removing decorators and using auto_instrument_package Function wrapping

Author: mashraf-222

examples/run_all_traces.py

@@ -29,13 +29,9 @@
 import pandas as pd
 
 from src.telemetry import setup_telemetry
-from src.numerical.optimization import gradient_descent
-from src.algorithms.graph import graph_traversal, find_node_clusters, PathFinder, calculate_node_betweenness
-from src.algorithms.dynamic_programming import fibonacci, matrix_sum, matrix_chain_order, coin_change, knapsack
-from src.data_processing.dataframe import dataframe_filter, groupby_mean, dataframe_merge
-from src.statistics.descriptive import describe, correlation
+from src.telemetry.auto_instrumentation import auto_instrument_package
 
-# Initialize OpenTelemetry with auto-instrumentation
+# Initialize OpenTelemetry with auto-instrumentation FIRST
 # Uses environment variables if set, otherwise defaults to console exporter
 print("=" * 80)
 print("Initializing OpenTelemetry with auto-instrumentation...")
@@ -51,13 +47,58 @@
 exporter_type = os.getenv("OTEL_TRACES_EXPORTER") or os.getenv("OTEL_EXPORTER_TYPE", "console")
 exporter_endpoint = os.getenv("OTEL_EXPORTER_OTLP_ENDPOINT", "http://localhost:4317")
 
-setup_telemetry(
-    service_name="optimize-me",
-    service_version="0.1.0",
-    exporter_type=exporter_type,
-    exporter_endpoint=exporter_endpoint,
-    use_auto_instrumentation=True,  # Use auto-instrumentation (standard pattern)
+# Detect if running under opentelemetry-instrument
+# opentelemetry-instrument sets up its own TracerProvider, so we shouldn't replace it
+from opentelemetry import trace
+existing_provider = trace.get_tracer_provider()
+is_opentelemetry_instrument = (
+    existing_provider is not None 
+    and not isinstance(existing_provider, trace.NoOpTracerProvider)
+    and type(existing_provider).__name__ in ("ProxyTracerProvider", "TracerProvider")
+)
+
+if is_opentelemetry_instrument:
+    print("Detected opentelemetry-instrument - using existing OpenTelemetry setup")
+    print(f"Exporter: {exporter_type}, Endpoint: {exporter_endpoint}")
+    # Still call setup_telemetry to ensure console exporter is added if needed
+    # setup_telemetry handles ProxyTracerProvider by creating a new TracerProvider
+    # This ensures we have a console exporter configured for immediate output
+    setup_telemetry(
+        service_name="optimize-me",
+        service_version="0.1.0",
+        exporter_type=exporter_type,  # Use console when OTEL_TRACES_EXPORTER=console
+        exporter_endpoint=exporter_endpoint,
+        use_auto_instrumentation=True,
+    )
+    if exporter_type == "console":
+        print("✅ Console exporter configured - traces will appear below")
+else:
+    # Call setup_telemetry to create TracerProvider and configure exporter
+    setup_telemetry(
+        service_name="optimize-me",
+        service_version="0.1.0",
+        exporter_type=exporter_type,
+        exporter_endpoint=exporter_endpoint,
+        use_auto_instrumentation=True,  # Use auto-instrumentation (standard pattern)
+    )
+
+# Auto-instrument all custom functions in src package
+# This automatically traces all functions without requiring decorators
+# IMPORTANT: This must happen BEFORE importing the modules
+print("\nSetting up auto-instrumentation for custom functions...")
+auto_instrument_package(
+    'src',
+    include_private=False,  # Don't instrument private functions (starting with _)
+    exclude_modules=['src.tests', 'src.telemetry']  # Exclude test and telemetry modules
 )
+print("✅ Auto-instrumentation enabled - all functions will be traced automatically")
+
+# NOW import modules - functions will be automatically wrapped
+from src.numerical.optimization import gradient_descent
+from src.algorithms.graph import graph_traversal, find_node_clusters, PathFinder, calculate_node_betweenness
+from src.algorithms.dynamic_programming import fibonacci, matrix_sum, matrix_chain_order, coin_change, knapsack
+from src.data_processing.dataframe import dataframe_filter, groupby_mean, dataframe_merge
+from src.statistics.descriptive import describe, correlation
 
 print("\n" + "=" * 80)
 print("RUNNING ALL INSTRUMENTED FUNCTIONS")
@@ -202,11 +243,15 @@
 # Force flush spans to ensure they're exported (especially for console exporter)
 try:
     from opentelemetry import trace
+    from opentelemetry.sdk.trace import TracerProvider
     provider = trace.get_tracer_provider()
-    if hasattr(provider, "force_flush"):
-        provider.force_flush()
-except Exception:
-    pass
+    if isinstance(provider, TracerProvider):
+        if hasattr(provider, "force_flush"):
+            provider.force_flush(timeout_millis=5000)  # Wait up to 5 seconds
+            print("\n✅ Spans flushed to exporter")
+except Exception as e:
+    if os.getenv("OTEL_LOG_LEVEL", "").upper() == "DEBUG":
+        print(f"\n⚠️ Could not flush spans: {e}")
 
 print()
 

src/algorithms/dynamic_programming.py

@@ -1,21 +1,16 @@
 from __future__ import annotations
 
-from src.telemetry.decorators import trace_function
 
-
-@trace_function
 def fibonacci(n):
     if n <= 1:
         return n
     return fibonacci(n - 1) + fibonacci(n - 2)
 
 
-@trace_function
 def matrix_sum(matrix: list[list[int]]) -> list[int]:
     return [sum(matrix[i]) for i in range(len(matrix)) if sum(matrix[i]) > 0]
 
 
-@trace_function
 def matrix_chain_order(matrices: list[tuple[int, int]]) -> int:
     """
     Find the minimum number of operations needed to multiply a chain of matrices.
@@ -47,7 +42,6 @@ def dp(i: int, j: int) -> int:
     return dp(0, n - 1)
 
 
-@trace_function
 def coin_change(coins: list[int], amount: int, index: int) -> int:
     if amount == 0:
         return 1
@@ -59,7 +53,6 @@ def coin_change(coins: list[int], amount: int, index: int) -> int:
     )
 
 
-@trace_function
 def knapsack(weights: list[int], values: list[int], capacity: int, n: int) -> int:
     if n == 0 or capacity == 0:
         return 0

src/algorithms/graph.py

@@ -2,10 +2,7 @@
 
 import networkx as nx
 
-from src.telemetry.decorators import trace_function
 
-
-@trace_function
 def graph_traversal(graph: dict[int, dict[int]], node: int) -> dict[int]:
     visited = []
 
@@ -24,7 +21,6 @@ class PathFinder:
     def __init__(self, graph: dict[str, list[str]]):
         self.graph = graph
 
-    @trace_function
     def find_shortest_path(self, start: str, end: str) -> list[str]:
         if start not in self.graph or end not in self.graph:
             return []
@@ -106,7 +102,6 @@ def find_node_with_highest_degree(
     return max_degree_node
 
 
-@trace_function
 def find_node_clusters(nodes: list[dict], edges: list[dict]) -> list[list[dict]]:
     """Find connected components (clusters) in the graph."""
     # Create node ID to node mapping for easy lookup
@@ -154,7 +149,6 @@ def find_node_clusters(nodes: list[dict], edges: list[dict]) -> list[list[dict]]
     return clusters
 
 
-@trace_function
 def calculate_node_betweenness(
     nodes: list[str], edges: list[dict[str, str]]
 ) -> dict[str, float]:

src/data_processing/dataframe.py

@@ -3,10 +3,7 @@
 import numpy as np
 import pandas as pd
 
-from src.telemetry.decorators import trace_function
 
-
-@trace_function
 def dataframe_filter(df: pd.DataFrame, column: str, value: Any) -> pd.DataFrame:
     indices = []
     for i in range(len(df)):
@@ -15,7 +12,6 @@ def dataframe_filter(df: pd.DataFrame, column: str, value: Any) -> pd.DataFrame:
     return df.iloc[indices].reset_index(drop=True)
 
 
-@trace_function
 def groupby_mean(df: pd.DataFrame, group_col: str, value_col: str) -> dict[Any, float]:
     sums = {}
     counts = {}
@@ -34,7 +30,6 @@ def groupby_mean(df: pd.DataFrame, group_col: str, value_col: str) -> dict[Any,
     return result
 
 
-@trace_function
 def dataframe_merge(
     left: pd.DataFrame, right: pd.DataFrame, left_on: str, right_on: str
 ) -> pd.DataFrame:

src/numerical/optimization.py

@@ -1,9 +1,6 @@
 import numpy as np
 
-from src.telemetry.decorators import trace_function
 
-
-@trace_function(span_name="gradient_descent", capture_args=["iterations", "learning_rate"])
 def gradient_descent(
     X: np.ndarray, y: np.ndarray, learning_rate: float = 0.01, iterations: int = 1000
 ) -> np.ndarray:

src/statistics/descriptive.py

@@ -3,10 +3,7 @@
 import numpy as np
 import pandas as pd
 
-from src.telemetry.decorators import trace_function
 
-
-@trace_function
 def describe(series: pd.Series) -> dict[str, float]:
     values = [v for v in series if not pd.isna(v)]
     n = len(values)
@@ -44,7 +41,6 @@ def percentile(p):
     }
 
 
-@trace_function
 def correlation(df: pd.DataFrame) -> dict[Tuple[str, str], float]:
     numeric_columns = [
         col for col in df.columns if np.issubdtype(df[col].dtype, np.number)