Remove all "labels"-related code

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"
-)
-
 PARALLEL_UPLOAD_PROCESSING_BY_REPO = Feature("parallel_upload_processing")
 
 CARRYFORWARD_BASE_SEARCH_RANGE_BY_OWNER = Feature("carryforward_base_search_range")

services/report/__init__.py

@@ -50,10 +50,7 @@
     RAW_UPLOAD_RAW_REPORT_COUNT,
     RAW_UPLOAD_SIZE,
 )
-from services.report.raw_upload_processor import (
-    SessionAdjustmentResult,
-    process_raw_upload,
-)
+from services.report.raw_upload_processor import process_raw_upload
 from services.repository import get_repo_provider_service
 from services.yaml.reader import get_paths_from_flags, read_yaml_field
 
@@ -73,7 +70,7 @@ class ProcessingResult:
     session: Session
     report: Report | None = None
     error: ProcessingError | None = None
-    session_adjustment: SessionAdjustmentResult | None = None
+    deleted_sessions: set[int] | None = None
 
 
 @dataclass
@@ -717,7 +714,7 @@ def build_report_from_raw_content(
                 upload=upload,
             )
             result.report = process_result.report
-            result.session_adjustment = process_result.session_adjustment
+            result.deleted_sessions = process_result.deleted_sessions
 
             log.info(
                 "Successfully processed report",
@@ -812,11 +809,10 @@ def update_upload_with_processing_result(
             # delete all the carryforwarded `Upload` records corresponding to `Session`s
             # which have been removed from the report.
             # we always have a `session_adjustment` in the non-error case.
-            assert processing_result.session_adjustment
-            deleted_sessions = (
-                processing_result.session_adjustment.fully_deleted_sessions
+            assert processing_result.deleted_sessions is not None
+            delete_uploads_by_sessionid(
+                upload, list(processing_result.deleted_sessions)
             )
-            delete_uploads_by_sessionid(upload, deleted_sessions)
 
         else:
             error = processing_result.error

services/report/languages/pycoverage.py

@@ -1,7 +1,7 @@
 import sentry_sdk
 
 from services.report.languages.base import BaseLanguageProcessor
-from services.report.report_builder import ReportBuilderSession, SpecialLabelsEnum
+from services.report.report_builder import ReportBuilderSession
 
 COVERAGE_HIT = 1
 COVERAGE_MISS = 0
@@ -16,8 +16,6 @@ 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)
-
         for filename, file_coverage in content["files"].items():
             _file = report_builder_session.create_coverage_file(filename)
             if _file is None:
@@ -28,79 +26,6 @@ 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,
-                        ),
-                    )
+                    _line = report_builder_session.create_coverage_line(cov)
+                    _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)

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)