diff --git a/src/documents/tests/test_zone_ocr.py b/src/documents/tests/test_zone_ocr.py index 5040363ac6..da65f81874 100644 --- a/src/documents/tests/test_zone_ocr.py +++ b/src/documents/tests/test_zone_ocr.py @@ -24,6 +24,7 @@ from documents.models import MatchingModel from documents.models import OcrTemplate from documents.models import OcrTemplateZone from documents.parsers import ParseError +from documents.versioning import get_root_document from documents.zone_ocr import MAX_PARSED_TEXT_LENGTH from documents.zone_ocr import ZoneExtraction from documents.zone_ocr import ZoneOcrReason @@ -39,11 +40,12 @@ from documents.zone_ocr import _ocr_text from documents.zone_ocr import _parse_date from documents.zone_ocr import _parse_monetary from documents.zone_ocr import _parse_number +from documents.zone_ocr import _plan_template from documents.zone_ocr import _prepare_field_update -from documents.zone_ocr import _process_template from documents.zone_ocr import _render_pages from documents.zone_ocr import _resolve_page_idx from documents.zone_ocr import _zone_language +from documents.zone_ocr import apply_zone_plans from documents.zone_ocr import extract_zone_preview from documents.zone_ocr import get_templates_for_content from documents.zone_ocr import get_templates_for_document @@ -66,6 +68,7 @@ if TYPE_CHECKING: from pytest_django.fixtures import SettingsWrapper from pytest_mock import MockerFixture + from documents.zone_ocr import ZoneOcrResult from paperless_testing.dirs import PaperlessDirs @@ -107,6 +110,18 @@ def ocr_outputs(*values: str) -> list[subprocess.CompletedProcess[str]]: ] +def plan_and_apply_template( + document: Document, + page_images: dict[int, Path], + template: OcrTemplate, + zones: list[OcrTemplateZone], + tmp_path: Path, +) -> list[ZoneOcrResult]: + """Plan one template over rendered pages and apply it to the root document.""" + plan = _plan_template(document.page_count, page_images, template, zones, tmp_path) + return apply_zone_plans(get_root_document(document), [plan], label="test") + + def make_visible_template(name: str, **kwargs: Any) -> OcrTemplate: """Build a 100x100 template matching "Northwind".""" kwargs.setdefault("match", "Northwind") @@ -602,7 +617,7 @@ class TestSavedZoneExtraction: ] save = mocker.patch.object(document, "save", wraps=document.save) - _process_template( + plan_and_apply_template( document, {0: page}, template, @@ -642,7 +657,7 @@ class TestSavedZoneExtraction: for value in ["New title", "ASN 42", "2026-04-13", "no"] ] - _process_template( + plan_and_apply_template( document, {0: page}, template, @@ -699,7 +714,7 @@ class TestSavedZoneExtraction: ] ] - _process_template( + plan_and_apply_template( document, {0: page}, template, @@ -729,8 +744,8 @@ class TestSavedZoneExtraction: WHEN: - The template is processed THEN: - - The error propagates, the write is rolled back and the in-memory - document is restored + - The template is reported as failed, the write is rolled back and + the in-memory document is restored """ field = CustomFieldFactory(name="Number", data_type="integer") zones = [ @@ -754,14 +769,17 @@ class TestSavedZoneExtraction: mocker.patch.object(document, "save", side_effect=fail_after_save) - with pytest.raises(RuntimeError, match="save failed"): - _process_template( - document, - {0: page}, - template, - zones, - paperless_dirs.scratch_dir, - ) + results = plan_and_apply_template( + document, + {0: page}, + template, + zones, + paperless_dirs.scratch_dir, + ) + + assert [result.reason for result in results] == [ + ZoneOcrReason.TEMPLATE_FAILED, + ] * 2 assert document.title == "Original" document.refresh_from_db() @@ -968,7 +986,7 @@ class TestSavedZoneExtraction: mock_zone_ocr_run.side_effect = read_crop - _process_template( + plan_and_apply_template( document, {0: page}, template, @@ -978,33 +996,6 @@ class TestSavedZoneExtraction: assert document.title == "extracted" - def test_missing_page_returns_skipped_result( - self, - paperless_dirs: PaperlessDirs, - document: Document, - template: OcrTemplate, - ) -> None: - """ - GIVEN: - - A zone on the last page and no rendered pages - WHEN: - - The template is processed - THEN: - - The zone result is skipped as a missing page with no value - """ - zone = make_zone(template, "Last", page=-1) - - results = _process_template( - document, - {}, - template, - [zone], - paperless_dirs.scratch_dir, - ) - - assert results[0].reason == ZoneOcrReason.MISSING_PAGE - assert results[0].value is None - def test_empty_combination_cannot_clear_a_field( self, mock_zone_ocr_run: MagicMock, @@ -1025,7 +1016,7 @@ class TestSavedZoneExtraction: template.combine_formats = {"title": "{Unknown}"} mock_zone_ocr_run.return_value.stdout = "extracted" - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, @@ -1065,7 +1056,7 @@ class TestSavedZoneExtraction: for value in ["", "rejected", "abc", "orphan"] ] - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, @@ -1320,7 +1311,7 @@ class TestSavedZoneExtraction: make_zone(template, "ASN", "asn") mock_zone_ocr_run.side_effect = ocr_outputs("New title", "42") - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, @@ -1357,7 +1348,7 @@ class TestSavedZoneExtraction: make_zone(template, "ASN", "asn") mock_zone_ocr_run.side_effect = ocr_outputs("42") - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, @@ -1399,7 +1390,7 @@ class TestSavedZoneExtraction: preview = extract_zone_preview(page, zone, 100, 50, 1, is_image=True) assert preview.raw_text == raw_text assert preview.value == raw_text.upper() - _process_template( + plan_and_apply_template( document, {0: page}, template, @@ -1466,7 +1457,7 @@ class TestSavedZoneExtraction: texts = ["05", "2024"] if first_text is None else [first_text, "05", "2024"] mock_zone_ocr_run.side_effect = ocr_outputs(*texts) - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, @@ -1504,7 +1495,7 @@ class TestSavedZoneExtraction: template.combine_formats = {"title": "{First} / {Second}"} mock_zone_ocr_run.side_effect = ocr_outputs("", "two") - results = _process_template( + results = plan_and_apply_template( document, {0: page}, template, diff --git a/src/documents/zone_ocr.py b/src/documents/zone_ocr.py index 71b8557423..c74d7df801 100644 --- a/src/documents/zone_ocr.py +++ b/src/documents/zone_ocr.py @@ -494,20 +494,6 @@ def _resolve_page_idx(page_value: int | None, page_count: int | None) -> int | N return max((page_value or 1) - 1, 0) -def _process_template( - document: Document, - page_images: dict[int, Path], - template: OcrTemplate, - zones: list[OcrTemplateZone], - tmp_path: Path, -) -> list[ZoneOcrResult]: - """Extract and combine zone values, then apply a template's updates.""" - plan = _plan_template(document.page_count, page_images, template, zones, tmp_path) - if plan.updates: - _apply_template_updates(get_root_document(document), plan.updates) - return plan.results - - def _plan_template( page_count: int | None, page_images: dict[int, Path],