fix(seer-rpc): Don't extract function name, simplify ifs (#89938)

Have been debugging why the issue fetching query sometimes breaks, and
it seems like it's related to the more complicated parsers. Wanna see
whether a simpler query works.

Seer never uses the `function_name`, so it can be safely removed

Author: kddubey

src/sentry/seer/fetch_issues/fetch_issues.py

@@ -21,7 +21,10 @@
 from sentry.models.group import Group, GroupStatus
 from sentry.models.project import Project
 from sentry.models.repository import Repository
-from sentry.seer.fetch_issues.more_parsing import patch_parsers_more
+from sentry.seer.fetch_issues.more_parsing import (
+    patch_parsers_more,
+    simple_function_name_conditions,
+)
 from sentry.snuba.dataset import Dataset
 from sentry.snuba.referrer import Referrer
 from sentry.utils.snuba import raw_snql_query
@@ -90,13 +93,6 @@ def _get_issues_for_file(
     if not projects:
         return []
 
-    # Gets the appropriate parser for formatting the snuba query given the file extension.
-    # The extension is never replaced in reverse codemapping.
-    file_extension = sentry_filenames[0].split(".")[-1]
-    if file_extension not in patch_parsers_more:
-        return []
-    language_parser = patch_parsers_more[file_extension]
-
     # Fetch an initial, candidate set of groups.
     group_ids: list[int] = list(
         Group.objects.filter(
@@ -109,7 +105,6 @@ def _get_issues_for_file(
         .values_list("id", flat=True)
     )[:MAX_RECENT_ISSUES]
     project_ids = [project.id for project in projects]
-    multi_if = language_parser.generate_multi_if(function_names)
 
     # Fetch the latest event for each group, along with some other event data we'll need for
     # filtering by function names and file names.
@@ -163,7 +158,6 @@ def _get_issues_for_file(
                 Column("group_id"),
                 Column("event_id"),
                 Column("title"),
-                Function("multiIf", multi_if, "function_name"),
             ]
         )
         .set_where(
@@ -182,9 +176,7 @@ def _get_issues_for_file(
                                     Op.IN,
                                     sentry_filenames,
                                 ),
-                                language_parser.generate_function_name_conditions(
-                                    function_names, stackframe_idx
-                                ),
+                                simple_function_name_conditions(function_names, stackframe_idx),
                             ],
                         )
                         for stackframe_idx in range(-STACKFRAME_COUNT, 0)  # first n frames
@@ -235,15 +227,11 @@ def _add_event_details(
     serialized_events: list[EventSerializerResponse] = serialize(
         events, serializer=EventSerializer()
     )
-    group_id_to_group_dict = {
-        group_dict["group_id"]: group_dict for group_dict in issues_result_set
-    }
     return [
         {  # Structured like seer.automation.models.IssueDetails
             "id": int(event_dict["groupID"]),
             "title": event_dict["title"],
             "events": [event_dict],
-            "function_name": group_id_to_group_dict[int(event_dict["groupID"])]["function_name"],
         }
         for event_dict in serialized_events
         if event_dict["groupID"] is not None

src/sentry/seer/fetch_issues/more_parsing.py

@@ -1,13 +1,23 @@
 import re
 
+from snuba_sdk import Condition, Op
+
 from sentry.integrations.source_code_management.language_parsers import (
     PATCH_PARSERS,
     LanguageParser,
     PythonParser,
     SimpleLanguageParser,
+    stackframe_function_name,
 )
 
 
+def simple_function_name_conditions(
+    function_names: list[str],
+    stack_frame_idx: int,
+):
+    return Condition(stackframe_function_name(stack_frame_idx), Op.IN, function_names)
+
+
 class PythonParserMore(PythonParser):
     @classmethod
     def extract_functions_from_rest_of_patch(cls, patch: str) -> set[str]:

tests/sentry/seer/fetch_issues/test_fetch_issues.py

@@ -37,10 +37,8 @@ def test_simple(self):
         )
 
         issue_ids = [issue["id"] for issue in issues_with_event_details]
-        function_names = [issue["function_name"] for issue in issues_with_event_details]
         assert group_id != self.group_id
         assert set(issue_ids) == {group_id, self.group_id}
-        assert set(function_names) == {"planet", "world"}
 
         # Check that each issue is structured like an IssueDetails
         assert all(len(issue["events"]) == 1 for issue in issues_with_event_details)
@@ -57,6 +55,26 @@ def test_simple(self):
             serialized_event = serialize(event, serializer=EventSerializer())
             assert event_dict == serialized_event
 
+    def test_javascript_simple(self):
+        group_id = [
+            self._create_event(
+                function_names=["component.blue", "world"],
+                filenames=["foo.js", "baz.js"],
+                user_id=str(i),
+            )
+            for i in range(6)
+        ][0].group.id
+
+        issues_with_event_details = get_issues_with_event_details_for_file(
+            projects=[self.project],
+            sentry_filenames=["baz.js"],
+            function_names=["world", "planet"],
+        )
+
+        issue_ids = [issue["id"] for issue in issues_with_event_details]
+        assert group_id != self.group_id
+        assert set(issue_ids) == {group_id}
+
     # The rest are mostly copied from tests/sentry/integrations/github/tasks/test_open_pr_comment.py
     def test_filename_mismatch(self):
         group_id = self._create_event(
@@ -85,10 +103,8 @@ def test_not_first_frame(self):
 
         issues = get_issues_with_event_details_for_file([self.project], ["baz.py"], ["world"])
         issue_ids = [issue["id"] for issue in issues]
-        function_names = [issue["function_name"] for issue in issues]
         assert group_id != self.group_id
         assert set(issue_ids) == {self.group_id, group_id}
-        assert function_names == ["world", "world"]
 
     def test_not_within_frame_limit(self):
         function_names = ["world"] + ["a" for _ in range(STACKFRAME_COUNT)]