feat(tracing): better otel adapter

pyproject.toml

@@ -9,7 +9,7 @@ dependencies = [
     "llama-index-embeddings-azure-openai>=0.3.8",
     "llama-index-llms-azure-openai>=0.3.2",
     "openinference-instrumentation-llama-index>=4.3.0",
-    "uipath>=2.1.10, <2.2.0",
+    "uipath>=2.1.28.dev1005170000,<2.1.28.dev1005180000",
 ]
 classifiers = [
     "Development Status :: 3 - Alpha",
@@ -91,3 +91,5 @@ url = "https://test.pypi.org/simple/"
 publish-url = "https://test.pypi.org/legacy/"
 explicit = true
 
+[tool.uv.sources]
+uipath = { index = "testpypi" }

src/uipath_llamaindex/_cli/_tracing/_oteladapter.py

@@ -1,101 +1,26 @@
-import json
 import logging
 from typing import Any, Dict
 
 from opentelemetry.sdk.trace.export import (
     SpanExportResult,
 )
-from uipath.tracing import LlmOpsHttpExporter
+from uipath.tracing import CommonSpanProcessor, LlmOpsHttpExporter
 
 logger = logging.getLogger(__name__)
 
 
 class LlamaIndexExporter(LlmOpsHttpExporter):
-    # Mapping of old attribute names to new attribute names or (new name, function)
-    ATTRIBUTE_MAPPING = {
-        "input.value": ("input", lambda s: json.loads(s)),
-        "output.value": ("output", lambda s: json.loads(s)),
-        "llm.model_name": "model",
-    }
-
-    # Mapping of span types
-    SPAN_TYPE_MAPPING = {
-        "LLM": "completion",
-        # Add more mappings as needed
-    }
-
     def __init__(self, **kwargs):
         super().__init__(**kwargs)
+        self._processor = CommonSpanProcessor()
 
     def _send_with_retries(
         self, url: str, payload: list[Dict[str, Any]], max_retries: int = 4
     ) -> SpanExportResult:
-        # Transform attributes in each span's payload before sending
-        for span_data in payload:
-            if "Attributes" in span_data and isinstance(span_data["Attributes"], str):
-                try:
-                    # Parse the JSON string to a dictionary
-                    attributes = json.loads(span_data["Attributes"])
-
-                    if "openinference.span.kind" in attributes:
-                        # Remove the span kind attribute
-                        span_type = attributes["openinference.span.kind"]
-                        # Map span type using SPAN_TYPE_MAPPING
-                        span_data["SpanType"] = self.SPAN_TYPE_MAPPING.get(
-                            span_type, span_type
-                        )
-                        del attributes["openinference.span.kind"]
-
-                    # Apply the transformation logic
-                    for old_key, mapping in self.ATTRIBUTE_MAPPING.items():
-                        if old_key in attributes:
-                            if isinstance(mapping, tuple):
-                                new_key, func = mapping
-                                try:
-                                    attributes[new_key] = func(attributes[old_key])
-                                except Exception:
-                                    attributes[new_key] = attributes[old_key]
-                            else:
-                                new_key = mapping
-                                attributes[new_key] = attributes[old_key]
-                            del attributes[old_key]
-
-                    # Transform token usage data if present
-                    token_keys = {
-                        "llm.token_count.prompt": "promptTokens",
-                        "llm.token_count.completion": "completionTokens",
-                        "llm.token_count.total": "totalTokens",
-                    }
-
-                    # Check if any token count keys exist
-                    if any(key in attributes for key in token_keys):
-                        usage = {}
-                        for old_key, new_key in token_keys.items():
-                            if old_key in attributes:
-                                usage[new_key] = attributes[old_key]
-                                del attributes[old_key]
-
-                        # Add default values for BYO execution fields
-                        usage["isByoExecution"] = False
-                        # usage["executionDeploymentType"] = "PAYGO"
-
-                        # Add usage to attributes
-                        attributes["usage"] = usage
-
-                        # Clean up any other token count attributes
-                        keys_to_remove = [
-                            k for k in attributes if k.startswith("llm.token_count.")
-                        ]
-                        for key in keys_to_remove:
-                            del attributes[key]
-
-                    # Convert back to JSON string
-                    span_data["attributes"] = json.dumps(attributes)
-                except json.JSONDecodeError as e:
-                    logger.warning(f"Failed to parse attributes JSON: {e}")
-
+        processed_payload = [self._processor.process_span(span) for span in payload]
         return super()._send_with_retries(
             url=url,
-            payload=payload,
+            payload=processed_payload,
             max_retries=max_retries,
         )
+