fix: weaviate create collection tracing (#298)

Author: darshit-s3

pyproject.toml

@@ -18,7 +18,7 @@ classifiers=[
     "Operating System :: OS Independent",
 ]
 dependencies = [
-  'trace-attributes==7.0.1',
+  'trace-attributes==7.0.3',
   'opentelemetry-api>=1.25.0',
   'opentelemetry-sdk>=1.25.0',
   'opentelemetry-instrumentation>=0.47b0',

src/examples/weaviate_example/query_text.py

@@ -15,13 +15,14 @@
 import requests
 import weaviate
 import weaviate.classes as wvc
+import weaviate.classes.config as wc
+from dotenv import load_dotenv
 from weaviate.classes.aggregate import GroupByAggregate
 from weaviate.classes.query import Filter, HybridFusion, MetadataQuery
 from weaviate.collections.classes.grpc import Move
 
 import langtrace_python_sdk.langtrace as langtrace
 from langtrace_python_sdk import with_langtrace_root_span
-from dotenv import load_dotenv
 
 load_dotenv()
 # Set these environment variables
@@ -44,11 +45,16 @@ def create():
     if not client.collections.get("Question"):
         questions = client.collections.create(
             name="Question",
-            # Set the vectorizer to "text2vec-openai" to use the OpenAI API for vector-related operations
-            vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(),
-            # Ensure the `generative-openai` module is used for generative queries
-            generative_config=wvc.config.Configure.Generative.openai(),
-        )
+        properties=[
+            wc.Property(name="answer", data_type=wc.DataType.TEXT),
+            wc.Property(name="question", data_type=wc.DataType.TEXT),
+            wc.Property(name="category", data_type=wc.DataType.TEXT),
+        ],
+        # Define the vectorizer module
+        vectorizer_config=wc.Configure.Vectorizer.text2vec_openai(),
+        # Define the generative module
+        generative_config=wc.Configure.Generative.openai(),
+    )
 
 
 @with_langtrace_root_span("insert")

src/langtrace_python_sdk/instrumentation/weaviate/instrumentation.py

@@ -37,7 +37,7 @@ class WeaviateInstrumentation(BaseInstrumentor):
     """
 
     def instrumentation_dependencies(self) -> Collection[str]:
-        return ["weaviate-client >= 4.6.1", "trace-attributes >= 4.0.2"]
+        return ["weaviate-client >= 4.6.1", "trace-attributes >= 7.0.3"]
 
     def _instrument(self, **kwargs):
         tracer_provider = kwargs.get("tracer_provider")

src/langtrace_python_sdk/instrumentation/weaviate/patch.py

@@ -16,21 +16,21 @@
 
 import json
 
+from importlib_metadata import version as v
 from langtrace.trace_attributes import DatabaseSpanAttributes
-from langtrace_python_sdk.utils.llm import get_span_name
 from opentelemetry import baggage, trace
 from opentelemetry.trace import SpanKind
-from opentelemetry.trace.status import Status, StatusCode
 from opentelemetry.trace.propagation import set_span_in_context
+from opentelemetry.trace.status import Status, StatusCode
+
+from langtrace_python_sdk.constants import LANGTRACE_SDK_NAME
 from langtrace_python_sdk.constants.instrumentation.common import (
     LANGTRACE_ADDITIONAL_SPAN_ATTRIBUTES_KEY,
     SERVICE_PROVIDERS,
 )
 from langtrace_python_sdk.constants.instrumentation.weaviate import APIS
+from langtrace_python_sdk.utils.llm import get_span_name
 from langtrace_python_sdk.utils.misc import extract_input_params, to_iso_format
-from importlib_metadata import version as v
-
-from langtrace_python_sdk.constants import LANGTRACE_SDK_NAME
 
 # Predefined metadata response attributes
 METADATA_ATTRIBUTES = [
@@ -45,6 +45,22 @@
 ]
 
 
+def extract_inputs(args, kwargs):
+    extracted_params = {}
+    kwargs_without_properties = {k: v for k, v in kwargs.items() if k != "properties"}
+    extracted_params.update(extract_input_params(args, kwargs_without_properties))
+
+    if kwargs.get("properties", None):
+        extracted_params["properties"] = []
+        for each_prop in kwargs.get("properties"):
+            if hasattr(each_prop, "_to_dict"):
+                # append properties to extracted_params
+                extracted_params["properties"].append(each_prop._to_dict())
+
+        extracted_params["properties"] = json.dumps(extracted_params["properties"])
+    return extracted_params
+
+
 def extract_metadata(metadata):
     # Extraction response Query metadata
     extracted_metadata = {
@@ -126,7 +142,7 @@ def traced_method(wrapped, instance, args, kwargs):
             "db.system": "weaviate",
             "db.operation": api["OPERATION"],
             "db.collection.name": collection_name,
-            "db.query": json.dumps(extract_input_params(args, kwargs)),
+            "db.query": json.dumps(extract_inputs(args, kwargs)),
             **(extra_attributes if extra_attributes is not None else {}),
         }
 
@@ -143,7 +159,6 @@ def traced_method(wrapped, instance, args, kwargs):
             try:
                 # Attempt to call the original method
                 result = wrapped(*args, **kwargs)
-                print(result)
                 if api["OPERATION"] in ["query", "generate"]:
                     span.add_event(
                         name="db.response",

src/langtrace_python_sdk/version.py

@@ -1 +1 @@
-__version__ = "2.2.24"
+__version__ = "2.2.25"