test: cover the Word (.docx) report export

Add tests for the docx helpers and the report export, skipping the module when python-docx is unavailable, and add python-docx to the test environments so CI still exercises the export.

Author: ghostiee-11

lumen/tests/ai/test_docx.py

@@ -0,0 +1,153 @@
+from io import BytesIO
+
+import param
+import pytest
+
+try:
+    import lumen.ai  # noqa
+except ModuleNotFoundError:
+    pytest.skip("lumen.ai could not be imported, skipping tests.", allow_module_level=True)
+
+# python-docx is an optional dependency; skip the whole module without it.
+pytest.importorskip("docx")
+
+from docx import Document
+from panel.pane import Markdown
+
+from lumen.ai.report import Action, Report, Section
+
+
+class ChartAction(Action):
+    """Emits a pipeline-backed LumenEditor so the story catalog has real views."""
+
+    source = param.Parameter()
+
+    label = param.String(default="Chart")
+
+    async def _execute(self, context, **kwargs):
+        from lumen.ai.editors import LumenEditor
+        from lumen.pipeline import Pipeline
+
+        pipeline = Pipeline(source=self.source, table='tiny')
+        return [LumenEditor(component=pipeline, title=self.label)], {'text': self.label}
+
+
+class A(Action):
+
+    order = param.List()
+
+    async def _execute(self, context, **kwargs):
+        return [Markdown("A done")], {'text': 'A'}
+
+
+class B(Action):
+
+    order = param.List()
+
+    async def _execute(self, context, **kwargs):
+        return [Markdown("B done")], {'text': 'B'}
+
+
+def _docx_text(data: bytes) -> str:
+    doc = Document(BytesIO(data))
+    return "\n".join(p.text for p in doc.paragraphs)
+
+
+def test_docx_add_markdown_headings_bold_bullets():
+    from lumen.ai.export import docx_add_markdown
+
+    doc = Document()
+    docx_add_markdown(doc, "# Title\n\nSome **bold** text\n\n- item one\n- item two")
+
+    text = "\n".join(p.text for p in doc.paragraphs)
+    assert "Title" in text
+    assert "bold" in text
+    assert "item one" in text
+    assert any(p.style.name.startswith("Heading") for p in doc.paragraphs if p.text == "Title")
+    assert any(run.bold for p in doc.paragraphs for run in p.runs if run.text == "bold")
+
+
+def test_docx_add_table_from_dataframe():
+    import pandas as pd
+
+    from lumen.ai.export import docx_add_table
+
+    doc = Document()
+    df = pd.DataFrame({"category": ["A", "B"], "value": [1, 2]})
+    docx_add_table(doc, df)
+
+    assert len(doc.tables) == 1
+    table = doc.tables[0]
+    assert [c.text for c in table.rows[0].cells] == ["category", "value"]
+    assert len(table.rows) == 3  # header + 2 data rows
+
+
+def test_docx_add_chart_embeds_png():
+    from PIL import Image
+
+    from lumen.ai.export import docx_add_chart
+
+    class FakeEditor:
+        export_formats = ("png",)
+
+        def export(self, fmt):
+            buf = BytesIO()
+            Image.new("RGB", (4, 4), "white").save(buf, "PNG")
+            buf.seek(0)
+            return buf
+
+    doc = Document()
+    assert docx_add_chart(doc, FakeEditor()) is True
+    assert len(doc.inline_shapes) == 1
+
+
+async def test_report_to_docx_headings_and_text():
+    report = Report(
+        Section(A(order=[]), title='Section A'),
+        Section(B(order=[]), title='Section B'),
+        title='My Report',
+    )
+    await report.execute()
+
+    text = _docx_text(report.to_docx())
+    assert "My Report" in text
+    assert "Section A" in text
+    assert "A done" in text
+    assert "B done" in text
+
+
+async def test_report_to_docx_includes_story(llm, tiny_source):
+    from lumen.ai.agents.story import Story, StoryBlock
+
+    report = Report(
+        Section(ChartAction(source=tiny_source, label='Chart A'), title='Section A'),
+        title='My Report', llm=llm,
+    )
+    await report.execute()
+    llm.set_responses([Story(chain_of_thought="c", title="Big Picture", blocks=[
+        StoryBlock(prose="Overall narrative."), StoryBlock(view=1),
+    ])])
+    await report._annotate_report()
+
+    text = _docx_text(report.to_docx())
+    assert "Big Picture" in text
+    assert "Overall narrative." in text
+
+
+def test_export_menu_has_word_option():
+    report = Report(Section(A(order=[]), title='S'), title='R')
+    assert "docx" in [item.get("format") for item in report._export.items]
+    assert "docx" in [item.get("format") for item in report._dial.items]
+
+
+async def test_export_report_docx_returns_bytesio():
+    import io
+
+    report = Report(Section(A(order=[]), title='S'), title='My Report')
+    await report.execute()
+
+    result = await report._export_report({"format": "docx"})
+
+    assert isinstance(result, io.BytesIO)
+    assert report._download.filename == "My Report.docx"
+    Document(result)  # the bytes parse as a valid .docx

pixi.toml

@@ -10,8 +10,8 @@ install = 'python -m pip install --no-deps --disable-pip-version-check -e .'
 PYTHONIOENCODING = "utf-8"
 
 [environments]
-test-312 = ["py312", "test-core", "test", "ai", "sql", "xarray"]
-test-313 = ["py313", "test-core", "test", "ai", "sql", "xarray"]
+test-312 = ["py312", "test-core", "test", "ai", "sql", "xarray", "docx"]
+test-313 = ["py313", "test-core", "test", "ai", "sql", "xarray", "docx"]
 test-core = ["py313", "test-core"]
 docs = ["py312", "doc", "ai", "sql"]
 build = ["py312", "build"]
@@ -69,6 +69,9 @@ duckdb = "*"
 intake-sql = "*"
 sqlalchemy = "*"
 
+[feature.docx.dependencies]
+python-docx = "*"
+
 [feature.xarray.dependencies]
 xarray = "*"
 netcdf4 = "*"