🚧 Add export feature and refined exclusion info

2026-09-13 12:04:42 +02:00
parent 1d67330f27
commit 3bbfd6c161
10 changed files with 637 additions and 16 deletions
@@ -0,0 +1,124 @@
+"""Tests for exclusion reporting and log-level policy (Sections 7 and 22).
+
+Excluded paths and the rule that matched them are logged at DEBUG level;
+the INFO summary carries only the in-scope/excluded split. Commit messages
+never contain exclusion details (Section 7).
+"""
+
+from __future__ import annotations
+
+import logging
+
+from git_auto_tracker.changes.filters import (
+    Exclusion,
+    exclusion_reason,
+    filter_change_set,
+    inclusion_reason,
+)
+from git_auto_tracker.changes.models import ChangeRecord, ChangeSet, ChangeType
+from git_auto_tracker.config import FileFilters
+
+
+class TestReasonFunctions:
+    def test_exclusion_reason_names_the_pattern(self) -> None:
+        filters = FileFilters(include=["**/*"], exclude=["runtime/**", "**/*.log"])
+        assert exclusion_reason("runtime/cache.dat", filters) == "matched exclude 'runtime/**'"
+        assert exclusion_reason("logs/a.log", filters) == "matched exclude '**/*.log'"
+        assert exclusion_reason("keep.xml", filters) is None
+
+    def test_inclusion_reason_when_out_of_scope(self) -> None:
+        filters = FileFilters(include=["Stations/**"], exclude=[])
+        assert inclusion_reason("System/global.xml", filters) == "no include pattern matched"
+        assert inclusion_reason("Stations/PS1.xml", filters) is None
+
+    def test_empty_include_means_everything_included(self) -> None:
+        filters = FileFilters(include=[], exclude=["**/*.tmp"])
+        assert inclusion_reason("anything.xml", filters) is None
+
+    def test_legacy_predicates_still_work(self) -> None:
+        filters = FileFilters(include=["**/*"], exclude=["**/*.log"])
+        from git_auto_tracker.changes.filters import is_excluded, is_included
+
+        assert is_included("a.xml", filters) is True
+        assert is_excluded("a.log", filters) is True
+        assert is_excluded("a.xml", filters) is False
+
+
+class TestFilterChangeSet:
+    def _change_set(self) -> ChangeSet:
+        return ChangeSet(
+            added=[
+                ChangeRecord(path="runtime/cache.dat", change_type=ChangeType.ADDED),
+                ChangeRecord(path="Stations/PS1.xml", change_type=ChangeType.ADDED),
+            ],
+            modified=[
+                ChangeRecord(path="logs/app.log", change_type=ChangeType.MODIFIED),
+                ChangeRecord(path="System/global.xml", change_type=ChangeType.MODIFIED),
+            ],
+        )
+
+    def test_exclusions_carry_reasons(self) -> None:
+        filters = FileFilters(include=["**/*"], exclude=["runtime/**", "**/*.log"])
+        kept, excluded = filter_change_set(self._change_set(), filters)
+
+        assert [r.path for r in kept.added] == ["Stations/PS1.xml"]
+        assert [r.path for r in kept.modified] == ["System/global.xml"]
+        by_path = {e.path: e.reason for e in excluded}
+        assert by_path["runtime/cache.dat"] == "matched exclude 'runtime/**'"
+        assert by_path["logs/app.log"] == "matched exclude '**/*.log'"
+
+    def test_out_of_scope_excluded_with_include_reason(self) -> None:
+        filters = FileFilters(include=["Stations/**"], exclude=["**/*.log"])
+        _, excluded = filter_change_set(self._change_set(), filters)
+        by_path = {e.path: e.reason for e in excluded}
+        assert by_path["System/global.xml"] == "no include pattern matched"
+        assert by_path["runtime/cache.dat"] == "no include pattern matched"
+        assert by_path["logs/app.log"] == "matched exclude '**/*.log'"
+
+    def test_exclusion_is_dataclass(self) -> None:
+        record = Exclusion(path="x.tmp", reason="matched exclude '**/*.tmp'")
+        assert record.path == "x.tmp"
+        assert "x.tmp" in record.reason or "**/*.tmp" in record.reason
+
+
+class TestLogLevels:
+    """INFO summary only; per-file detail at DEBUG (application wiring)."""
+
+    def test_collect_logs_split_and_debug_details(
+        self, git_repo, tracker_config, tmp_path, caplog
+    ) -> None:
+        from git_auto_tracker.application import ApplicationController
+        from git_auto_tracker.config import AppConfig
+        from git_auto_tracker.logging.setup import setup_logging
+
+        (git_repo / "junk.log").write_text("noise", encoding="utf-8")
+        (git_repo / "real.xml").write_text("<r/>", encoding="utf-8")
+
+        config = AppConfig.from_dict(tracker_config, config_dir=tmp_path)
+        setup_logging(
+            config.logging,
+            config_dir=tmp_path,
+            repository_path=config.repository_path,
+            override_level="DEBUG",
+        )
+        # The app logger does not propagate (setup_logging), so caplog's
+        # root-attached handler never sees the records; attach it directly.
+        app_logger = logging.getLogger("git_auto_tracker")
+        app_logger.addHandler(caplog.handler)
+        try:
+            controller = ApplicationController(config)
+            outcome = controller.collect()
+        finally:
+            app_logger.removeHandler(caplog.handler)
+
+        assert outcome.change_count == 1  # real.xml only
+        debug_text = caplog.text
+
+        # INFO summary: the split, no per-file exclusion detail
+        assert "Detected 2 change(s): 1 in scope, 1 excluded by filters" in debug_text
+        # DEBUG detail: path plus the matching rule
+        assert "Excluded: junk.log (matched exclude '**/*.log')" in debug_text
+        # Author resolution details land at DEBUG too
+        assert "Author resolution detail: real.xml -> source=" in debug_text
+        # The exclusion must NOT appear in the commit message (Section 7)
+        assert "junk.log" not in (outcome.commit_message or "")
@@ -0,0 +1,118 @@
+"""Unit tests for history export parsing (exporting.history).
+
+The commit-message grammar is defined by message_builder (Section 15);
+these tests pin the parser against it.
+"""
+
+from __future__ import annotations
+
+from git_auto_tracker.exporting.history import (
+    HistoryRecord,
+    parse_message,
+    to_rows,
+)
+
+_MESSAGE = """SCADA automatic change capture - 2026-09-12 08:00:03+02:00
+
+Summary:
+  Added: 1
+  Modified: 1
+  Deleted: 1
+  Renamed: 0
+
+Changes:
+  A ; Stations/PS101/pump2.xml ; DOMAIN\\jsmith ; 2026-09-12 07:41:22+02:00
+  M ; System/global.xml ; UNKNOWN ; 2026-09-12 07:45:01+02:00
+  D ; Stations/PS027/old-config.xml
+
+Tracker-Version: 1.0.0
+Tracker-Run: 01M2ADPMSHSA2TZKVH6PSVDVA0
+Tracker-Host: SCADA-SRV-01
+Tracker-Changes: 3
+"""
+
+
+class TestParseMessage:
+    def test_full_message(self) -> None:
+        record = parse_message(_MESSAGE, ";")
+        assert record is not None
+        assert record.collected_at == "2026-09-12 08:00:03+02:00"
+        assert record.run_id == "01M2ADPMSHSA2TZKVH6PSVDVA0"
+        assert record.host == "SCADA-SRV-01"
+        assert len(record.rows) == 3
+
+        added = record.rows[0]
+        assert added["letter"] == "A"
+        assert added["path"] == "Stations/PS101/pump2.xml"
+        assert added["author"] == "DOMAIN\\jsmith"
+        assert added["mtime"] == "2026-09-12 07:41:22+02:00"
+
+        modified = record.rows[1]
+        assert modified["letter"] == "M"
+        assert modified["author"] == "UNKNOWN"
+
+        deleted = record.rows[2]
+        assert deleted["letter"] == "D"
+        assert deleted["path"] == "Stations/PS027/old-config.xml"
+        assert deleted["author"] == ""  # bare line (FR-013)
+
+    def test_paths_with_semicolon_inside_are_not_broken(self) -> None:
+        # A path containing ';' is ambiguous by design; the parser splits on
+        # ' ; ' so a lone ';' inside a filename keeps the line intact here.
+        message = _MESSAGE.replace(
+            "Stations/PS101/pump2.xml", " Stations/PS101;backup.xml".strip()
+        )
+        record = parse_message(message, ";")
+        assert record is not None
+        assert record.rows[0]["path"] == "Stations/PS101;backup.xml"
+
+    def test_non_tracker_message_returns_none(self) -> None:
+        assert parse_message("initial import\n", ";") is None
+        assert parse_message("", ";") is None
+
+    def test_custom_separator(self) -> None:
+        message = _MESSAGE.replace(" ; ", " | ")
+        record = parse_message(message, "|")
+        assert record is not None
+        assert len(record.rows) == 3
+        assert record.rows[0]["path"] == "Stations/PS101/pump2.xml"
+
+    def test_wrong_separator_yields_no_rows_but_is_tracker_commit(self) -> None:
+        # Written with ';', parsed with '|': recognized but not splittable.
+        record = parse_message(_MESSAGE, "|")
+        assert record is not None
+        assert record.collected_at  # tracker commit detected
+        assert record.rows == []    # but no reliable rows
+
+    def test_rename_line(self) -> None:
+        message = _MESSAGE.replace(
+            "  D ; Stations/PS027/old-config.xml",
+            "  R ; Stations/PS027/old.xml -> Stations/PS027/new.xml",
+        )
+        record = parse_message(message, ";")
+        assert record is not None
+        row = record.rows[2]
+        assert row["letter"] == "R"
+        assert row["previous_path"] == "Stations/PS027/old.xml"
+        assert row["path"] == "Stations/PS027/new.xml"
+
+
+class TestToRows:
+    def test_long_format(self) -> None:
+        record = parse_message(_MESSAGE, ";")
+        assert record is not None
+        record.commit = "a1b2c3d"
+        rows = to_rows([record])
+        assert len(rows) == 3
+        first = rows[0]
+        assert first["commit"] == "a1b2c3d"
+        assert first["run_id"] == "01M2ADPMSHSA2TZKVH6PSVDVA0"
+        assert first["letter"] == "A"
+        assert set(first) == {
+            "commit", "collected_at", "run_id", "host",
+            "letter", "path", "previous_path", "author", "mtime",
+        }
+
+    def test_empty(self) -> None:
+        assert to_rows([]) == []
+        assert to_rows([HistoryRecord(commit="x")]) == []