open-document-lib 1.2.0__tar.gz → 1.4.0__tar.gz

{open_document_lib-1.2.0/open_document_lib.egg-info → open_document_lib-1.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: open-document-lib
-Version: 1.2.0
+Version: 1.4.0
 Summary: Standard-library toolkit for reading, editing, and writing OpenDocument Format files (ODT, ODP, ODS, ODG)
 Author: Patrick Leiverkus
 License: MIT
@@ -30,6 +30,8 @@ Provides-Extra: scholarly
 Requires-Dist: bibtexparser>=1.4; extra == "scholarly"
 Provides-Extra: validate
 Requires-Dist: lxml>=4.9; extra == "validate"
+Provides-Extra: render
+Requires-Dist: Pillow>=10; extra == "render"
 Provides-Extra: dev
 Requires-Dist: pytest>=8.0; extra == "dev"
 Requires-Dist: ruff>=0.9; extra == "dev"
@@ -37,6 +39,7 @@ Requires-Dist: pytest-cov>=5.0; extra == "dev"
 Requires-Dist: hypothesis>=6.0; extra == "dev"
 Requires-Dist: build>=1.0; extra == "dev"
 Requires-Dist: mypy>=1.8; extra == "dev"
+Requires-Dist: Pillow>=10; extra == "dev"
 Dynamic: license-file
 
 # open-document-lib
@@ -190,6 +193,14 @@ helpers) is internal and may change without notice.
 | `find_pandoc() -> str \| None` | Locate the `pandoc` binary. |
 | `latex_to_mathml(latex) -> bytes` | Convert a LaTeX snippet to MathML via Pandoc. |
 
+### Rendering
+
+| Signature | Description |
+|---|---|
+| `render_to_pdf(odf_path, outdir) -> Path` | Render an ODF file to PDF via LibreOffice (isolated profile). |
+| `pdf_to_pngs(pdf_path, outdir, dpi=150) -> list[Path]` | Render each PDF page to a PNG via `pdftoppm` (Poppler). |
+| `build_contact_sheet(images, output_path, columns=0) -> Path` | Compose page thumbnails into one labelled grid image. Requires Pillow (`pip install open-document-lib[render]`). |
+
 ## License
 
 MIT. See [LICENSE](https://github.com/leiverkus/open-document-skills/blob/main/LICENSE).

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/README.md

@@ -17,7 +17,7 @@ Four self-contained skills for Codex, Claude Code, and OpenCode that teach an ag
 
 ```bash
 # Generate, edit, validate, version — all from the agent shell:
-python skills/odt/scripts/create_minimal_odt.py spec.json doc.odt
+python skills/odt/scripts/create_from_markdown.py article.md doc.odt   # rich text from Markdown
 python skills/odt/scripts/replace_text.py doc.odt "{{NAME}}" "Patrick" -o out.odt
 python skills/odt/scripts/pack_fodt.py out.odt -o out.fodt   # diff-friendly XML
 python skills/odt/scripts/validate_refs.py out.odt
@@ -27,7 +27,7 @@ python skills/odt/scripts/validate_refs.py out.odt
 
 | Skill | LibreOffice app | Smithery | Triggers |
 | --- | --- | --- | --- |
-| [`odt`](skills/odt) | Writer | [smithery.ai/skills/leiverkus/odt](https://smithery.ai/skills/leiverkus/odt) | edit ODT, footnotes, citations (BibTeX/CSL-JSON), bookmarks, cross-references, figure/table sequences, MathML formulas, render to PDF |
+| [`odt`](skills/odt) | Writer | [smithery.ai/skills/leiverkus/odt](https://smithery.ai/skills/leiverkus/odt) | author from Markdown, edit ODT, footnotes, citations (BibTeX/CSL-JSON), bookmarks, cross-references, figure/table sequences, MathML formulas, render to PDF |
 | [`odp`](skills/odp) | Impress | [smithery.ai/skills/leiverkus/odp](https://smithery.ai/skills/leiverkus/odp) | clone slide, edit notes, add image, animations (entrance/exit/emphasis/motion), slide transitions, master-page customization (background, header/footer, logo), render deck |
 | [`ods`](skills/ods) | Calc | [smithery.ai/skills/leiverkus/ods](https://smithery.ai/skills/leiverkus/ods) | set cells/formulas, named ranges, dropdowns + data validation, embedded charts (bar/line/pie/scatter), export CSV, recalculate |
 | [`odg`](skills/odg) | Draw | [smithery.ai/skills/leiverkus/odg](https://smithery.ai/skills/leiverkus/odg) | edit labels, add shape image, glue points, connectors with shape binding, groups, flowcharts, org charts, export SVG/PNG |

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/docs/library-api.md

@@ -149,6 +149,14 @@ helpers) is internal and may change without notice.
 | `find_pandoc() -> str \| None` | Locate the `pandoc` binary. |
 | `latex_to_mathml(latex) -> bytes` | Convert a LaTeX snippet to MathML via Pandoc. |
 
+### Rendering
+
+| Signature | Description |
+|---|---|
+| `render_to_pdf(odf_path, outdir) -> Path` | Render an ODF file to PDF via LibreOffice (isolated profile). |
+| `pdf_to_pngs(pdf_path, outdir, dpi=150) -> list[Path]` | Render each PDF page to a PNG via `pdftoppm` (Poppler). |
+| `build_contact_sheet(images, output_path, columns=0) -> Path` | Compose page thumbnails into one labelled grid image. Requires Pillow (`pip install open-document-lib[render]`). |
+
 ## License
 
 MIT. See [LICENSE](https://github.com/leiverkus/open-document-skills/blob/main/LICENSE).

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/odf_lib/__init__.py

@@ -23,6 +23,7 @@ from odf_lib.odf_common import (
     ODF_NAMESPACES,
     VERSION,
     apply_strict_schema_check,
+    build_contact_sheet,
     clear_children,
     copy_into_package,
     copy_with_multiple_members,
@@ -42,6 +43,8 @@ from odf_lib.odf_common import (
     pack_dir_as_odf,
     pack_flat_odf,
     parse_xml_from_zip,
+    pdf_to_pngs,
+    render_to_pdf,
     replace_pattern_with_element_in_element,
     replace_text_in_element,
     sniff_image_mime,
@@ -105,4 +108,8 @@ __all__ = [
     "find_soffice",
     "find_pandoc",
     "latex_to_mathml",
+    # Rendering
+    "render_to_pdf",
+    "pdf_to_pngs",
+    "build_contact_sheet",
 ]

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/odf_lib/odf_common.py

@@ -19,7 +19,7 @@ from datetime import datetime, timezone
 from pathlib import Path
 from xml.etree import ElementTree as ET
 
-VERSION = "1.2.0"  # keep in sync with pyproject.toml (see CONTRIBUTING.md)
+VERSION = "1.4.0"  # keep in sync with pyproject.toml (see CONTRIBUTING.md)
 
 ODF_NAMESPACES: dict[str, str] = {
     "office": "urn:oasis:names:tc:opendocument:xmlns:office:1.0",
@@ -1630,3 +1630,149 @@ def local_name(tag: str) -> str:
         The local name part (e.g. ``"text"``).
     """
     return tag.split("}", 1)[1] if tag.startswith("{") else tag
+
+
+def render_to_pdf(odf_path: Path, outdir: Path) -> Path:
+    """Render an ODF file to PDF with LibreOffice.
+
+    Uses an isolated, throwaway user profile so concurrent renders do not
+    clash. The PDF is written to *outdir* as ``<stem>.pdf``.
+
+    Args:
+        odf_path: Source ODF file.
+        outdir: Directory for the PDF (created if absent).
+
+    Returns:
+        Path to the rendered PDF.
+
+    Raises:
+        SystemExit: If LibreOffice fails or the PDF is not produced.
+    """
+    import subprocess
+
+    outdir.mkdir(parents=True, exist_ok=True)
+    soffice: str = find_soffice()
+    profile: str = tempfile.mkdtemp(prefix="odf-soffice-")
+    try:
+        result = subprocess.run(
+            [
+                soffice,
+                f"-env:UserInstallation=file://{profile}",
+                "--headless",
+                "--convert-to",
+                "pdf",
+                "--outdir",
+                str(outdir),
+                str(odf_path),
+            ],
+            text=True,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.STDOUT,
+        )
+    finally:
+        shutil.rmtree(profile, ignore_errors=True)
+    pdf: Path = outdir / f"{odf_path.stem}.pdf"
+    if result.returncode != 0 or not pdf.exists():
+        raise SystemExit(f"LibreOffice failed to render {odf_path.name}:\n{result.stdout}")
+    return pdf
+
+
+def pdf_to_pngs(pdf_path: Path, outdir: Path, dpi: int = 150) -> list[Path]:
+    """Render every page of a PDF to a PNG with ``pdftoppm`` (Poppler).
+
+    Args:
+        pdf_path: Source PDF.
+        outdir: Directory for the page PNGs (created if absent).
+        dpi: Render resolution.
+
+    Returns:
+        Sorted list of ``<stem>-N.png`` page images.
+
+    Raises:
+        SystemExit: If ``pdftoppm`` is not installed or rendering fails.
+    """
+    import subprocess
+
+    pdftoppm: str | None = shutil.which("pdftoppm")
+    if pdftoppm is None:
+        raise SystemExit(
+            "pdftoppm not found — install Poppler (e.g. 'brew install poppler' or 'apt install poppler-utils')"
+        )
+    outdir.mkdir(parents=True, exist_ok=True)
+    prefix: Path = outdir / pdf_path.stem
+    result = subprocess.run(
+        [pdftoppm, "-png", "-r", str(dpi), str(pdf_path), str(prefix)],
+        text=True,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.STDOUT,
+    )
+    if result.returncode != 0:
+        raise SystemExit(f"pdftoppm failed for {pdf_path.name}:\n{result.stdout}")
+    return sorted(outdir.glob(f"{pdf_path.stem}-*.png"))
+
+
+def build_contact_sheet(images: list[Path], output_path: Path, columns: int = 0) -> Path:
+    """Compose page thumbnails into a single labelled grid image.
+
+    A contact sheet shows every page of a document at once, so layout and
+    cross-page consistency can be judged in a single glance. Requires Pillow.
+
+    Args:
+        images: Page images, in order.
+        output_path: Destination PNG.
+        columns: Grid columns; 0 picks 2 for landscape pages, 3 for portrait.
+
+    Returns:
+        Path to the contact sheet PNG.
+
+    Raises:
+        SystemExit: If Pillow is not installed or *images* is empty.
+    """
+    try:
+        from PIL import Image, ImageDraw, ImageFont
+    except ImportError:
+        raise SystemExit(
+            "Contact sheet rendering requires Pillow. Install with:\n"
+            "  pip install open-document-lib[render]\n"
+            "Or render per-page PNGs with --png and inspect them individually."
+        ) from None
+
+    if not images:
+        raise SystemExit("build_contact_sheet: no page images to compose")
+
+    thumb_w = 480
+    pad = 16
+    label_h = 22
+    bg = (245, 245, 247)
+
+    thumbs: list[Image.Image] = []
+    for path in images:
+        img = Image.open(path).convert("RGB")
+        ratio = thumb_w / img.width
+        thumbs.append(img.resize((thumb_w, max(1, round(img.height * ratio))), Image.LANCZOS))
+
+    if columns <= 0:
+        columns = 2 if thumbs[0].width >= thumbs[0].height else 3
+    columns = min(columns, len(thumbs))
+    rows = (len(thumbs) + columns - 1) // columns
+
+    cell_w = thumb_w + pad
+    cell_h = max(t.height for t in thumbs) + pad + label_h
+    sheet = Image.new("RGB", (columns * cell_w + pad, rows * cell_h + pad), bg)
+    draw = ImageDraw.Draw(sheet)
+    try:
+        font = ImageFont.truetype("/System/Library/Fonts/Supplemental/Arial.ttf", 15)
+    except OSError:
+        font = ImageFont.load_default()
+
+    for index, thumb in enumerate(thumbs):
+        col = index % columns
+        row = index // columns
+        x = pad + col * cell_w
+        y = pad + row * cell_h
+        draw.text((x, y), f"Page {index + 1}", fill=(60, 60, 70), font=font)
+        sheet.paste(thumb, (x, y + label_h))
+
+    output_path.parent.mkdir(parents=True, exist_ok=True)
+    sheet.save(output_path)
+    return output_path

{open_document_lib-1.2.0 → open_document_lib-1.4.0/open_document_lib.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: open-document-lib
-Version: 1.2.0
+Version: 1.4.0
 Summary: Standard-library toolkit for reading, editing, and writing OpenDocument Format files (ODT, ODP, ODS, ODG)
 Author: Patrick Leiverkus
 License: MIT
@@ -30,6 +30,8 @@ Provides-Extra: scholarly
 Requires-Dist: bibtexparser>=1.4; extra == "scholarly"
 Provides-Extra: validate
 Requires-Dist: lxml>=4.9; extra == "validate"
+Provides-Extra: render
+Requires-Dist: Pillow>=10; extra == "render"
 Provides-Extra: dev
 Requires-Dist: pytest>=8.0; extra == "dev"
 Requires-Dist: ruff>=0.9; extra == "dev"
@@ -37,6 +39,7 @@ Requires-Dist: pytest-cov>=5.0; extra == "dev"
 Requires-Dist: hypothesis>=6.0; extra == "dev"
 Requires-Dist: build>=1.0; extra == "dev"
 Requires-Dist: mypy>=1.8; extra == "dev"
+Requires-Dist: Pillow>=10; extra == "dev"
 Dynamic: license-file
 
 # open-document-lib
@@ -190,6 +193,14 @@ helpers) is internal and may change without notice.
 | `find_pandoc() -> str \| None` | Locate the `pandoc` binary. |
 | `latex_to_mathml(latex) -> bytes` | Convert a LaTeX snippet to MathML via Pandoc. |
 
+### Rendering
+
+| Signature | Description |
+|---|---|
+| `render_to_pdf(odf_path, outdir) -> Path` | Render an ODF file to PDF via LibreOffice (isolated profile). |
+| `pdf_to_pngs(pdf_path, outdir, dpi=150) -> list[Path]` | Render each PDF page to a PNG via `pdftoppm` (Poppler). |
+| `build_contact_sheet(images, output_path, columns=0) -> Path` | Compose page thumbnails into one labelled grid image. Requires Pillow (`pip install open-document-lib[render]`). |
+
 ## License
 
 MIT. See [LICENSE](https://github.com/leiverkus/open-document-skills/blob/main/LICENSE).

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/open_document_lib.egg-info/SOURCES.txt

@@ -24,6 +24,7 @@ tests/test_footnotes.py
 tests/test_install.py
 tests/test_lib_odf_common.py
 tests/test_libreoffice_integration.py
+tests/test_markdown_to_odt.py
 tests/test_math.py
 tests/test_meta_lifecycle.py
 tests/test_odg_connectors.py

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/open_document_lib.egg-info/requires.txt

@@ -6,10 +6,14 @@ pytest-cov>=5.0
 hypothesis>=6.0
 build>=1.0
 mypy>=1.8
+Pillow>=10
 
 [odf]
 odfpy>=1.4.0
 
+[render]
+Pillow>=10
+
 [scholarly]
 bibtexparser>=1.4
 

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "open-document-lib"
-version = "1.2.0"
+version = "1.4.0"
 description = "Standard-library toolkit for reading, editing, and writing OpenDocument Format files (ODT, ODP, ODS, ODG)"
 readme = "docs/library-api.md"
 requires-python = ">=3.10"
@@ -47,7 +47,8 @@ Issues = "https://github.com/leiverkus/open-document-skills/issues"
 odf = ["odfpy>=1.4.0"]
 scholarly = ["bibtexparser>=1.4"]
 validate = ["lxml>=4.9"]
-dev = ["pytest>=8.0", "ruff>=0.9", "pytest-cov>=5.0", "hypothesis>=6.0", "build>=1.0", "mypy>=1.8"]
+render = ["Pillow>=10"]
+dev = ["pytest>=8.0", "ruff>=0.9", "pytest-cov>=5.0", "hypothesis>=6.0", "build>=1.0", "mypy>=1.8", "Pillow>=10"]
 
 [tool.setuptools]
 packages = ["odf_lib"]

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/tests/test_lib_odf_common.py

@@ -15,6 +15,7 @@ from xml.etree import ElementTree as ET
 
 from odf_lib.odf_common import (
     VERSION,
+    build_contact_sheet,
     clear_children,
     copy_into_package,
     copy_with_multiple_members,
@@ -559,5 +560,53 @@ class LibOdfCommonTests(unittest.TestCase):
                 find_soffice()
 
 
+try:
+    from PIL import Image as _PILImage
+
+    HAVE_PILLOW = True
+except ImportError:
+    HAVE_PILLOW = False
+
+
+@unittest.skipUnless(HAVE_PILLOW, "Pillow not installed")
+class ContactSheetTests(unittest.TestCase):
+    def _make_pages(self, tmp: Path, count: int, size: tuple[int, int] = (320, 240)) -> list[Path]:
+        paths: list[Path] = []
+        for i in range(count):
+            path = tmp / f"page-{i + 1}.png"
+            _PILImage.new("RGB", size, (200, 200 + i * 10, 220)).save(path)
+            paths.append(path)
+        return paths
+
+    def test_contact_sheet_composes_a_valid_grid(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            pages = self._make_pages(tmp_path, 5)
+            out = tmp_path / "sheet.png"
+            result = build_contact_sheet(pages, out)
+            self.assertEqual(result, out)
+            self.assertTrue(out.exists())
+            with _PILImage.open(out) as sheet:
+                self.assertEqual(sheet.format, "PNG")
+                # 5 landscape pages -> 2 columns, 3 rows: sheet is wider than one cell.
+                self.assertGreater(sheet.width, 480)
+                self.assertGreater(sheet.height, 480)
+
+    def test_contact_sheet_respects_explicit_columns(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            pages = self._make_pages(tmp_path, 4)
+            wide = build_contact_sheet(pages, tmp_path / "wide.png", columns=4)
+            narrow = build_contact_sheet(pages, tmp_path / "narrow.png", columns=1)
+            with _PILImage.open(wide) as w, _PILImage.open(narrow) as n:
+                self.assertGreater(w.width, n.width)
+                self.assertGreater(n.height, w.height)
+
+    def test_contact_sheet_rejects_empty_input(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            with self.assertRaises(SystemExit):
+                build_contact_sheet([], Path(tmp) / "sheet.png")
+
+
 if __name__ == "__main__":
     unittest.main()

{open_document_lib-1.2.0 → open_document_lib-1.4.0}/tests/test_libreoffice_integration.py

@@ -88,6 +88,65 @@ class LibreOfficeIntegrationTests(unittest.TestCase):
             self.assertTrue(pdf.exists())
             self.assertGreater(pdf.stat().st_size, 0)
 
+    def test_ods_render_to_pdf(self) -> None:
+        """The new ODS render.py must produce a PDF."""
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            scripts = SKILLS / "ods" / "scripts"
+            ods = tmp_path / "book.ods"
+            outdir = tmp_path / "qa"
+            run_script(scripts / "create_minimal_ods.py", FIXTURES / "ods_workbook.json", ods)
+            run_script(scripts / "render.py", ods, "--outdir", outdir)
+            pdf = outdir / "book.pdf"
+            self.assertTrue(pdf.exists())
+            self.assertGreater(pdf.stat().st_size, 0)
+
+    def test_contact_sheet_render(self) -> None:
+        """render.py --contact-sheet must compose all pages into one PNG."""
+        import shutil
+
+        if not shutil.which("pdftoppm"):
+            self.skipTest("pdftoppm (Poppler) not available")
+        try:
+            import PIL  # noqa: F401
+        except ImportError:
+            self.skipTest("Pillow not installed")
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            scripts = SKILLS / "odp" / "scripts"
+            spec = tmp_path / "deck.json"
+            spec.write_text(
+                json.dumps({"slides": [{"name": f"S{i}", "title": f"Slide {i}"} for i in range(1, 4)]}),
+                encoding="utf-8",
+            )
+            odp = tmp_path / "deck.odp"
+            run_script(scripts / "create_minimal_odp.py", spec, odp)
+            outdir = tmp_path / "qa"
+            run_script(scripts / "render.py", odp, "--outdir", outdir, "--contact-sheet")
+            sheet = outdir / "deck-contact.png"
+            self.assertTrue(sheet.exists())
+            self.assertGreater(sheet.stat().st_size, 0)
+
+    def test_markdown_to_odt_renders_to_pdf(self) -> None:
+        """An ODT built from Markdown must render to a non-empty PDF."""
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            scripts = SKILLS / "odt" / "scripts"
+            src = tmp_path / "doc.md"
+            src.write_text(
+                "# Title\n\nText with **bold** and a [link](https://x.io).\n\n"
+                "- one\n- two\n\n| A | B |\n|---|---|\n| 1 | 2 |\n",
+                encoding="utf-8",
+            )
+            odt = tmp_path / "doc.odt"
+            run_script(scripts / "create_from_markdown.py", src, odt)
+            run_script(scripts / "validate_refs.py", odt, "--strict")
+            outdir = tmp_path / "qa"
+            run_script(scripts / "render.py", odt, "--outdir", outdir)
+            pdf = outdir / "doc.pdf"
+            self.assertTrue(pdf.exists())
+            self.assertGreater(pdf.stat().st_size, 0)
+
     def test_branded_odg_diagram_renders_to_pdf(self) -> None:
         """A base ODG with the branded diagram styles.xml injected must render
         to a non-empty PDF."""

open_document_lib-1.4.0/tests/test_markdown_to_odt.py

@@ -0,0 +1,262 @@
+"""Tests for the Markdown → ODT authoring path — the stdlib parser and the
+inline rich-text emitter (text:span, text:a, footnotes, tables, images)."""
+
+from __future__ import annotations
+
+import sys
+import tempfile
+import unittest
+import zipfile
+from pathlib import Path
+from xml.etree import ElementTree as ET
+
+from helpers import SKILLS, run_script
+
+ODT_SCRIPTS = SKILLS / "odt" / "scripts"
+sys.path.insert(0, str(ODT_SCRIPTS))
+
+import md_parser as md  # noqa: E402
+
+NS = {
+    "office": "urn:oasis:names:tc:opendocument:xmlns:office:1.0",
+    "text": "urn:oasis:names:tc:opendocument:xmlns:text:1.0",
+    "table": "urn:oasis:names:tc:opendocument:xmlns:table:1.0",
+    "draw": "urn:oasis:names:tc:opendocument:xmlns:drawing:1.0",
+    "xlink": "http://www.w3.org/1999/xlink",
+    "style": "urn:oasis:names:tc:opendocument:xmlns:style:1.0",
+}
+
+
+def q(prefix: str, local: str) -> str:
+    return f"{{{NS[prefix]}}}{local}"
+
+
+def member(path: Path, name: str) -> ET.Element:
+    with zipfile.ZipFile(path) as archive:
+        return ET.fromstring(archive.read(name))
+
+
+# --------------------------------------------------------------------------
+# Parser unit tests
+# --------------------------------------------------------------------------
+
+
+class ParserBlockTests(unittest.TestCase):
+    def test_headings_all_levels(self) -> None:
+        doc = md.parse("\n\n".join(f"{'#' * n} Level {n}" for n in range(1, 7)))
+        self.assertEqual([type(b).__name__ for b in doc.children], ["Heading"] * 6)
+        self.assertEqual([b.level for b in doc.children], [1, 2, 3, 4, 5, 6])
+
+    def test_nested_bullet_list(self) -> None:
+        doc = md.parse("- a\n- b\n  - b1\n  - b2\n- c")
+        lst = doc.children[0]
+        self.assertIsInstance(lst, md.ListNode)
+        self.assertFalse(lst.ordered)
+        self.assertEqual(len(lst.items), 3)
+        nested = [b for b in lst.items[1].children if isinstance(b, md.ListNode)]
+        self.assertEqual(len(nested), 1)
+        self.assertEqual(len(nested[0].items), 2)
+
+    def test_ordered_list_start(self) -> None:
+        doc = md.parse("3. three\n4. four")
+        lst = doc.children[0]
+        self.assertIsInstance(lst, md.ListNode)
+        self.assertTrue(lst.ordered)
+        self.assertEqual(lst.start, 3)
+
+    def test_fenced_code_block(self) -> None:
+        doc = md.parse("```python\nx = 1\ny = 2\n```")
+        block = doc.children[0]
+        self.assertIsInstance(block, md.CodeBlock)
+        self.assertEqual(block.language, "python")
+        self.assertEqual(block.text, "x = 1\ny = 2")
+
+    def test_blockquote(self) -> None:
+        doc = md.parse("> quoted line one\n> quoted line two")
+        quote = doc.children[0]
+        self.assertIsInstance(quote, md.BlockQuote)
+        self.assertIsInstance(quote.children[0], md.Paragraph)
+
+    def test_thematic_break(self) -> None:
+        doc = md.parse("text\n\n---\n\nmore")
+        self.assertIsInstance(doc.children[1], md.ThematicBreak)
+
+    def test_gfm_table_with_alignment(self) -> None:
+        doc = md.parse("| A | B | C |\n|:--|:-:|--:|\n| 1 | 2 | 3 |")
+        table = doc.children[0]
+        self.assertIsInstance(table, md.Table)
+        self.assertEqual(table.alignments, ["left", "center", "right"])
+        self.assertEqual(len(table.header), 3)
+        self.assertEqual(len(table.rows), 1)
+
+    def test_block_image(self) -> None:
+        doc = md.parse("![alt text](pic.png)")
+        image = doc.children[0]
+        self.assertIsInstance(image, md.BlockImage)
+        self.assertEqual(image.src, "pic.png")
+
+
+class ParserInlineTests(unittest.TestCase):
+    def _inline(self, text: str) -> list[md.Inline]:
+        doc = md.parse(text)
+        para = doc.children[0]
+        assert isinstance(para, md.Paragraph)
+        return para.children
+
+    def test_bold_italic_code(self) -> None:
+        nodes = self._inline("a **bold** and *italic* and `code` end")
+        kinds = [type(n).__name__ for n in nodes]
+        self.assertIn("Strong", kinds)
+        self.assertIn("Emphasis", kinds)
+        self.assertIn("Code", kinds)
+
+    def test_bold_italic_combined(self) -> None:
+        nodes = self._inline("***both***")
+        self.assertEqual(len(nodes), 1)
+        self.assertIsInstance(nodes[0], md.Strong)
+        self.assertIsInstance(nodes[0].children[0], md.Emphasis)
+
+    def test_inline_link(self) -> None:
+        nodes = self._inline("see [the docs](https://example.org) now")
+        link = next(n for n in nodes if isinstance(n, md.Link))
+        self.assertEqual(link.href, "https://example.org")
+
+    def test_reference_link(self) -> None:
+        nodes = self._inline("see [the docs][ref] now\n\n[ref]: https://ref.example")
+        link = next(n for n in nodes if isinstance(n, md.Link))
+        self.assertEqual(link.href, "https://ref.example")
+
+    def test_footnote_ref_and_def(self) -> None:
+        doc = md.parse("text with a note.[^x]\n\n[^x]: the note body")
+        para = doc.children[0]
+        assert isinstance(para, md.Paragraph)
+        self.assertTrue(any(isinstance(n, md.FootnoteRef) for n in para.children))
+        self.assertIn("x", doc.footnotes)
+
+    def test_underscore_in_word_is_literal(self) -> None:
+        nodes = self._inline("a snake_case_name here")
+        self.assertEqual(len(nodes), 1)
+        self.assertIsInstance(nodes[0], md.Text)
+
+    def test_backslash_escape(self) -> None:
+        nodes = self._inline(r"not \*emphasised\* here")
+        self.assertEqual(len(nodes), 1)
+        self.assertIsInstance(nodes[0], md.Text)
+        self.assertEqual(nodes[0].value, "not *emphasised* here")
+
+
+# --------------------------------------------------------------------------
+# End-to-end MD → ODT
+# --------------------------------------------------------------------------
+
+SAMPLE_MD = """# Report Title
+
+A paragraph with **bold**, *italic*, `code`, and a [link](https://example.org).
+A footnote follows.[^a]
+
+## Section Two
+
+- one
+- two
+  - nested
+
+1. first
+2. second
+
+> a quoted line
+
+```python
+print("hi")
+```
+
+| Name | Score |
+|:-----|------:|
+| Alice | 42 |
+
+---
+
+Done.
+
+[^a]: The footnote text.
+"""
+
+
+class EndToEndTests(unittest.TestCase):
+    def _build(self, tmp: Path, markdown: str = SAMPLE_MD) -> Path:
+        src = tmp / "doc.md"
+        src.write_text(markdown, encoding="utf-8")
+        odt = tmp / "doc.odt"
+        run_script(ODT_SCRIPTS / "create_from_markdown.py", src, odt)
+        return odt
+
+    def test_inline_runs_become_spans(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            content = member(self._build(Path(tmp)), "content.xml")
+            spans = list(content.iter(q("text", "span")))
+            styles = {s.attrib.get(q("text", "style-name")) for s in spans}
+            self.assertIn("Strong", styles)
+            self.assertIn("Emphasis", styles)
+            self.assertIn("Code", styles)
+
+    def test_link_becomes_text_a(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            content = member(self._build(Path(tmp)), "content.xml")
+            anchor = next(content.iter(q("text", "a")))
+            self.assertEqual(anchor.attrib.get(q("xlink", "href")), "https://example.org")
+
+    def test_headings_carry_outline_levels(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            content = member(self._build(Path(tmp)), "content.xml")
+            levels = sorted({h.attrib.get(q("text", "outline-level")) for h in content.iter(q("text", "h"))})
+            self.assertEqual(levels, ["1", "2"])
+
+    def test_lists_and_table_and_footnote(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            content = member(self._build(Path(tmp)), "content.xml")
+            self.assertGreaterEqual(len(list(content.iter(q("text", "list")))), 2)
+            self.assertEqual(len(list(content.iter(q("table", "table")))), 1)
+            notes = list(content.iter(q("text", "note")))
+            self.assertEqual(len(notes), 1)
+            self.assertTrue(notes[0].attrib.get(q("text", "id")))
+
+    def test_code_block_uses_codeblock_style(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            content = member(self._build(Path(tmp)), "content.xml")
+            code_paras = [
+                p for p in content.iter(q("text", "p")) if p.attrib.get(q("text", "style-name")) == "CodeBlock"
+            ]
+            self.assertGreater(len(code_paras), 0)
+
+    def test_title_from_first_heading(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            meta = member(self._build(Path(tmp)), "meta.xml")
+            titles = [t.text for t in meta.iter("{http://purl.org/dc/elements/1.1/}title")]
+            self.assertEqual(titles, ["Report Title"])
+
+    def test_embedded_image(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            tmp_path = Path(tmp)
+            png = tmp_path / "p.png"
+            png.write_bytes(
+                bytes.fromhex(
+                    "89504e470d0a1a0a0000000d49484452000000040000000408060000"
+                    "00b6f8b4570000000a49444154789c63600000000200013fd7e2d6"
+                    "0000000049454e44ae426082"
+                )
+            )
+            odt = self._build(tmp_path, f"# Pic\n\n![logo]({png})\n")
+            with zipfile.ZipFile(odt) as archive:
+                pictures = [n for n in archive.namelist() if n.startswith("Pictures/")]
+            self.assertEqual(len(pictures), 1)
+            content = member(odt, "content.xml")
+            self.assertEqual(len(list(content.iter(q("draw", "image")))), 1)
+
+    def test_output_passes_strict_validation(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            odt = self._build(Path(tmp))
+            result = run_script(ODT_SCRIPTS / "validate_refs.py", odt, "--strict", check=False)
+            self.assertIn('"status": "ok"', result.stdout)
+
+
+if __name__ == "__main__":
+    unittest.main()