rdocx: write schema-valid, Word-like markup for equations, notes, links, cells, numbering, TOC and replace (#432)

README.md

@@ -39,7 +39,7 @@ rows are the enforced release-mode bounds plus one dated observation.
 
 | Measurement | Value | Version | Platform | Build mode | Input | Command | Statistic | Measured on |
 |---|---|---|---|---|---|---|---|---|
-| Crates.io archive: rdocx | 4,105,912 compressed bytes, 54,118,469 member bytes, 37 members | 0.16.4 | macOS 26.6.2, Apple M5 Max, arm64 | `cargo publish --workspace --dry-run --allow-dirty --locked` | Tracked `rdocx` package inventory | `python3 scripts/readme_doctests.py --record-measurements` | gzip archive bytes, tar member bytes, tar member count | 2026-10-28 |
+| Crates.io archive: rdocx | 4,112,736 compressed bytes, 54,143,654 member bytes, 37 members | 0.16.4 | macOS 26.6.2, Apple M5 Max, arm64 | `cargo publish --workspace --dry-run --allow-dirty --locked` | Tracked `rdocx` package inventory | `python3 scripts/readme_doctests.py --record-measurements` | gzip archive bytes, tar member bytes, tar member count | 2026-10-28 |
 | Large-document layout throughput | minimum 250 pages/s, observed 31,019.1 pages/s | rdocx 0.14.0 | macOS 26.6.2, Apple M5 Max, arm64 | release, one test thread | 1,000 one-page paragraphs with deterministic fonts | `cargo test -p rdocx --test regression_test --release a_thousand_page_document_paginates_and_renders_within_the_declared_limits -- --ignored --exact --nocapture --test-threads=1` | pages per wall-clock second | 2026-09-19 |
 | Large-document layout peak allocation | maximum 64 MiB, observed 29.03 MiB | rdocx 0.14.0 | macOS 26.6.2, Apple M5 Max, arm64 | release, one test thread | 1,000 one-page paragraphs with deterministic fonts | `cargo test -p rdocx --test regression_test --release a_thousand_page_document_paginates_and_renders_within_the_declared_limits -- --ignored --exact --nocapture --test-threads=1` | peak live allocation | 2026-09-19 |
 | Large-document PDF throughput | minimum 1,000 pages/s, observed 60,058.0 pages/s | rdocx 0.14.0 | macOS 26.6.2, Apple M5 Max, arm64 | release, one test thread | 1,000 deterministic layout pages | `cargo test -p rdocx --test regression_test --release a_thousand_page_document_paginates_and_renders_within_the_declared_limits -- --ignored --exact --nocapture --test-threads=1` | pages per wall-clock second | 2026-09-19 |

crates/rdocx-oxml/README.md

@@ -17,7 +17,7 @@ schema order, and retains unmodelled XML alongside typed edits.
 
 | Measurement | Value | Version | Platform | Build mode | Input | Command | Statistic | Measured on |
 |---|---|---|---|---|---|---|---|---|
-| Crates.io archive: rdocx-oxml | 565,559 compressed bytes, 3,484,893 member bytes, 32 members | 0.16.4 | macOS 26.6.2, Apple M5 Max, arm64 | `cargo publish --workspace --dry-run --allow-dirty --locked` | Tracked `rdocx-oxml` package inventory | `python3 scripts/readme_doctests.py --record-measurements` | gzip archive bytes, tar member bytes, tar member count | 2026-10-28 |
+| Crates.io archive: rdocx-oxml | 565,548 compressed bytes, 3,484,892 member bytes, 32 members | 0.16.4 | macOS 26.6.2, Apple M5 Max, arm64 | `cargo publish --workspace --dry-run --allow-dirty --locked` | Tracked `rdocx-oxml` package inventory | `python3 scripts/readme_doctests.py --record-measurements` | gzip archive bytes, tar member bytes, tar member count | 2026-10-28 |
 
 ## Use it when
 

crates/rdocx-oxml/src/math.rs

@@ -1028,7 +1028,7 @@ impl MathRadical {
             degree: MathArgument::default(),
             base,
             hide_degree: true,
-            degree_present: false,
+            degree_present: true,
             preservation: Preservation::default(),
         }
     }

crates/rdocx-py/src/paragraph.rs

@@ -621,11 +621,23 @@ impl PyParagraph {
                 }
                 None => None,
             };
+            let link_style = match url {
+                Some(_) => Some(
+                    candidate
+                        .ensure_hyperlink_style()
+                        .map_err(|error| crate::rdocx_to_pyerr(py, error))?,
+                ),
+                None => None,
+            };
             let append = |paragraph: &mut rdocx::Paragraph<'_>| {
                 let run_index = paragraph.run_count();
                 match &relationship_id {
                     Some(relationship_id) => {
-                        paragraph.add_hyperlink_with_tooltip(text, relationship_id, tooltip)?;
+                        let mut run =
+                            paragraph.add_hyperlink_with_tooltip(text, relationship_id, tooltip)?;
+                        if let Some(link_style) = &link_style {
+                            run.set_style(link_style);
+                        }
                     }
                     None => {
                         paragraph.add_run(text);

crates/rdocx-py/tests/test_core.py

@@ -5265,7 +5265,8 @@ def test_fx201_note_reference_creation_uses_checked_paragraph():
     reference = document.add_paragraph("reference")
     retired_run = reference.runs[0]
     note = document.add_footnote(reference, "note content")
-    assert note.paragraphs[0].text == "note content"
+    # The space run Word writes after the note's reference mark.
+    assert note.paragraphs[0].text == " note content"
     assert reference.text == "reference"
     with pytest.raises(rdocx.StaleElementError):
         _ = retired_run.text
@@ -5695,7 +5696,7 @@ def test_fx201_checked_note_cell_boundary_preserves_other_owners_and_refuses_sta
     table.cell(0, 0).text = "cell reference"
     reference = document.tables[0].cell(0, 0).paragraphs[0].runs[0]
     note = getattr(document, "add_" + family)(reference, family + " content")
-    assert note.paragraphs[0].text == family + " content"
+    assert note.paragraphs[0].text == " " + family + " content"
     before = document.to_bytes()
     with pytest.raises(rdocx.StaleElementError):
         getattr(document, "add_" + family)(reference, "stale refused")
@@ -10048,7 +10049,7 @@ def test_form_protection_and_permission_composition():
         (cell_run, "cell held"),
         (header, "header held"),
         (header_run, "header held"),
-        (note, "note held"),
+        (note, " note held"),
         (note_run, "note held"),
     )
     before = document.to_bytes()
@@ -10094,7 +10095,7 @@ def test_form_protection_and_permission_composition():
     assert document.paragraphs[0].text == body.text == "body held"
     assert document.tables[0].cell(0, 0).text == cell.text == "cell held"
     assert document.section(0).header.paragraphs[0].text == header.text == "header held"
-    assert _story_paragraph(document, "footnote").text == note.text == "note held"
+    assert _story_paragraph(document, "footnote").text == note.text == " note held"
     removed = document.remove_document_protection()
     assert removed is not None and _protection_fields(removed) == _protection_fields(full)
     with pytest.raises(AttributeError):
@@ -15251,3 +15252,68 @@ def test_fx268_picture_normalization_reuses_actual_checked_rounds(publisher, par
     assert staging[5] <= 3 * pieces + 20, staging
     assert picture.bytes == _one_pixel_png()
     assert len(rdocx.Document.from_bytes(document.to_bytes()).pictures()) == 1
+
+
+def test_issue432_python_links_take_the_hyperlink_character_style():
+    import rdocx
+
+    def link_runs(document):
+        xml = document.part_xml("word/document.xml")
+        return re.findall(r"<w:hyperlink\b.*?</w:hyperlink>", xml, re.S)
+
+    styled = '<w:rStyle w:val="Hyperlink"/>'
+    document = rdocx.Document()
+    document.add_paragraph("See ").add_hyperlink("the site", "https://example.com")
+    document.add_heading("Target", 1)
+    document.add_paragraph("And ")
+    document.add_heading_hyperlink(document.paragraphs[2], document.paragraphs[1], "Target")
+    links = link_runs(document)
+    assert len(links) == 2 and all(styled in link for link in links), links
+    styles = document.part_xml("word/styles.xml")
+    assert styles.count('w:styleId="Hyperlink"') == 1, styles
+
+    template = rdocx.Document()
+    template.add_style("Hyperlink", style_type="character", color="FF0000", underline=True)
+    for document in (
+        rdocx.Document.from_markdown("See [the site](https://example.com).", template=template)[0],
+        rdocx.Document.from_html('<p>See <a href="https://example.com">the site</a>.</p>')[0],
+    ):
+        links = link_runs(document)
+        assert links and all(styled in link for link in links), links
+    # The run of a plain add_run is left alone.
+    document = rdocx.Document()
+    document.add_paragraph("plain").add_run(" text")
+    assert "Hyperlink" not in document.part_xml("word/document.xml")
+
+
+def test_issue432_a_cell_with_only_paragraph_properties_merges_as_empty():
+    import rdocx
+
+    def build():
+        document = rdocx.Document()
+        table = document.add_table(1, 3, style="Table Grid")
+        table.cell(0, 0).text = "Total"
+        table.cell(0, 1).paragraphs[0].alignment = rdocx.WD_ALIGN_PARAGRAPH.RIGHT
+        table.cell(0, 2).text = "Kept"
+        return document
+
+    document = build()
+    document.tables[0].set_cell_grid_span(0, 0, 2)
+    assert [p.text for p in document.tables[0].cell(0, 0).paragraphs] == ["Total"]
+    assert [c.text for c in document.tables[0].rows[0].cells] == ["Total", "Kept"]
+
+    document = build()
+    merged = document.tables[0].cell(0, 0).merge(document.tables[0].cell(0, 1))
+    assert [p.text for p in merged.paragraphs] == ["Total"]
+
+    document = rdocx.Document()
+    table = document.add_table(2, 1)
+    table.cell(0, 0).text = "Total"
+    table.cell(1, 0).paragraphs[0].alignment = rdocx.WD_ALIGN_PARAGRAPH.RIGHT
+    merged = document.tables[0].cell(0, 0).merge(document.tables[0].cell(1, 0))
+    assert [p.text for p in merged.paragraphs] == ["Total"]
+
+    # A cell that holds text is still refused.
+    document = build()
+    with pytest.raises(Exception, match="nonempty cell"):
+        document.tables[0].set_cell_grid_span(0, 1, 2)