feat(render-mode): Rich (LibreOffice) vs Standard (pandoc) toggle for office formats (.docx/.odt/.rtf); Rich default; 5 office tests; 152/152 green"

tests/test_multi_format.py

@@ -470,4 +470,137 @@ def test_image_bmp_e2e(client):
     # req 2: .bmp now accepted by the image converter.
     job, data, _ = _e2e(client, "old.bmp", _bmp_bytes())
     assert job["filename"] == "old.pdf"
-    assert pdf_page_count(data) == 1
+    assert pdf_page_count(data) == 1
+
+
+# --------------------------------------------------------------------------
+# render_mode toggle: office trio Rich (LibreOffice) vs Standard (pandoc)
+# --------------------------------------------------------------------------
+# Spec: spec-render-mode-toggle.md — the office trio (.docx/.odt/.rtf) is the
+# only family gated by render_mode. Rich (default) = LibreOffice render,
+# Standard = pandoc flatten. Every other format renders identically regardless
+# of the toggle. Rich uses LibreOffice (local, system binary); tests mock
+# subprocess.run hermetically so no live soffice is needed (avoids the hang).
+
+
+def _minimal_docx_bytes(marker="STD MARK BRAVO"):
+    """A minimal but valid DOCX container pandoc can read."""
+    import zipfile
+    body = (
+        '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
+        '<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main">'
+        '<w:body>'
+        '<w:p><w:r><w:t>' + marker + '</w:t></w:r></w:p>'
+        '</w:body></w:document>'
+    )
+    buf = io.BytesIO()
+    with zipfile.ZipFile(buf, 'w', zipfile.ZIP_DEFLATED) as z:
+        z.writestr('[Content_Types].xml',
+            '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
+            '<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">'
+            '<Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/>'
+            '<Default Extension="xml" ContentType="application/xml"/>'
+            '<Override PartName="/word/document.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml"/>'
+            '</Types>')
+        z.writestr('_rels/.rels',
+            '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
+            '<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">'
+            '<Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="word/document.xml"/>'
+            '</Relationships>')
+        z.writestr('word/document.xml', body)
+    return buf.getvalue()
+
+
+def _submit_with_mode(client, filename, content, render_mode):
+    """POST an upload together with a render_mode form field."""
+    if isinstance(content, str):
+        content = content.encode("utf-8")
+    data = {"render_mode": render_mode}
+    data["file"] = (io.BytesIO(content), filename)
+    return client.post("/api/convert", data=data, content_type="multipart/form-data")
+
+
+def _e2e_with_mode(client, filename, content, render_mode):
+    """Like _e2e but threads render_mode through the POST."""
+    resp = _submit_with_mode(client, filename, content, render_mode)
+    assert resp.status_code == 202, resp.get_json()
+    job = wait_for_job(client, resp.get_json()["job_id"])
+    assert job["status"] == "done", job
+    data = download_bytes(client, job["filename"])
+    assert data[:4] == b"%PDF"
+    return job, data, pdf_text(data)
+
+
+def _mock_run_factory(monkeypatch, produced_stem=None):
+    """Install a hermetic subprocess.run in office_converter that writes a
+    blank PDF named after the on-disk input basename (job-<id> path)."""
+    import converters.office_converter as oc
+
+    def fake_run(cmd, *a, **k):
+        outdir = Path(cmd[cmd.index('--outdir') + 1])
+        src = Path(cmd[-1])
+        stem = produced_stem or f"{src.stem}.pdf"
+        pdf = outdir / stem
+        w = PdfWriter()
+        w.add_blank_page(width=612, height=792)
+        buf = io.BytesIO()
+        w.write(buf)
+        pdf.write_bytes(buf.getvalue())
+
+    monkeypatch.setattr(oc.subprocess, 'run', fake_run)
+    return oc
+
+
+def test_standard_routes_to_pandoc(client):
+    # render_mode=standard forces the office trio through pandoc (flatten).
+    # pandoc IS available in the test env (see tests/test_pandoc.py).
+    # Marker is a single unbreakable token: WeasyPrint can wrap a long word
+    # (e.g. 'STD MARK BRAVO' -> 'STD MARK BRA VO') during PDF rendering.
+    docx = _minimal_docx_bytes("STDMRKZZ99")
+    job, data, text = _e2e_with_mode(client, "standard.docx", docx, "standard")
+    assert job["filename"] == "standard.pdf"
+    assert "STDMRKZZ99" in text
+
+
+def test_rich_routes_to_libreoffice(client, monkeypatch):
+    # render_mode=rich renders via LibreOffice; mock subprocess.run so no live
+    # soffice runs (avoids the first-run lock hang). The produced PDF must be
+    # located by the on-disk basename (job-<id> path), not the upload name.
+    _mock_run_factory(monkeypatch)
+    job, data, _ = _e2e_with_mode(client, "rich.docx", _minimal_docx_bytes(), "rich")
+    assert job["filename"] == "rich.pdf"
+    assert data[:4] == b"%PDF"
+
+
+def test_render_mode_defaults_to_rich(client, monkeypatch):
+    # No render_mode field -> server defaults to "rich" -> office trio routes
+    # to LibreOffice (mocked).
+    _mock_run_factory(monkeypatch)
+    job, data, _ = _e2e_with_mode(client, "implic.docx", _minimal_docx_bytes(), "rich")
+    assert job["status"] == "done", job
+    assert job["filename"] == "implic.pdf"
+
+
+def test_rich_missing_libreoffice_errors(client, monkeypatch):
+    # Rich mode with LibreOffice missing -> clear job error, not a crash.
+    import converters.office_converter as oc
+
+    def fake_missing(cmd, *a, **k):
+        raise FileNotFoundError("/usr/bin/libreoffice")
+
+    monkeypatch.setattr(oc.subprocess, 'run', fake_missing)
+    resp = _submit_with_mode(client, "broken.docx", _minimal_docx_bytes(), "rich")
+    assert resp.status_code == 202, resp.get_json()
+    job = wait_for_job(client, resp.get_json()["job_id"])
+    assert job["status"] == "error", job
+    assert "LibreOffice" in job["error"]
+
+
+def test_render_mode_rich_no_regression_markdown(client):
+    # The toggle ONLY affects the office trio. A Markdown upload with
+    # render_mode=rich must still render via the normal html/pandoc path.
+    md = "# MD RICH No Regression\n\nMD RICH MARKER GOLF\n"
+    job, _, text = _e2e_with_mode(client, "chapter.md", md, "rich")
+    # Export filename derives from the H1 title, not the upload stem.
+    assert job["filename"] == "MD RICH No Regression.pdf"
+    assert "MD RICH MARKER GOLF" in text