Improve PPTX converter, fix lint/type errors, and harden release workflow

.github/workflows/release.yml

@@ -30,4 +30,10 @@ jobs:
       - name: Create GitHub Release
         env:
           GH_TOKEN: ${{ github.token }}
-        run: gh release create ${{ github.ref_name }} dist/* --generate-notes
+          TAG_NAME: ${{ github.ref_name }}
+        run: |
+          if gh release view "$TAG_NAME" > /dev/null 2>&1; then
+            gh release upload "$TAG_NAME" dist/* --clobber
+          else
+            gh release create "$TAG_NAME" dist/* --generate-notes
+          fi

Makefile

@@ -11,7 +11,7 @@ format:
 	uv run ruff format src/
 
 typecheck:
-	-uv run ty check
+	uv run ty check
 
 setup-hooks:
 	git config core.hooksPath .githooks

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "uv_build"
 
 [project]
 name = "leafpress"
-version = "0.8.0"
+version = "0.8.1"
 description = "Convert MkDocs sites to PDF, Word, HTML, ODT, EPUB, and Markdown documents with branding"
 readme = "README.md"
 authors = [{name = "Shane Hutchins", email = "hutchins@users.noreply.github.com"}]

src/leafpress/importer/converter_pptx.py

@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-import re
+from collections import Counter
 from pathlib import Path
 
 from pptx import Presentation
@@ -71,12 +71,15 @@ def import_pptx(
         except Exception as e:
             raise PptxImportError(f"Failed to open PPTX: {e}") from e
 
-        warnings: list[str] = []
+        skipped_counts: Counter[str] = Counter()
         sections: list[str] = []
         for slide_num, slide in enumerate(prs.slides, start=1):
-            slide_md = _convert_slide(slide, slide_num, image_handler, include_notes, warnings)
+            slide_md = _convert_slide(
+                slide, slide_num, image_handler, include_notes, skipped_counts
+            )
             sections.append(slide_md)
 
+    warnings = _aggregate_warnings(skipped_counts)
     markdown = "\n\n".join(sections)
     markdown = postprocess_markdown(markdown)
 
@@ -91,12 +94,21 @@ def import_pptx(
     )
 
 
+def _aggregate_warnings(skipped_counts: Counter[str]) -> list[str]:
+    """Build aggregated warning messages from per-type skip counts."""
+    warnings: list[str] = []
+    for label, count in sorted(skipped_counts.items()):
+        plural = "s" if count > 1 else ""
+        warnings.append(f"Presentation has {count} unsupported {label}{plural} — skipped")
+    return warnings
+
+
 def _convert_slide(
     slide,
     slide_num: int,
     image_handler: ImageHandler | None,
     include_notes: bool,
-    warnings: list[str],
+    skipped_counts: Counter[str],
 ) -> str:
     """Convert a single slide to markdown."""
     parts: list[str] = []
@@ -115,7 +127,7 @@ def _convert_slide(
     for shape in slide.shapes:
         if shape is title_shape:
             continue
-        shape_md = _convert_shape(shape, image_handler, slide_label, warnings)
+        shape_md = _convert_shape(shape, image_handler, slide_label, skipped_counts)
         if shape_md:
             parts.append(shape_md)
 
@@ -140,11 +152,11 @@ def _convert_shape(
     shape,
     image_handler: ImageHandler | None,
     slide_label: str,
-    warnings: list[str],
+    skipped_counts: Counter[str],
 ) -> str:
     """Convert a single shape to markdown."""
     if shape.shape_type == MSO_SHAPE_TYPE.GROUP:
-        return _convert_group(shape, image_handler, slide_label, warnings)
+        return _convert_group(shape, image_handler, slide_label, skipped_counts)
 
     if shape.shape_type == MSO_SHAPE_TYPE.TABLE:
         return _table_to_markdown(shape.table)
@@ -156,12 +168,11 @@ def _convert_shape(
         return _text_frame_to_markdown(shape.text_frame)
 
     # Shapes with no text, table, image, or group content are skipped.
-    # Warn for shape types that may contain meaningful content.
+    # Count per type for aggregated warnings.
     shape_type = shape.shape_type
     if shape_type in _WARN_SHAPE_TYPES:
         label = _WARN_SHAPE_TYPES[shape_type]
-        name = shape.name or "unnamed"
-        warnings.append(f"Unsupported {label} '{name}' on slide '{slide_label}' — skipped")
+        skipped_counts[label] += 1
 
     return ""
 
@@ -170,12 +181,12 @@ def _convert_group(
     group_shape,
     image_handler: ImageHandler | None,
     slide_label: str,
-    warnings: list[str],
+    skipped_counts: Counter[str],
 ) -> str:
     """Recursively convert shapes inside a group."""
     parts: list[str] = []
     for shape in group_shape.shapes:
-        md = _convert_shape(shape, image_handler, slide_label, warnings)
+        md = _convert_shape(shape, image_handler, slide_label, skipped_counts)
         if md:
             parts.append(md)
     return "\n\n".join(parts)
@@ -187,7 +198,8 @@ def _convert_image(shape, image_handler: ImageHandler) -> str:
     image_bytes = image.blob
     content_type = image.content_type
     src = image_handler.save_image(image_bytes, content_type)
-    return f"![]({src})"
+    alt = shape.name or ""
+    return f"![{alt}]({src})"
 
 
 def _text_frame_to_markdown(text_frame) -> str:
@@ -214,15 +226,14 @@ def _runs_to_markdown(paragraph) -> str:
         if not text:
             continue
 
-        # Apply formatting
+        # Apply hyperlink first, then wrap with formatting so
+        # bold/italic are preserved: **[text](url)** instead of lost.
+        if run.hyperlink and run.hyperlink.address:
+            text = f"[{text}]({run.hyperlink.address})"
         if run.font.bold:
             text = f"**{text}**"
         if run.font.italic:
             text = f"*{text}*"
-        if run.hyperlink and run.hyperlink.address:
-            # Strip formatting wrappers to put inside link text
-            display = re.sub(r"^\*{1,2}(.*?)\*{1,2}$", r"\1", text)
-            text = f"[{display}]({run.hyperlink.address})"
 
         parts.append(text)
     return "".join(parts)

src/leafpress/source.py

@@ -60,7 +60,7 @@ def resolve_source(
 def _clone_repo(url: str, branch: str | None) -> Path:
     """Clone a git repo to a temporary directory."""
     tmp_dir = Path(tempfile.mkdtemp(prefix="leafpress_"))
-    clone_kwargs: dict[str, object] = {"depth": 50}
+    clone_kwargs: dict[str, str | int] = {"depth": 50}
     if branch:
         clone_kwargs["branch"] = branch
 

tests/test_cibutler_integration.py

@@ -15,8 +15,8 @@
 from leafpress.mkdocs_parser import MkDocsConfig, flatten_nav, parse_mkdocs_config
 from leafpress.pdf.renderer import PdfRenderer
 
-CIBUTLER_DOCS = Path("/Users/hutchins/projects/osdu/cibutler/docs")
-CIBUTLER_REPO = Path("/Users/hutchins/projects/osdu/cibutler")
+CIBUTLER_DOCS = Path("/Users/hutchins/projects/cibutler/docs")
+CIBUTLER_REPO = Path("/Users/hutchins/projects/cibutler")
 
 pytestmark = pytest.mark.skipif(
     not CIBUTLER_DOCS.exists(),

tests/test_cli.py

@@ -3,10 +3,11 @@
 from pathlib import Path
 
 from docx import Document as DocxDocument
-from leafpress.cli import cli
 from pptx import Presentation
 from typer.testing import CliRunner
 
+from leafpress.cli import cli
+
 runner = CliRunner()
 
 

tests/test_docx_renderer.py

@@ -3,6 +3,7 @@
 from pathlib import Path
 
 import pytest
+
 from leafpress.config import WatermarkConfig, load_config
 from leafpress.docx.renderer import DocxRenderer
 from leafpress.git_info import extract_git_info

tests/test_docx_styles.py

@@ -7,10 +7,14 @@
 from leafpress.docx.styles import _parse_hex_color, apply_branding_styles
 
 
-def _make_branding(**kwargs: object) -> BrandingConfig:
-    defaults = {"company_name": "TestCo", "project_name": "TestProject"}
-    defaults.update(kwargs)
-    return BrandingConfig(**defaults)
+def _make_branding(
+    primary_color: str = "#1a73e8",
+) -> BrandingConfig:
+    return BrandingConfig(
+        company_name="TestCo",
+        project_name="TestProject",
+        primary_color=primary_color,
+    )
 
 
 def test_apply_branding_styles() -> None:

tests/test_html_styles.py

@@ -4,10 +4,16 @@
 from leafpress.html.styles import generate_html_css
 
 
-def _make_branding(**kwargs: object) -> BrandingConfig:
-    defaults = {"company_name": "TestCo", "project_name": "TestProject"}
-    defaults.update(kwargs)
-    return BrandingConfig(**defaults)
+def _make_branding(
+    primary_color: str = "#1a73e8",
+    accent_color: str = "#ffffff",
+) -> BrandingConfig:
+    return BrandingConfig(
+        company_name="TestCo",
+        project_name="TestProject",
+        primary_color=primary_color,
+        accent_color=accent_color,
+    )
 
 
 def test_generate_css_with_branding() -> None:

tests/test_import_pptx.py

@@ -210,11 +210,12 @@ def test_table_extraction(tmp_path: Path) -> None:
 
 
 def test_image_extraction(tmp_path: Path) -> None:
-    """Embedded images are extracted to assets/ and referenced in markdown."""
+    """Embedded images are extracted to assets/ with alt text from shape name."""
     pptx_path = _make_image_pptx(tmp_path)
     result = import_pptx(pptx_path)
     md = result.markdown_path.read_text()
-    assert "![](assets/" in md
+    assert "![" in md
+    assert "](assets/" in md
     assert len(result.images) == 1
     assert result.images[0].exists()
 
@@ -329,8 +330,9 @@ def test_runs_to_markdown_plain() -> None:
     assert _runs_to_markdown(para) == "hello world"
 
 
-def test_convert_shape_warns_on_chart() -> None:
-    """Chart shapes produce a warning."""
+def test_convert_shape_counts_chart() -> None:
+    """Chart shapes increment the skipped counter."""
+    from collections import Counter
     from unittest.mock import MagicMock
 
     from leafpress.importer.converter_pptx import _convert_shape
@@ -339,30 +341,27 @@ def test_convert_shape_warns_on_chart() -> None:
     shape.shape_type = MSO_SHAPE_TYPE.CHART
     shape.name = "Chart 1"
     shape.has_text_frame = False
-    warnings: list[str] = []
+    skipped_counts: Counter[str] = Counter()
 
-    result = _convert_shape(shape, None, "Revenue Slide", warnings)
+    result = _convert_shape(shape, None, "Revenue Slide", skipped_counts)
     assert result == ""
-    assert len(warnings) == 1
-    assert "chart" in warnings[0].lower()
-    assert "Chart 1" in warnings[0]
-    assert "Revenue Slide" in warnings[0]
-    assert "skipped" in warnings[0].lower()
+    assert skipped_counts["chart"] == 1
 
 
-def test_convert_shape_no_warn_on_freeform() -> None:
+def test_convert_shape_no_count_on_freeform() -> None:
     """Freeform shapes (decorative lines, etc.) are silently skipped."""
+    from collections import Counter
     from unittest.mock import MagicMock
 
     from leafpress.importer.converter_pptx import _convert_shape
 
     shape = MagicMock()
     shape.shape_type = MSO_SHAPE_TYPE.FREEFORM
     shape.has_text_frame = False
-    warnings: list[str] = []
+    skipped_counts: Counter[str] = Counter()
 
-    _convert_shape(shape, None, "Slide 1", warnings)
-    assert len(warnings) == 0
+    _convert_shape(shape, None, "Slide 1", skipped_counts)
+    assert len(skipped_counts) == 0
 
 
 def test_runs_to_markdown_hyperlink() -> None:
@@ -382,7 +381,7 @@ def test_runs_to_markdown_hyperlink() -> None:
 
 
 def test_runs_to_markdown_bold_hyperlink() -> None:
-    """Bold text with a hyperlink unwraps formatting for the link text."""
+    """Bold text with a hyperlink preserves both formatting and link."""
     from unittest.mock import MagicMock
 
     para = MagicMock()
@@ -394,7 +393,56 @@ def test_runs_to_markdown_bold_hyperlink() -> None:
     para.runs = [run]
 
     md = _runs_to_markdown(para)
-    assert "[bold link](https://example.com)" in md
+    assert "**[bold link](https://example.com)**" in md
+
+
+def test_runs_to_markdown_italic_hyperlink() -> None:
+    """Italic text with a hyperlink preserves both formatting and link."""
+    from unittest.mock import MagicMock
+
+    para = MagicMock()
+    run = MagicMock()
+    run.text = "italic link"
+    run.font.bold = False
+    run.font.italic = True
+    run.hyperlink.address = "https://example.com"
+    para.runs = [run]
+
+    md = _runs_to_markdown(para)
+    assert "*[italic link](https://example.com)*" in md
+
+
+def test_aggregate_warnings_multiple_charts() -> None:
+    """Multiple charts across slides produce a single aggregated warning."""
+    from collections import Counter
+    from unittest.mock import MagicMock
+
+    from leafpress.importer.converter_pptx import _aggregate_warnings, _convert_shape
+
+    skipped_counts: Counter[str] = Counter()
+    for i in range(3):
+        shape = MagicMock()
+        shape.shape_type = MSO_SHAPE_TYPE.CHART
+        shape.name = f"Chart {i}"
+        shape.has_text_frame = False
+        _convert_shape(shape, None, f"Slide {i}", skipped_counts)
+
+    warnings = _aggregate_warnings(skipped_counts)
+    assert len(warnings) == 1
+    assert "3 unsupported charts" in warnings[0]
+
+
+def test_aggregate_warnings_mixed_types() -> None:
+    """Different unsupported types each get their own aggregated warning."""
+    from collections import Counter
+
+    from leafpress.importer.converter_pptx import _aggregate_warnings
+
+    skipped_counts: Counter[str] = Counter({"chart": 2, "media": 1})
+    warnings = _aggregate_warnings(skipped_counts)
+    assert len(warnings) == 2
+    assert any("2 unsupported charts" in w for w in warnings)
+    assert any("1 unsupported media" in w for w in warnings)
 
 
 def test_runs_to_markdown_empty_run() -> None:
@@ -587,7 +635,7 @@ def test_image_extracted(self) -> None:
         """Embedded image extracted to assets/."""
         assert len(self.result.images) >= 1
         assert self.result.images[0].exists()
-        assert "![](assets/" in self.content
+        assert "](assets/" in self.content
 
     def test_slide_count(self) -> None:
         """All 5 slides appear in output."""

tests/test_markdown_renderer.py

@@ -56,7 +56,7 @@ def test_extension_load_failure_includes_error_message(sample_mkdocs_dir: Path)
     )
     failed = [(ext, ok, msg) for ext, ok, msg in renderer.extension_load_results if not ok]
     assert len(failed) >= 1
-    ext, ok, msg = failed[0]
+    ext, _ok, msg = failed[0]
     assert ext == "nonexistent_extension_xyz"
     assert msg != ""