Strip leading comments from SQL queries when generating the span name.

Author: Daniel Rogers

CHANGELOG.md

@@ -13,6 +13,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   ([#1424](https://github.com/open-telemetry/opentelemetry-python-contrib/pull/1424))
 - `opentelemetry-instrumentation-fastapi` Add support for regular expression matching and sanitization of HTTP headers.
   ([#1403](https://github.com/open-telemetry/opentelemetry-python-contrib/pull/1403))
+- `opentelemetry-instrumentation-dbapi` Strip leading comments from SQL queries when generating the span name.
+  ([#1434](https://github.com/open-telemetry/opentelemetry-python-contrib/pull/1434))
 
 ### Fixed
 

instrumentation/opentelemetry-instrumentation-dbapi/src/opentelemetry/instrumentation/dbapi/__init__.py

@@ -39,6 +39,7 @@
 
 import functools
 import logging
+import re
 import typing
 
 import wrapt
@@ -368,6 +369,7 @@ def __init__(self, db_api_integration: DatabaseApiIntegration) -> None:
             else {}
         )
         self._connect_module = self._db_api_integration.connect_module
+        self._leading_comment_remover = re.compile(r"^/\*.*?\*/")
 
     def _populate_span(
         self,
@@ -397,7 +399,8 @@ def _populate_span(
 
     def get_operation_name(self, cursor, args):  # pylint: disable=no-self-use
         if args and isinstance(args[0], str):
-            return args[0].split()[0]
+            # Strip leading comments so we get the operation name.
+            return self._leading_comment_remover.sub("", args[0]).split()[0]
         return ""
 
     def get_statement(self, cursor, args):  # pylint: disable=no-self-use

instrumentation/opentelemetry-instrumentation-dbapi/tests/test_dbapi_integration.py

@@ -88,11 +88,17 @@ def test_span_name(self):
         query"""
         )
         cursor.execute("tab\tseparated query")
+        cursor.execute("/* leading comment */ query")
+        cursor.execute("/* leading comment */ query /* trailing comment */")
+        cursor.execute("query /* trailing comment */")
         spans_list = self.memory_exporter.get_finished_spans()
-        self.assertEqual(len(spans_list), 3)
+        self.assertEqual(len(spans_list), 6)
         self.assertEqual(spans_list[0].name, "Test")
         self.assertEqual(spans_list[1].name, "multi")
         self.assertEqual(spans_list[2].name, "tab")
+        self.assertEqual(spans_list[3].name, "query")
+        self.assertEqual(spans_list[4].name, "query")
+        self.assertEqual(spans_list[5].name, "query")
 
     def test_span_succeeded_with_capture_of_statement_parameters(self):
         connection_props = {

instrumentation/opentelemetry-instrumentation-psycopg2/src/opentelemetry/instrumentation/psycopg2/__init__.py

@@ -216,7 +216,8 @@ def get_operation_name(self, cursor, args):
             statement = statement.as_string(cursor)
 
         if isinstance(statement, str):
-            return statement.split()[0]
+            # Strip leading comments so we get the operation name.
+            return self._leading_comment_remover.sub("", statement).split()[0]
 
         return ""
 

instrumentation/opentelemetry-instrumentation-psycopg2/tests/test_psycopg2_integration.py

@@ -116,6 +116,32 @@ def test_instrumentor(self):
         spans_list = self.memory_exporter.get_finished_spans()
         self.assertEqual(len(spans_list), 1)
 
+    def test_span_name(self):
+        Psycopg2Instrumentor().instrument()
+
+        cnx = psycopg2.connect(database="test")
+
+        cursor = cnx.cursor()
+
+        cursor.execute("Test query", ("param1Value", False))
+        cursor.execute(
+            """multi
+        line
+        query"""
+        )
+        cursor.execute("tab\tseparated query")
+        cursor.execute("/* leading comment */ query")
+        cursor.execute("/* leading comment */ query /* trailing comment */")
+        cursor.execute("query /* trailing comment */")
+        spans_list = self.memory_exporter.get_finished_spans()
+        self.assertEqual(len(spans_list), 6)
+        self.assertEqual(spans_list[0].name, "Test")
+        self.assertEqual(spans_list[1].name, "multi")
+        self.assertEqual(spans_list[2].name, "tab")
+        self.assertEqual(spans_list[3].name, "query")
+        self.assertEqual(spans_list[4].name, "query")
+        self.assertEqual(spans_list[5].name, "query")
+
     # pylint: disable=unused-argument
     def test_not_recording(self):
         mock_tracer = mock.Mock()