Remove labels-index feature flag (#852)

rollouts/__init__.py

@@ -4,12 +4,6 @@
 FLAKY_TEST_DETECTION = Feature("flaky_test_detection")
 FLAKY_SHADOW_MODE = Feature("flaky_shadow_mode")
 
-# Eventually we want all repos to use this
-# This flag will just help us with the rollout process
-USE_LABEL_INDEX_IN_REPORT_PROCESSING_BY_REPO_ID = Feature(
-    "use_label_index_in_report_processing"
-)
-
 INTERMEDIATE_REPORTS_IN_REDIS = Feature("intermediate_reports_in_redis")
 
 CARRYFORWARD_BASE_SEARCH_RANGE_BY_OWNER = Feature("carryforward_base_search_range")

services/report/__init__.py

@@ -665,11 +665,7 @@ def build_report_from_raw_content(
         log.debug("Retrieved report for processing from url %s", archive_url)
         try:
             result.report = process_raw_upload(
-                self.current_yaml,
-                raw_report,
-                flags,
-                session,
-                upload=upload,
+                self.current_yaml, raw_report, flags, session
             )
 
             log.info(

services/report/languages/pycoverage.py

@@ -16,7 +16,7 @@ def matches_content(self, content: dict, first_line: str, name: str) -> bool:
     def process(
         self, content: dict, report_builder_session: ReportBuilderSession
     ) -> None:
-        labels_table = LabelsTable(report_builder_session, content)
+        labels_table = content.get("labels_table", {})
 
         for filename, file_coverage in content["files"].items():
             _file = report_builder_session.create_coverage_file(filename)
@@ -28,79 +28,26 @@ def process(
             ] + [(COVERAGE_MISS, ln) for ln in file_coverage["missing_lines"]]
             for cov, ln in lines_and_coverage:
                 if ln > 0:
-                    label_list_of_lists: list[list[str]] | list[list[int]] = []
-                    if report_builder_session.should_use_label_index:
-                        label_list_of_lists = [
-                            [single_id]
-                            for single_id in labels_table._get_list_of_label_ids(
-                                report_builder_session.label_index,
-                                file_coverage.get("contexts", {}).get(str(ln), []),
-                            )
-                        ]
-                    else:
-                        label_list_of_lists = [
-                            [labels_table._normalize_label(testname)]
-                            for testname in file_coverage.get("contexts", {}).get(
-                                str(ln), []
-                            )
-                        ]
-                    _file.append(
-                        ln,
-                        report_builder_session.create_coverage_line(
-                            cov,
-                            labels_list_of_lists=label_list_of_lists,
-                        ),
+                    label_list_of_lists = [
+                        [_normalize_label(labels_table, testname)]
+                        for testname in file_coverage.get("contexts", {}).get(
+                            str(ln), []
+                        )
+                    ]
+                    _line = report_builder_session.create_coverage_line(
+                        cov,
+                        labels_list_of_lists=label_list_of_lists,
                     )
+                    _file.append(ln, _line)
             report_builder_session.append(_file)
 
 
-class LabelsTable:
-    def __init__(
-        self, report_builder_session: ReportBuilderSession, content: dict
-    ) -> None:
-        self.labels_table: dict[str, str] = {}
-        self.reverse_table: dict[str, int] = {}
-        self.are_labels_already_encoded = False
-
-        # Compressed pycoverage files will include a labels_table
-        if "labels_table" in content:
-            self.labels_table = content["labels_table"]
-            # We can pre-populate some of the indexes that will be used
-            for idx, testname in self.labels_table.items():
-                clean_label = self._normalize_label(testname)
-                report_builder_session.label_index[int(idx)] = clean_label
-            self.are_labels_already_encoded = True
-
-    def _normalize_label(self, testname: int | float | str) -> str:
-        if isinstance(testname, int) or isinstance(testname, float):
-            # This is from a compressed report.
-            # Pull label from the labels_table
-            # But the labels_table keys are strings, because of JSON format
-            testname = self.labels_table[str(testname)]
-        if testname == "":
-            return SpecialLabelsEnum.CODECOV_ALL_LABELS_PLACEHOLDER.corresponding_label
-        return testname.split("|", 1)[0]
-
-    def _get_list_of_label_ids(
-        self,
-        current_label_idx: dict[int, str],
-        line_contexts: list[str | int],
-    ) -> list[int]:
-        if self.are_labels_already_encoded:
-            # The line contexts already include indexes in the table.
-            # We can re-use the table and don't have to do anything with contexts.
-            return sorted(map(int, line_contexts))
-
-        # In this case we do need to fix the labels
-        label_ids_for_line = set()
-        for testname in line_contexts:
-            clean_label = self._normalize_label(testname)
-            if clean_label in self.reverse_table:
-                label_ids_for_line.add(self.reverse_table[clean_label])
-            else:
-                label_id = max([*current_label_idx.keys(), 0]) + 1
-                current_label_idx[label_id] = clean_label
-                self.reverse_table[clean_label] = label_id
-                label_ids_for_line.add(label_id)
-
-        return sorted(label_ids_for_line)
+def _normalize_label(labels_table: dict[str, str], testname: int | float | str) -> str:
+    if isinstance(testname, int) or isinstance(testname, float):
+        # This is from a compressed report.
+        # Pull label from the labels_table
+        # But the labels_table keys are strings, because of JSON format
+        testname = labels_table[str(testname)]
+    if testname == "":
+        return SpecialLabelsEnum.CODECOV_ALL_LABELS_PLACEHOLDER.corresponding_label
+    return testname.split("|", 1)[0]

services/report/languages/tests/unit/__init__.py

@@ -6,7 +6,6 @@ def create_report_builder_session(
     path_fixer: PathFixer | None = None,
     filename: str = "filename",
     current_yaml: dict | None = None,
-    should_use_label_index: bool = False,
 ) -> ReportBuilderSession:
     def fixes(filename, bases_to_try=None):
         return filename
@@ -16,6 +15,5 @@ def fixes(filename, bases_to_try=None):
         ignored_lines={},
         sessionid=0,
         current_yaml=current_yaml,
-        should_use_label_index=should_use_label_index,
     )
     return report_builder.create_report_builder_session(filename)