实现 ClinDB 第一批清洗组件

tests/test_html_table_layout.py

@@ -0,0 +1,63 @@
+from __future__ import annotations
+
+import pytest
+
+from mdpolish import Pipeline, RunStatus
+from mdpolish.components import HtmlTableLayoutComponent
+
+TABLE = '<table class="x"><tr><td colspan="2">A</td></tr><tr><td>B</td><td>C</td></tr></table>'
+
+
+def transform(markdown: str):  # type: ignore[no-untyped-def]
+    return Pipeline([HtmlTableLayoutComponent()]).transform(markdown)
+
+
+def test_expands_rows_without_changing_tags_attributes_or_cells() -> None:
+    result = transform(f"before\n{TABLE}\nafter")
+
+    assert result.status is RunStatus.SUCCESS
+    assert result.output_markdown == (
+        "before\n<table class=\"x\">\n"
+        "  <tr><td colspan=\"2\">A</td></tr>\n"
+        "  <tr><td>B</td><td>C</td></tr>\n"
+        "</table>\nafter"
+    )
+    assert len(result.changes) == 1
+
+
+@pytest.mark.parametrize("line_ending", ["\n", "\r\n", "\r"])
+def test_uses_the_documents_single_line_ending_style(line_ending: str) -> None:
+    markdown = f"before{line_ending}{TABLE}{line_ending}after"
+    result = transform(markdown)
+    assert result.output_markdown is not None
+    assert f"<table class=\"x\">{line_ending}  <tr>" in result.output_markdown
+
+
+def test_table_only_document_uses_lf() -> None:
+    assert transform(TABLE).output_markdown == (
+        '<table class="x">\n'
+        '  <tr><td colspan="2">A</td></tr>\n'
+        "  <tr><td>B</td><td>C</td></tr>\n"
+        "</table>"
+    )
+
+
+@pytest.mark.parametrize(
+    "markdown",
+    [
+        "before\n<table><tr><td>A</td></tr>\r\nafter",
+        "<table>\n  <tr><td>A</td></tr>\n</table>",
+        "<table><tr><td>A</tr></table>",
+        "<table><tbody><tr><td>A</td></tr></tbody></table>",
+        "<table><tr><td><em>A</em></td></tr></table>",
+    ],
+)
+def test_mixed_multiline_or_non_strict_tables_are_preserved(markdown: str) -> None:
+    assert transform(markdown).output_markdown == markdown
+
+
+def test_successful_output_is_idempotent() -> None:
+    pipeline = Pipeline([HtmlTableLayoutComponent()])
+    first = pipeline.transform(TABLE)
+    assert first.output_markdown is not None
+    assert pipeline.transform(first.output_markdown).changes == ()