ragaai-catalyst 2.1.7b0__py3-none-any.whl → 2.1.7.1__py3-none-any.whl

ragaai_catalyst/tracers/exporters/dynamic_trace_exporter.py

@@ -14,7 +14,7 @@ class DynamicTraceExporter(SpanExporter):
     certain properties to be updated dynamically during execution.
     """
     
-    def __init__(self, tracer_type, files_to_zip, project_name, project_id, dataset_name, user_details, base_url, custom_model_cost, timeout=120, post_processor = None):
+    def __init__(self, tracer_type, files_to_zip, project_name, project_id, dataset_name, user_details, base_url, custom_model_cost, timeout=120, post_processor = None, max_upload_workers = 30,user_context = None, external_id=None):
         """
         Initialize the DynamicTraceExporter.
         
@@ -27,6 +27,7 @@ class DynamicTraceExporter(SpanExporter):
             user_details: User details
             base_url: Base URL for API
             post_processor: Post processing function before uploading trace
+            max_upload_workers: Maximum number of upload workers
         """
         self._exporter = RAGATraceExporter(
             tracer_type=tracer_type,
@@ -38,7 +39,10 @@ class DynamicTraceExporter(SpanExporter):
             base_url=base_url,
             custom_model_cost=custom_model_cost,
             timeout=timeout,
-            post_processor= post_processor
+            post_processor= post_processor,
+            max_upload_workers = max_upload_workers,
+            user_context = user_context,
+            external_id=external_id
         )
         
         # Store the initial values
@@ -50,6 +54,8 @@ class DynamicTraceExporter(SpanExporter):
         self._base_url = base_url
         self._custom_model_cost = custom_model_cost
         self._post_processor = post_processor
+        self._max_upload_workers = max_upload_workers
+        self._user_context = user_context
 
     
     def export(self, spans):
@@ -107,6 +113,8 @@ class DynamicTraceExporter(SpanExporter):
         self._exporter.base_url = self._base_url
         self._exporter.custom_model_cost = self._custom_model_cost
         self._exporter.post_processor = self._post_processor
+        self._exporter.max_upload_workers = self._max_upload_workers
+        self._exporter.user_context = self._user_context
     
     # Getter and setter methods for dynamic properties
     
@@ -165,3 +173,20 @@ class DynamicTraceExporter(SpanExporter):
     @custom_model_cost.setter
     def custom_model_cost(self, value):
         self._custom_model_cost = value
+    
+    @property
+    def max_upload_workers(self):
+        return self._max_upload_workers
+    
+    @max_upload_workers.setter
+    def max_upload_workers(self, value):
+        self._max_upload_workers = value
+
+    @property
+    def user_context(self):
+        return self._user_context
+    
+    @user_context.setter
+    def user_context(self, value):
+        self._user_context = value
+

ragaai_catalyst/tracers/exporters/ragaai_trace_exporter.py

@@ -26,7 +26,7 @@ logging_level = (
 
 
 class RAGATraceExporter(SpanExporter):
-    def __init__(self, tracer_type, files_to_zip, project_name, project_id, dataset_name, user_details, base_url, custom_model_cost, timeout=120, post_processor = None):
+    def __init__(self, tracer_type, files_to_zip, project_name, project_id, dataset_name, user_details, base_url, custom_model_cost, timeout=120, post_processor = None, max_upload_workers = 30,user_context = None, external_id=None):
         self.trace_spans = dict()
         self.tmp_dir = tempfile.gettempdir()
         self.tracer_type = tracer_type
@@ -40,6 +40,9 @@ class RAGATraceExporter(SpanExporter):
         self.system_monitor = SystemMonitor(dataset_name)
         self.timeout = timeout
         self.post_processor = post_processor
+        self.max_upload_workers = max_upload_workers
+        self.user_context = user_context
+        self.external_id = external_id
 
     def export(self, spans):
         for span in spans:
@@ -85,12 +88,33 @@ class RAGATraceExporter(SpanExporter):
                 ragaai_trace_details = self.prepare_trace(spans, trace_id)
         except Exception as e:
             print(f"Error converting trace {trace_id}: {e}")
+            return  # Exit early if conversion fails
+            
+        # Check if trace details are None (conversion failed)
+        if ragaai_trace_details is None:
+            logger.error(f"Cannot upload trace {trace_id}: conversion failed and returned None")
+            return  # Exit early if conversion failed
+            
         # Upload the trace if upload_trace function is provided
         try:
             if self.post_processor!=None:
                 ragaai_trace_details['trace_file_path'] = self.post_processor(ragaai_trace_details['trace_file_path'])
             if self.tracer_type == "langchain":
-                asyncio.run(self.upload_rag_trace(ragaai_trace_details, additional_metadata, trace_id))
+                # Check if we're already in an event loop
+                try:
+                    loop = asyncio.get_event_loop()
+                    if loop.is_running():
+                        # We're in a running event loop (like in Colab/Jupyter)
+                        # Create a future and run the coroutine
+                        future = asyncio.ensure_future(self.upload_rag_trace(ragaai_trace_details, additional_metadata, trace_id))
+                        # We don't wait for it to complete as this would block the event loop
+                        logger.info(f"Scheduled async upload for trace {trace_id} in existing event loop")
+                    else:
+                        # No running event loop, use asyncio.run()
+                        asyncio.run(self.upload_rag_trace(ragaai_trace_details, additional_metadata, trace_id))
+                except RuntimeError:
+                    # No event loop exists, create one
+                    asyncio.run(self.upload_rag_trace(ragaai_trace_details, additional_metadata, trace_id))
             else:
                 self.upload_trace(ragaai_trace_details, trace_id)
         except Exception as e:
@@ -184,12 +208,14 @@ class RAGATraceExporter(SpanExporter):
     
     async def upload_rag_trace(self, ragaai_trace, additional_metadata, trace_id):
         try:
+            ragaai_trace[0]['external_id'] = self.external_id
             trace_file_path = os.path.join(self.tmp_dir, f"{trace_id}.json")
             with open(trace_file_path, 'w') as f:
                 json.dump(ragaai_trace, f, indent=2)
-            
+            logger.info(f"Trace file saved at {trace_file_path}")
+
             # Create a ThreadPoolExecutor with max_workers=30
-            with concurrent.futures.ThreadPoolExecutor(max_workers=30) as executor:
+            with concurrent.futures.ThreadPoolExecutor(max_workers=self.max_upload_workers) as executor:
                 # Create a partial function with all the necessary arguments
                 upload_func = partial(
                     UploadTraces(
@@ -232,7 +258,7 @@ class RAGATraceExporter(SpanExporter):
     
     def prepare_rag_trace(self, spans, trace_id):
         try:            
-            ragaai_trace, additional_metadata = rag_trace_json_converter(spans, self.custom_model_cost, trace_id, self.user_details, self.tracer_type)
+            ragaai_trace, additional_metadata = rag_trace_json_converter(spans, self.custom_model_cost, trace_id, self.user_details, self.tracer_type,self.user_context)
             ragaai_trace["metadata"]["recorded_on"] = datetime.datetime.now().astimezone().isoformat()
             ragaai_trace["metadata"]["log_source"] = "langchain_tracer"
 

ragaai_catalyst/tracers/tracer.py

@@ -24,6 +24,7 @@ from ragaai_catalyst.tracers.exporters.raga_exporter import RagaExporter
 from ragaai_catalyst.tracers.utils import get_unique_key
 # from ragaai_catalyst.tracers.llamaindex_callback import LlamaIndexTracer
 from ragaai_catalyst.tracers.llamaindex_instrumentation import LlamaIndexInstrumentationTracer
+from openinference.instrumentation.langchain import LangChainInstrumentor
 from ragaai_catalyst import RagaAICatalyst
 from ragaai_catalyst.tracers.agentic_tracing import AgenticTracing
 from ragaai_catalyst.tracers.agentic_tracing.tracers.llm_tracer import LLMTracerMixin
@@ -59,6 +60,8 @@ class Tracer(AgenticTracing):
         },
         interval_time=2,
         # auto_instrumentation=True/False  # to control automatic instrumentation of everything
+        max_upload_workers=30,
+        external_id=None
 
     ):
         """
@@ -141,6 +144,11 @@ class Tracer(AgenticTracing):
         self.user_context = ""  # Initialize user_context to store context from add_context
         self.file_tracker = TrackName()
         self.post_processor = None
+        self.max_upload_workers = max_upload_workers
+        self.user_details = self._pass_user_data()
+        self.update_llm_cost = update_llm_cost
+        self.auto_instrumentation = auto_instrumentation
+        self.external_id = external_id
         
         try:
             response = requests.get(
@@ -458,6 +466,30 @@ class Tracer(AgenticTracing):
             self.dynamic_exporter._exporter.post_processor = post_processor_func
             self.dynamic_exporter._post_processor = post_processor_func
         logger.info("Registered post process as: "+str(post_processor_func))
+
+    
+    def set_external_id(self, external_id):
+        current_params = {
+            'project_name': self.project_name,
+            'dataset_name': self.dataset_name,
+            'trace_name': self.trace_name,
+            'tracer_type': self.tracer_type,
+            'pipeline': self.pipeline,
+            'metadata': self.metadata,
+            'description': self.description,
+            'timeout': self.timeout,
+            'update_llm_cost': self.update_llm_cost,
+            'auto_instrumentation': self.auto_instrumentation,
+            'interval_time': self.interval_time,
+            'max_upload_workers': self.max_upload_workers
+        }
+
+        # Reinitialize self with new external_id and stored parameters
+        self.__init__(
+            external_id=external_id,
+            **current_params
+        )
+
     
 
     def set_dataset_name(self, dataset_name):
@@ -483,15 +515,20 @@ class Tracer(AgenticTracing):
             # Also update the user_details in the dynamic exporter
             self.dynamic_exporter.user_details = self.user_details
         else:
-            # Store current parameters
             current_params = {
-                'project_name': self.project_name,
-                'tracer_type': self.tracer_type,
-                'pipeline': self.pipeline,
-                'metadata': self.metadata,
-                'description': self.description,
-                'timeout': self.timeout
-            }
+            'project_name': self.project_name,
+            'trace_name': self.trace_name,
+            'tracer_type': self.tracer_type,
+            'pipeline': self.pipeline,
+            'metadata': self.metadata,
+            'description': self.description,
+            'timeout': self.timeout,
+            'update_llm_cost': self.update_llm_cost,
+            'auto_instrumentation': self.auto_instrumentation,
+            'interval_time': self.interval_time,
+            'max_upload_workers': self.max_upload_workers,
+            'external_id': self.external_id
+        }
             
             # Reinitialize self with new dataset_name and stored parameters
             self.__init__(
@@ -501,7 +538,7 @@ class Tracer(AgenticTracing):
 
     def _improve_metadata(self, metadata, tracer_type):
         if metadata is None:
-            metadata = {}
+            metadata = {"metadata": {}}
         metadata.setdefault("log_source", f"{tracer_type}_tracer")
         metadata.setdefault("recorded_on", str(datetime.datetime.now()))
         return metadata
@@ -763,7 +800,10 @@ class Tracer(AgenticTracing):
             base_url=self.base_url,
             custom_model_cost=self.model_custom_cost,
             timeout = self.timeout,
-            post_processor= self.post_processor
+            post_processor= self.post_processor,
+            max_upload_workers = self.max_upload_workers,
+            user_context = self.user_context,
+            external_id=self.external_id
         )
         
         # Set up tracer provider
@@ -772,7 +812,15 @@ class Tracer(AgenticTracing):
         
         # Instrument all specified instrumentors
         for instrumentor_class, args in instrumentors:
-            instrumentor_class().instrument(tracer_provider=tracer_provider, *args)
+            # Create an instance of the instrumentor
+            instrumentor = instrumentor_class()
+            
+            # Uninstrument only if it is already instrumented
+            if isinstance(instrumentor, LangChainInstrumentor) and instrumentor._is_instrumented_by_opentelemetry:
+                instrumentor.uninstrument()
+            
+            # Instrument with the provided tracer provider and arguments
+            instrumentor.instrument(tracer_provider=tracer_provider, *args)
             
     def update_file_list(self):
         """
@@ -807,6 +855,7 @@ class Tracer(AgenticTracing):
         
         # Convert string context to string if needed
         if isinstance(context, str):
+            self.dynamic_exporter.user_context = context
             self.user_context = context
         else:
             raise TypeError("context must be a string")

ragaai_catalyst/tracers/utils/rag_trace_json_converter.py

@@ -11,9 +11,8 @@ logging_level = (
     logger.setLevel(logging.DEBUG) if os.getenv("DEBUG") == "1" else logging.INFO
 )
 
-def rag_trace_json_converter(input_trace, custom_model_cost, trace_id, user_details, tracer_type):
+def rag_trace_json_converter(input_trace, custom_model_cost, trace_id, user_details, tracer_type,user_context):
     trace_aggregate = {}
-    
     def get_prompt(input_trace):
         if tracer_type == "langchain":
             for span in input_trace:
@@ -46,7 +45,9 @@ def rag_trace_json_converter(input_trace, custom_model_cost, trace_id, user_deta
         return None
     
     def get_context(input_trace):
-        if tracer_type == "langchain":
+        if user_context.strip():
+            return user_context
+        elif tracer_type == "langchain":
             for span in input_trace:
                 if span["name"] == "VectorStoreRetriever":
                     return span["attributes"].get("retrieval.documents.1.document.content")
@@ -82,6 +83,7 @@ def rag_trace_json_converter(input_trace, custom_model_cost, trace_id, user_deta
     else:
         additional_metadata = get_additional_metadata(input_trace, custom_model_cost, model_cost)
     
+    trace_aggregate["metadata"] = user_details.get("trace_user_detail", {}).get("metadata")
     trace_aggregate["metadata"].update(additional_metadata)
     additional_metadata.pop("total_cost")
     additional_metadata.pop("total_latency")

{ragaai_catalyst-2.1.7b0.dist-info → ragaai_catalyst-2.1.7.1.dist-info}/METADATA

@@ -1,9 +1,9 @@
 Metadata-Version: 2.4
 Name: ragaai_catalyst
-Version: 2.1.7b0
+Version: 2.1.7.1
 Summary: RAGA AI CATALYST
 Author-email: Kiran Scaria <kiran.scaria@raga.ai>, Kedar Gaikwad <kedar.gaikwad@raga.ai>, Dushyant Mahajan <dushyant.mahajan@raga.ai>, Siddhartha Kosti <siddhartha.kosti@raga.ai>, Ritika Goel <ritika.goel@raga.ai>, Vijay Chaurasia <vijay.chaurasia@raga.ai>, Tushar Kumar <tushar.kumar@raga.ai>
-Requires-Python: <3.13,>=3.10
+Requires-Python: <=3.13.2,>=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: aiohttp>=3.10.2
@@ -45,7 +45,7 @@ Requires-Dist: opentelemetry-sdk
 Requires-Dist: opentelemetry-exporter-otlp
 Requires-Dist: opentelemetry-proto>=1.12.0
 Provides-Extra: dev
-Requires-Dist: pytest; extra == "dev"
+Requires-Dist: pytest>=8.3.5; extra == "dev"
 Requires-Dist: pytest-cov; extra == "dev"
 Requires-Dist: black; extra == "dev"
 Requires-Dist: isort; extra == "dev"

{ragaai_catalyst-2.1.7b0.dist-info → ragaai_catalyst-2.1.7.1.dist-info}/RECORD

@@ -31,7 +31,7 @@ ragaai_catalyst/tracers/distributed.py,sha256=MwlBwIxCAng-OI-7Ove_rkE1mTLeuW4Jw-
 ragaai_catalyst/tracers/langchain_callback.py,sha256=CB75zzG3-DkYTELj0vI1MOHQTY0MuQJfoHIXz9Cl8S8,34568
 ragaai_catalyst/tracers/llamaindex_callback.py,sha256=ZY0BJrrlz-P9Mg2dX-ZkVKG3gSvzwqBtk7JL_05MiYA,14028
 ragaai_catalyst/tracers/llamaindex_instrumentation.py,sha256=Ys_jLkvVqo12bKgXDmkp4TxJu9HkBATrFE8cIcTYxWw,14329
-ragaai_catalyst/tracers/tracer.py,sha256=16fu4zAPP8N8xzBtcbvXlWaR8g2GIC8BhOJwcDrp4vo,36993
+ragaai_catalyst/tracers/tracer.py,sha256=LHRd35rTFgYovxsI4tyXHgsCCdccO1_O1txFLFyLUKc,39069
 ragaai_catalyst/tracers/upload_traces.py,sha256=PEE_JhAmOAMKyb-pl4ZoFWhIePxJm1zs93crrk94iEg,5887
 ragaai_catalyst/tracers/agentic_tracing/README.md,sha256=X4QwLb7-Jg7GQMIXj-SerZIgDETfw-7VgYlczOR8ZeQ,4508
 ragaai_catalyst/tracers/agentic_tracing/__init__.py,sha256=yf6SKvOPSpH-9LiKaoLKXwqj5sez8F_5wkOb91yp0oE,260
@@ -74,10 +74,10 @@ ragaai_catalyst/tracers/agentic_tracing/utils/trace_utils.py,sha256=W7Nw-IpugejI
 ragaai_catalyst/tracers/agentic_tracing/utils/unique_decorator.py,sha256=G027toV-Km20JjKrc-Y_PilQ8ABEKrBvvzgLTnqVg7I,5819
 ragaai_catalyst/tracers/agentic_tracing/utils/zip_list_of_unique_files.py,sha256=4TeCGsFF26249fV6dJHLTZDrRa93SG9oer4rudoF8Y4,19443
 ragaai_catalyst/tracers/exporters/__init__.py,sha256=wQbaqyeIjVZxYprHCKZ9BeiqxeXYBKjzEgP79LWNxCU,293
-ragaai_catalyst/tracers/exporters/dynamic_trace_exporter.py,sha256=l5wYX80lmMa39iOgZ7wopcH5km8t-koERL7B9wFy8Hg,5433
+ragaai_catalyst/tracers/exporters/dynamic_trace_exporter.py,sha256=v2V-y1l9t9gQQbqoi2jlEhm1xRQjR5yTyr6rxSriXCE,6303
 ragaai_catalyst/tracers/exporters/file_span_exporter.py,sha256=RgGteu-NVGprXKkynvyIO5yOjpbtA41R3W_NzCjnkwE,6445
 ragaai_catalyst/tracers/exporters/raga_exporter.py,sha256=6xvjWXyh8XPkHKSLLmAZUQSvwuyY17ov8pv2VdfI0qA,17875
-ragaai_catalyst/tracers/exporters/ragaai_trace_exporter.py,sha256=5TJfwD5LBLD3D9i_mrvkx2-Jp8JtWUgKTqnmAo2toCw,11364
+ragaai_catalyst/tracers/exporters/ragaai_trace_exporter.py,sha256=BzL5u7yOcWHHtPueoJclXcvexBQo1SFu5lHY8ANnzP8,12998
 ragaai_catalyst/tracers/instrumentators/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 ragaai_catalyst/tracers/utils/__init__.py,sha256=KeMaZtYaTojilpLv65qH08QmpYclfpacDA0U3wg6Ybw,64
 ragaai_catalyst/tracers/utils/convert_langchain_callbacks_output.py,sha256=e0URuRWCdzpxuBLfL82FOTMjbRuDAkW8aIRi7s7Nocc,1655
@@ -85,11 +85,11 @@ ragaai_catalyst/tracers/utils/convert_llama_instru_callback.py,sha256=8qLo7x4Zsn
 ragaai_catalyst/tracers/utils/extraction_logic_llama_index.py,sha256=ZhPs0YhVtB82-Pq9o1BvCinKE_WPvVxPTEcZjlJbFYM,2371
 ragaai_catalyst/tracers/utils/langchain_tracer_extraction_logic.py,sha256=XS2_x2qneqEx9oAighLg-LRiueWcESLwIC2r7eJT-Ww,3117
 ragaai_catalyst/tracers/utils/model_prices_and_context_window_backup.json,sha256=C3uwkibJ08C9sOX-54kulZYmJlIpZ-SQpfE6HNGrjbM,343502
-ragaai_catalyst/tracers/utils/rag_trace_json_converter.py,sha256=XSanJ3ChRn_Nqj5mlJhy-jFUCAhkw6lBuhhzUx6Mo9k,11239
+ragaai_catalyst/tracers/utils/rag_trace_json_converter.py,sha256=sIUWwzXlEGU8wkVVL0w944tXxsOVdUHHV9UCImvBVOU,11406
 ragaai_catalyst/tracers/utils/trace_json_converter.py,sha256=E0_QfciQMMpCtQYrNB4l8HJhlaFalr5bkMqkVRgQahY,14073
 ragaai_catalyst/tracers/utils/utils.py,sha256=ViygfJ7vZ7U0CTSA1lbxVloHp4NSlmfDzBRNCJuMhis,2374
-ragaai_catalyst-2.1.7b0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-ragaai_catalyst-2.1.7b0.dist-info/METADATA,sha256=Ccfgxo8Ehm7CeGQRZrFsDYPb_RsTymn6B0oG_y9JuZE,22139
-ragaai_catalyst-2.1.7b0.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-ragaai_catalyst-2.1.7b0.dist-info/top_level.txt,sha256=HpgsdRgEJMk8nqrU6qdCYk3di7MJkDL0B19lkc7dLfM,16
-ragaai_catalyst-2.1.7b0.dist-info/RECORD,,
+ragaai_catalyst-2.1.7.1.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+ragaai_catalyst-2.1.7.1.dist-info/METADATA,sha256=oQG1fW2asI1Jxz-Oglbns-GknOOBxpMsQK8SNieowsk,22149
+ragaai_catalyst-2.1.7.1.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+ragaai_catalyst-2.1.7.1.dist-info/top_level.txt,sha256=HpgsdRgEJMk8nqrU6qdCYk3di7MJkDL0B19lkc7dLfM,16
+ragaai_catalyst-2.1.7.1.dist-info/RECORD,,