfuse import Langfuse, propagate_attributes
 from langfuse._client.resource_manager import LangfuseResourceManager
@@ -15,6 +16,14 @@
     get_api,
 )
 
+LANGFUSE_SECRET_KEY = "sk-lf-04e91827-fcd4-457e-8e51-6cde2417bc67"
+LANGFUSE_PUBLIC_KEY = "pk-lf-33e3a6d2-2cac-48f6-8d00-5ccc04cbe804"
+LANGFUSE_BASE_URL = "http://localhost:3000"
+
+os.environ["LANGFUSE_SECRET_KEY"] = LANGFUSE_SECRET_KEY
+os.environ["LANGFUSE_PUBLIC_KEY"] = LANGFUSE_PUBLIC_KEY
+os.environ["LANGFUSE_BASE_URL"] = LANGFUSE_BASE_URL
+
 
 @pytest.mark.asyncio
 async def test_concurrency():
@@ -337,7 +346,6 @@ def test_create_text_score():
 
     # Ensure data is sent
     langfuse.flush()
-    sleep(2)
 
     # Create a text score
     score_id = create_uuid()
@@ -360,23 +368,26 @@ def test_create_text_score():
 
     # Ensure data is sent
     langfuse.flush()
-    sleep(2)
 
-    # Retrieve and verify
-    trace = api_wrapper.get_trace(trace_id)
+    # Retrieve and verify with retry
+    for attempt in Retrying(
+        stop=stop_after_delay(10), wait=wait_fixed(0.1), reraise=True
+    ):
+        with attempt:
+            trace = api_wrapper.get_trace(trace_id)
 
-    # Find the score we created by name
-    created_score = next(
-        (s for s in trace["scores"] if s["name"] == "this-is-a-score"), None
-    )
-    assert created_score is not None, "Score not found in trace"
-    assert created_score["id"] == score_id
-    assert created_score["dataType"] == "TEXT"
-    assert created_score["value"] is None
-    assert (
-        created_score["stringValue"]
-        == "This is a detailed text evaluation of the output quality."
-    )
+            # Find the score we created by name
+            created_score = next(
+                (s for s in trace["scores"] if s["name"] == "this-is-a-score"), None
+            )
+            assert created_score is not None, "Score not found in trace"
+            assert created_score["id"] == score_id
+            assert created_score["dataType"] == "TEXT"
+
+            assert (
+                created_score["stringValue"]
+                == "This is a detailed text evaluation of the output quality."
+            )
 
 
 def test_create_score_with_custom_timestamp():