Bilder standardmäßig als JPEG (--image-format), Output/ ignorieren

EPUB Band 1: 12,9 MB -> 5,0 MB, Band 2: 14,3 MB -> 5,3 MB.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

.gitignore

@@ -14,3 +14,6 @@ __pycache__/
 
 # Leere Datei durch System-Bug
 /3
+
+# Ergebnisse zum Ansehen
+Output/

pdf2md_v5.py

@@ -805,7 +805,7 @@ def convert_pages(source: str, run: str, pages: list[int], options: dict) -> str
         footer=options["keep_footers"],
         write_images=options["images"],
         image_path=options["image_path"],
-        image_format="png",
+        image_format=options["image_format"],
     )
     if run == RUN_OCR:
         kwargs.update(
@@ -984,6 +984,12 @@ def main() -> int:
         action="store_true",
         help="Grafiken speichern und in Markdown verlinken",
     )
+    parser.add_argument(
+        "--image-format",
+        choices=["jpg", "png"],
+        default="jpg",
+        help="Bildformat: jpg (klein, Standard) oder png (verlustfrei, groß)",
+    )
     parser.add_argument(
         "--epub",
         action="store_true",
@@ -1057,7 +1063,7 @@ def main() -> int:
             if args.images:
                 image_dir.mkdir(parents=True, exist_ok=True)
                 if epub_path:
-                    cover = image_dir / "cover.png"
+                    cover = image_dir / f"cover.{args.image_format}"
                     doc[0].get_pixmap(dpi=150).save(cover)
 
         options = dict(
@@ -1065,6 +1071,7 @@ def main() -> int:
             keep_footers=args.keep_footers,
             images=args.images,
             image_path=str(image_dir.resolve()) if args.images else "",
+            image_format=args.image_format,
             ocr_language=args.ocr_language,
             ocr_dpi=args.ocr_dpi,
         )

tests/test_e2e.py

@@ -110,7 +110,9 @@ def test_epub(pdfs, tmp_path):
     out = tmp_path / "buch.md"
     run(pdfs["digital"], out, "--epub")
     epub = tmp_path / "buch.epub"
-    assert (tmp_path / "buch_images" / "cover.png").is_file()
+    images = tmp_path / "buch_images"
+    assert (images / "cover.jpg").is_file()
+    assert {f.suffix for f in images.iterdir()} == {".jpg"}
     with zipfile.ZipFile(epub) as z:
         assert z.read("mimetype") == b"application/epub+zip"
         nav = z.read("EPUB/nav.xhtml").decode()
@@ -119,3 +121,10 @@ def test_epub(pdfs, tmp_path):
         assert title in nav
     assert "Erika Muster" in opf
     assert "cover" in opf
+
+
+def test_png_on_request(pdfs, tmp_path):
+    out = tmp_path / "png.md"
+    run(pdfs["digital"], out, "--images", "--image-format", "png")
+    md = out.read_text(encoding="utf-8")
+    assert re.search(r"!\[\]\(png_images/[^)]+\.png\)", md)