Extract Comparator class from CompareCommand

Move JSON loading, batching, and cross-referencing logic into a
dedicated Comparator class. CompareCommand.run() now delegates to
the Comparator and iterates over it to build the comment.

Author: Pierre-Sassoulas

pylint/testutils/_primer/comparator.py

@@ -0,0 +1,63 @@
+# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
+# For details: https://github.com/pylint-dev/pylint/blob/main/LICENSE
+# Copyright (c) https://github.com/pylint-dev/pylint/blob/main/CONTRIBUTORS.txt
+
+from __future__ import annotations
+
+import json
+from collections.abc import Generator
+from pathlib import Path
+
+from pylint.reporters.json_reporter import OldJsonExport
+from pylint.testutils._primer.primer_command import (
+    PackageData,
+    PackageMessages,
+)
+
+
+class Comparator:
+    def __init__(
+        self, base_file: str, new_file: str, batches: int | None = None
+    ) -> None:
+        main_data: PackageMessages
+        if batches is None:
+            main_data = self._load_json(base_file)
+            self.pr_data = self._load_json(new_file)
+        else:
+            main_data = {}
+            self.pr_data = {}
+            for idx in range(batches):
+                main_data.update(
+                    self._load_json(base_file.replace("BATCHIDX", "batch" + str(idx)))
+                )
+                self.pr_data.update(
+                    self._load_json(new_file.replace("BATCHIDX", "batch" + str(idx)))
+                )
+
+        self.missing_messages: PackageMessages = {}
+        for package, data in main_data.items():
+            package_missing_messages: list[OldJsonExport] = []
+            for message in data["messages"]:
+                try:
+                    self.pr_data[package]["messages"].remove(message)
+                except ValueError:
+                    package_missing_messages.append(message)
+            self.missing_messages[package] = PackageData(
+                commit=self.pr_data[package]["commit"],
+                messages=package_missing_messages,
+            )
+
+    def __iter__(
+        self,
+    ) -> Generator[tuple[str, PackageData, PackageData]]:
+        for package, missing_messages in self.missing_messages.items():
+            new_messages = self.pr_data[package]
+            if not missing_messages["messages"] and not new_messages["messages"]:
+                continue
+            yield package, missing_messages, new_messages
+
+    @staticmethod
+    def _load_json(file_path: Path | str) -> PackageMessages:
+        with open(file_path, encoding="utf-8") as f:
+            result: PackageMessages = json.load(f)
+        return result

pylint/testutils/_primer/primer_compare_command.py

@@ -3,79 +3,28 @@
 # Copyright (c) https://github.com/pylint-dev/pylint/blob/main/CONTRIBUTORS.txt
 from __future__ import annotations
 
-import json
-from pathlib import Path, PurePosixPath
+from pathlib import PurePosixPath
 
-from pylint.reporters.json_reporter import OldJsonExport
-from pylint.testutils._primer.primer_command import (
-    PackageData,
-    PackageMessages,
-    PrimerCommand,
-)
+from pylint.testutils._primer.comparator import Comparator
+from pylint.testutils._primer.primer_command import PackageData, PrimerCommand
 
 MAX_GITHUB_COMMENT_LENGTH = 65536
 
 
 class CompareCommand(PrimerCommand):
     def run(self) -> None:
-        if self.config.batches is None:
-            main_data = self._load_json(self.config.base_file)
-            pr_data = self._load_json(self.config.new_file)
-        else:
-            main_data = {}
-            pr_data = {}
-            for idx in range(self.config.batches):
-                main_data.update(
-                    self._load_json(
-                        self.config.base_file.replace("BATCHIDX", "batch" + str(idx))
-                    )
-                )
-                pr_data.update(
-                    self._load_json(
-                        self.config.new_file.replace("BATCHIDX", "batch" + str(idx))
-                    )
-                )
-
-        missing_messages_data, new_messages_data = self._cross_reference(
-            main_data, pr_data
+        comparator = Comparator(
+            self.config.base_file, self.config.new_file, self.config.batches
         )
-        comment = self._create_comment(missing_messages_data, new_messages_data)
+        comment = self._create_comment(comparator)
         with open(self.primer_directory / "comment.txt", "w", encoding="utf-8") as f:
             f.write(comment)
 
-    @staticmethod
-    def _cross_reference(
-        main_data: PackageMessages, pr_data: PackageMessages
-    ) -> tuple[PackageMessages, PackageMessages]:
-        missing_messages_data: PackageMessages = {}
-        for package, data in main_data.items():
-            package_missing_messages: list[OldJsonExport] = []
-            for message in data["messages"]:
-                try:
-                    pr_data[package]["messages"].remove(message)
-                except ValueError:
-                    package_missing_messages.append(message)
-            missing_messages_data[package] = PackageData(
-                commit=pr_data[package]["commit"], messages=package_missing_messages
-            )
-        return missing_messages_data, pr_data
-
-    @staticmethod
-    def _load_json(file_path: Path | str) -> PackageMessages:
-        with open(file_path, encoding="utf-8") as f:
-            result: PackageMessages = json.load(f)
-        return result
-
-    def _create_comment(
-        self, all_missing_messages: PackageMessages, all_new_messages: PackageMessages
-    ) -> str:
+    def _create_comment(self, comparator: Comparator) -> str:
         comment = ""
-        for package, missing_messages in all_missing_messages.items():
+        for package, missing_messages, new_messages in comparator:
             if len(comment) >= MAX_GITHUB_COMMENT_LENGTH:
                 break
-            new_messages = all_new_messages[package]
-            if not missing_messages["messages"] and not new_messages["messages"]:
-                continue
             comment += self._create_comment_for_package(
                 package, new_messages, missing_messages
             )

tests/testutils/_primer/test_comparator.py

@@ -0,0 +1,67 @@
+# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
+# For details: https://github.com/pylint-dev/pylint/blob/main/LICENSE
+# Copyright (c) https://github.com/pylint-dev/pylint/blob/main/CONTRIBUTORS.txt
+
+from pathlib import Path
+
+import pytest
+
+from pylint.testutils._primer.comparator import Comparator
+
+FIXTURES_PATH = Path(__file__).parent / "fixtures"
+
+
+@pytest.mark.parametrize(
+    ("fixture", "main_file", "pr_file", "batches", "expected_packages", "expected_missing", "expected_new"),
+    [
+        pytest.param(
+            "both_empty", "main.json", "pr.json", None, [], [], [],
+            id="both_empty",
+        ),
+        pytest.param(
+            "no_change", "main.json", "pr.json", None, [], [], [],
+            id="no_change",
+        ),
+        pytest.param(
+            "new_message", "main.json", "pr.json", None,
+            ["astroid"], [0], [1],
+            id="new_message",
+        ),
+        pytest.param(
+            "removed_message", "main.json", "pr.json", None,
+            ["astroid"], [1], [0],
+            id="removed_message",
+        ),
+        pytest.param(
+            "message_changed", "main.json", "pr.json", None,
+            ["astroid"], [1], [1],
+            id="message_changed",
+        ),
+        pytest.param(
+            "batched", "main_BATCHIDX.json", "pr_BATCHIDX.json", 2,
+            ["astroid"], [2], [0],
+            id="batched",
+        ),
+    ],
+)
+def test_comparator(
+    fixture: str,
+    main_file: str,
+    pr_file: str,
+    batches: int | None,
+    expected_packages: list[str],
+    expected_missing: list[int],
+    expected_new: list[int],
+) -> None:
+    directory = FIXTURES_PATH / fixture
+    comparator = Comparator(str(directory / main_file), str(directory / pr_file), batches)
+    results = list(comparator)
+    assert len(results) == len(expected_packages)
+    for (package, missing, new), exp_pkg, exp_miss, exp_new in zip(
+        results, expected_packages, expected_missing, expected_new
+    ):
+        assert package == exp_pkg
+        assert len(missing["messages"]) == exp_miss
+        assert len(new["messages"]) == exp_new
+    # Iterating twice yields the same results
+    assert results == list(comparator)