langfuse-haystack 0.1.0__tar.gz → 0.3.0__tar.gz

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/.gitignore

@@ -135,3 +135,12 @@ dmypy.json
 # Docs generation artifacts
 _readme_*.md
 .idea
+
+# macOS
+.DS_Store
+
+# http cache (requests-cache)
+**/http_cache.sqlite
+
+# ruff
+.ruff_cache

langfuse_haystack-0.3.0/CHANGELOG.md

@@ -0,0 +1,19 @@
+# Changelog
+
+## [integrations/langfuse-v0.2.0] - 2024-06-18
+
+## [integrations/langfuse-v0.1.0] - 2024-06-13
+
+### 🚀 Features
+
+- Langfuse integration (#686)
+
+### 🐛 Bug Fixes
+
+- Performance optimizations and value error when streaming in langfuse (#798)
+
+### ⚙️ Miscellaneous Tasks
+
+- Use ChatMessage to_openai_format, update unit tests, pydocs (#725)
+
+<!-- generated by git-cliff -->

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: langfuse-haystack
-Version: 0.1.0
+Version: 0.3.0
 Project-URL: Documentation, https://github.com/deepset-ai/haystack-core-integrations/tree/main/integrations/langfuse#readme
 Project-URL: Issues, https://github.com/deepset-ai/haystack-core-integrations/issues
 Project-URL: Source, https://github.com/deepset-ai/haystack-core-integrations/tree/main/integrations/langfuse
@@ -50,7 +50,33 @@ To enable tracing in your Haystack pipeline, add the `LangfuseConnector` to your
 You also need to set the `LANGFUSE_SECRET_KEY` and `LANGFUSE_PUBLIC_KEY` environment variables in order to connect to Langfuse account.
 You can get these keys by signing up for an account on the Langfuse website.
 
-Here's an example:
+⚠️ **Important:** To ensure proper tracing, always set environment variables before importing any Haystack components. This is crucial because Haystack initializes its internal tracing components during import.
+
+Here's the correct way to set up your script:
+
+```python
+import os
+
+# Set environment variables first
+os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
+os.environ["TOKENIZERS_PARALLELISM"] = "false"
+os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
+
+# Then import Haystack components
+from haystack.components.builders import ChatPromptBuilder
+from haystack.components.generators.chat import OpenAIChatGenerator
+from haystack.dataclasses import ChatMessage
+from haystack import Pipeline
+
+from haystack_integrations.components.connectors.langfuse import LangfuseConnector
+
+# Rest of your code...
+```
+
+Alternatively, an even better practice is to set these environment variables in your shell before running the script.
+
+
+Here's a full example:
 
 ```python
 import os
@@ -59,7 +85,7 @@ os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
 os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
 
-from haystack.components.builders import DynamicChatPromptBuilder
+from haystack.components.builders import ChatPromptBuilder
 from haystack.components.generators.chat import OpenAIChatGenerator
 from haystack.dataclasses import ChatMessage
 from haystack import Pipeline
@@ -69,7 +95,7 @@ from haystack_integrations.components.connectors.langfuse import LangfuseConnect
 if __name__ == "__main__":
     pipe = Pipeline()
     pipe.add_component("tracer", LangfuseConnector("Chat example"))
-    pipe.add_component("prompt_builder", DynamicChatPromptBuilder())
+    pipe.add_component("prompt_builder", ChatPromptBuilder())
     pipe.add_component("llm", OpenAIChatGenerator(model="gpt-3.5-turbo"))
 
     pipe.connect("prompt_builder.prompt", "llm.messages")
@@ -80,7 +106,7 @@ if __name__ == "__main__":
     ]
 
     response = pipe.run(
-        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "prompt_source": messages}}
+        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "template": messages}}
     )
     print(response["llm"]["replies"][0])
     print(response["tracer"]["trace_url"])

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/README.md

@@ -27,7 +27,33 @@ To enable tracing in your Haystack pipeline, add the `LangfuseConnector` to your
 You also need to set the `LANGFUSE_SECRET_KEY` and `LANGFUSE_PUBLIC_KEY` environment variables in order to connect to Langfuse account.
 You can get these keys by signing up for an account on the Langfuse website.
 
-Here's an example:
+⚠️ **Important:** To ensure proper tracing, always set environment variables before importing any Haystack components. This is crucial because Haystack initializes its internal tracing components during import.
+
+Here's the correct way to set up your script:
+
+```python
+import os
+
+# Set environment variables first
+os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
+os.environ["TOKENIZERS_PARALLELISM"] = "false"
+os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
+
+# Then import Haystack components
+from haystack.components.builders import ChatPromptBuilder
+from haystack.components.generators.chat import OpenAIChatGenerator
+from haystack.dataclasses import ChatMessage
+from haystack import Pipeline
+
+from haystack_integrations.components.connectors.langfuse import LangfuseConnector
+
+# Rest of your code...
+```
+
+Alternatively, an even better practice is to set these environment variables in your shell before running the script.
+
+
+Here's a full example:
 
 ```python
 import os
@@ -36,7 +62,7 @@ os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
 os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
 
-from haystack.components.builders import DynamicChatPromptBuilder
+from haystack.components.builders import ChatPromptBuilder
 from haystack.components.generators.chat import OpenAIChatGenerator
 from haystack.dataclasses import ChatMessage
 from haystack import Pipeline
@@ -46,7 +72,7 @@ from haystack_integrations.components.connectors.langfuse import LangfuseConnect
 if __name__ == "__main__":
     pipe = Pipeline()
     pipe.add_component("tracer", LangfuseConnector("Chat example"))
-    pipe.add_component("prompt_builder", DynamicChatPromptBuilder())
+    pipe.add_component("prompt_builder", ChatPromptBuilder())
     pipe.add_component("llm", OpenAIChatGenerator(model="gpt-3.5-turbo"))
 
     pipe.connect("prompt_builder.prompt", "llm.messages")
@@ -57,7 +83,7 @@ if __name__ == "__main__":
     ]
 
     response = pipe.run(
-        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "prompt_source": messages}}
+        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "template": messages}}
     )
     print(response["llm"]["replies"][0])
     print(response["tracer"]["trace_url"])

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/example/chat.py

@@ -3,7 +3,7 @@ import os
 os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
 
 from haystack import Pipeline
-from haystack.components.builders import DynamicChatPromptBuilder
+from haystack.components.builders import ChatPromptBuilder
 from haystack.components.generators.chat import OpenAIChatGenerator
 from haystack.dataclasses import ChatMessage
 from haystack_integrations.components.connectors.langfuse import LangfuseConnector
@@ -12,7 +12,7 @@ if __name__ == "__main__":
 
     pipe = Pipeline()
     pipe.add_component("tracer", LangfuseConnector("Chat example"))
-    pipe.add_component("prompt_builder", DynamicChatPromptBuilder())
+    pipe.add_component("prompt_builder", ChatPromptBuilder())
     pipe.add_component("llm", OpenAIChatGenerator(model="gpt-3.5-turbo"))
 
     pipe.connect("prompt_builder.prompt", "llm.messages")
@@ -22,8 +22,6 @@ if __name__ == "__main__":
         ChatMessage.from_user("Tell me about {{location}}"),
     ]
 
-    response = pipe.run(
-        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "prompt_source": messages}}
-    )
+    response = pipe.run(data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "template": messages}})
     print(response["llm"]["replies"][0])
     print(response["tracer"]["trace_url"])

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/pyproject.toml

@@ -10,9 +10,7 @@ readme = "README.md"
 requires-python = ">=3.8"
 license = "Apache-2.0"
 keywords = []
-authors = [
-  { name = "deepset GmbH", email = "info@deepset.ai" },
-]
+authors = [{ name = "deepset GmbH", email = "info@deepset.ai" }]
 classifiers = [
   "Development Status :: 4 - Beta",
   "Programming Language :: Python",
@@ -24,10 +22,7 @@ classifiers = [
   "Programming Language :: Python :: Implementation :: CPython",
   "Programming Language :: Python :: Implementation :: PyPy",
 ]
-dependencies = [
-  "haystack-ai>=2.1.0",
-  "langfuse"
-]
+dependencies = ["haystack-ai>=2.1.0", "langfuse"]
 
 [project.urls]
 Documentation = "https://github.com/deepset-ai/haystack-core-integrations/tree/main/integrations/langfuse#readme"
@@ -50,19 +45,16 @@ git_describe_command = 'git describe --tags --match="integrations/langfuse-v[0-9
 dependencies = [
   "coverage[toml]>=6.5",
   "pytest",
+  "pytest-rerunfailures",
   "haystack-pydoc-tools",
 ]
 [tool.hatch.envs.default.scripts]
 test = "pytest {args:tests}"
 test-cov = "coverage run -m pytest {args:tests}"
-cov-report = [
-  "- coverage combine",
-  "coverage report",
-]
-cov = [
-  "test-cov",
-  "cov-report",
-]
+test-cov-retry = "test-cov --reruns 3 --reruns-delay 30 -x"
+cov-report = ["- coverage combine", "coverage report"]
+cov = ["test-cov", "cov-report"]
+cov-retry = ["test-cov-retry", "cov-report"]
 docs = ["pydoc-markdown pydoc/config.yml"]
 
 
@@ -75,8 +67,8 @@ dependencies = ["black>=23.1.0", "mypy>=1.0.0", "ruff>=0.0.243"]
 
 [tool.hatch.envs.lint.scripts]
 typing = "mypy --install-types --non-interactive --explicit-package-bases {args:src/ tests}"
-style = ["ruff check {args:.}", "black --check --diff {args:.}"]
-fmt = ["black {args:.}", "ruff --fix {args:.}", "style"]
+style = ["ruff check {args:. --exclude tests/}", "black --check --diff {args:.}"]
+fmt = ["black {args:.}", "ruff --fix {args:. --exclude tests/}", "style"]
 all = ["style", "typing"]
 
 [tool.hatch.metadata]
@@ -160,12 +152,8 @@ parallel = false
 
 [tool.coverage.report]
 omit = ["*/tests/*", "*/__init__.py"]
-show_missing=true
-exclude_lines = [
-  "no cov",
-  "if __name__ == .__main__.:",
-  "if TYPE_CHECKING:",
-]
+show_missing = true
+exclude_lines = ["no cov", "if __name__ == .__main__.:", "if TYPE_CHECKING:"]
 
 [[tool.mypy.overrides]]
 module = [
@@ -179,7 +167,5 @@ ignore_missing_imports = true
 
 [tool.pytest.ini_options]
 addopts = "--strict-markers"
-markers = [
-  "integration: integration tests",
-]
+markers = ["integration: integration tests"]
 log_cli = true

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/src/haystack_integrations/components/connectors/langfuse/langfuse_connector.py

@@ -1,14 +1,14 @@
 from haystack import component, tracing
-from haystack_integrations.tracing.langfuse import LangfuseTracer
 
+from haystack_integrations.tracing.langfuse import LangfuseTracer
 from langfuse import Langfuse
 
 
 @component
 class LangfuseConnector:
     """
-    LangfuseConnector connects Haystack LLM framework with Langfuse in order to enable the tracing of operations
-    and data flow within various components of a pipeline.
+    LangfuseConnector connects Haystack LLM framework with [Langfuse](https://langfuse.com) in order to enable the
+    tracing of operations and data flow within various components of a pipeline.
 
     Simply add this component to your pipeline, but *do not* connect it to any other component. The LangfuseConnector
     will automatically trace the operations and data flow within the pipeline.
@@ -40,6 +40,7 @@ class LangfuseConnector:
 
     # ...
 
+
     @app.on_event("shutdown")
     async def shutdown_event():
         tracer.actual_tracer.flush()
@@ -53,27 +54,35 @@ class LangfuseConnector:
     os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"
 
     from haystack import Pipeline
-    from haystack.components.builders import DynamicChatPromptBuilder
+    from haystack.components.builders import ChatPromptBuilder
     from haystack.components.generators.chat import OpenAIChatGenerator
     from haystack.dataclasses import ChatMessage
-    from haystack_integrations.components.connectors.langfuse import LangfuseConnector
+    from haystack_integrations.components.connectors.langfuse import (
+        LangfuseConnector,
+    )
 
     if __name__ == "__main__":
-
         pipe = Pipeline()
         pipe.add_component("tracer", LangfuseConnector("Chat example"))
-        pipe.add_component("prompt_builder", DynamicChatPromptBuilder())
+        pipe.add_component("prompt_builder", ChatPromptBuilder())
         pipe.add_component("llm", OpenAIChatGenerator(model="gpt-3.5-turbo"))
 
         pipe.connect("prompt_builder.prompt", "llm.messages")
 
         messages = [
-            ChatMessage.from_system("Always respond in German even if some input data is in other languages."),
+            ChatMessage.from_system(
+                "Always respond in German even if some input data is in other languages."
+            ),
             ChatMessage.from_user("Tell me about {{location}}"),
         ]
 
         response = pipe.run(
-            data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "prompt_source": messages}}
+            data={
+                "prompt_builder": {
+                    "template_variables": {"location": "Berlin"},
+                    "template": messages,
+                }
+            }
         )
         print(response["llm"]["replies"][0])
         print(response["tracer"]["trace_url"])

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/src/haystack_integrations/tracing/langfuse/tracer.py

@@ -2,6 +2,7 @@ import contextlib
 import os
 from typing import Any, Dict, Iterator, Optional, Union
 
+from haystack.components.generators.openai_utils import _convert_message_to_openai_format
 from haystack.dataclasses import ChatMessage
 from haystack.tracing import Span, Tracer, tracer
 from haystack.tracing import utils as tracing_utils
@@ -9,6 +10,9 @@ from haystack.tracing import utils as tracing_utils
 import langfuse
 
 HAYSTACK_LANGFUSE_ENFORCE_FLUSH_ENV_VAR = "HAYSTACK_LANGFUSE_ENFORCE_FLUSH"
+_SUPPORTED_GENERATORS = ["AzureOpenAIGenerator", "OpenAIGenerator"]
+_SUPPORTED_CHAT_GENERATORS = ["AzureOpenAIChatGenerator", "OpenAIChatGenerator"]
+_ALL_SUPPORTED_GENERATORS = _SUPPORTED_GENERATORS + _SUPPORTED_CHAT_GENERATORS
 
 
 class LangfuseSpan(Span):
@@ -48,14 +52,14 @@ class LangfuseSpan(Span):
             return
         if key.endswith(".input"):
             if "messages" in value:
-                messages = [m.to_openai_format() for m in value["messages"]]
+                messages = [_convert_message_to_openai_format(m) for m in value["messages"]]
                 self._span.update(input=messages)
             else:
                 self._span.update(input=value)
         elif key.endswith(".output"):
             if "replies" in value:
                 if all(isinstance(r, ChatMessage) for r in value["replies"]):
-                    replies = [m.to_openai_format() for m in value["replies"]]
+                    replies = [_convert_message_to_openai_format(m) for m in value["replies"]]
                 else:
                     replies = value["replies"]
                 self._span.update(output=replies)
@@ -109,7 +113,7 @@ class LangfuseTracer(Tracer):
         tags = tags or {}
         span_name = tags.get("haystack.component.name", operation_name)
 
-        if tags.get("haystack.component.type") in ["OpenAIGenerator", "OpenAIChatGenerator"]:
+        if tags.get("haystack.component.type") in _ALL_SUPPORTED_GENERATORS:
             span = LangfuseSpan(self.current_span().raw_span().generation(name=span_name))
         else:
             span = LangfuseSpan(self.current_span().raw_span().span(name=span_name))
@@ -119,14 +123,14 @@ class LangfuseTracer(Tracer):
 
         yield span
 
-        if tags.get("haystack.component.type") == "OpenAIGenerator":
+        if tags.get("haystack.component.type") in _SUPPORTED_GENERATORS:
             meta = span._data.get("haystack.component.output", {}).get("meta")
             if meta:
                 # Haystack returns one meta dict for each message, but the 'usage' value
                 # is always the same, let's just pick the first item
                 m = meta[0]
                 span._span.update(usage=m.get("usage") or None, model=m.get("model"))
-        elif tags.get("haystack.component.type") == "OpenAIChatGenerator":
+        elif tags.get("haystack.component.type") in _SUPPORTED_CHAT_GENERATORS:
             replies = span._data.get("haystack.component.output", {}).get("replies")
             if replies:
                 meta = replies[0].meta

{langfuse_haystack-0.1.0 → langfuse_haystack-0.3.0}/tests/test_tracing.py

@@ -9,7 +9,7 @@ import pytest
 import requests
 
 from haystack import Pipeline
-from haystack.components.builders import DynamicChatPromptBuilder
+from haystack.components.builders import ChatPromptBuilder
 from haystack.components.generators.chat import OpenAIChatGenerator
 from haystack.dataclasses import ChatMessage
 from requests.auth import HTTPBasicAuth
@@ -26,7 +26,7 @@ def test_tracing_integration():
 
     pipe = Pipeline()
     pipe.add_component("tracer", LangfuseConnector(name="Chat example", public=True))  # public so anyone can verify run
-    pipe.add_component("prompt_builder", DynamicChatPromptBuilder())
+    pipe.add_component("prompt_builder", ChatPromptBuilder())
     pipe.add_component("llm", OpenAIChatGenerator(model="gpt-3.5-turbo"))
 
     pipe.connect("prompt_builder.prompt", "llm.messages")
@@ -36,9 +36,7 @@ def test_tracing_integration():
         ChatMessage.from_user("Tell me about {{location}}"),
     ]
 
-    response = pipe.run(
-        data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "prompt_source": messages}}
-    )
+    response = pipe.run(data={"prompt_builder": {"template_variables": {"location": "Berlin"}, "template": messages}})
     assert "Berlin" in response["llm"]["replies"][0].content
     assert response["tracer"]["trace_url"]
     url = "https://cloud.langfuse.com/api/public/traces/"