fix(html): reduce inline images to alt text under markdownify 1.x (convert_as_inline → parent_tags)

packages/markitdown/src/markitdown/converters/_markdownify.py

@@ -86,19 +86,24 @@ def convert_img(
         self,
         el: Any,
         text: str,
-        convert_as_inline: Optional[bool] = False,
+        parent_tags: Optional[Any] = None,
         **kwargs,
     ) -> str:
         """Same as usual converter, but removes data URIs"""
 
+        # markdownify >= 1.0 passes `parent_tags` (containing the pseudo-tag
+        # "_inline" for headings/table cells) instead of the old
+        # `convert_as_inline` boolean; check that so inline images still reduce
+        # to their alt text.
+        parent_tags = parent_tags or set()
         alt = el.attrs.get("alt", None) or ""
         src = el.attrs.get("src", None) or el.attrs.get("data-src", None) or ""
         title = el.attrs.get("title", None) or ""
         title_part = ' "%s"' % title.replace('"', r"\"") if title else ""
         # Remove all line breaks from alt
         alt = alt.replace("\n", " ")
         if (
-            convert_as_inline
+            "_inline" in parent_tags
             and el.parent.name not in self.options["keep_inline_images_in"]
         ):
             return alt

packages/markitdown/tests/test_module_misc.py

@@ -552,3 +552,29 @@ def test_markitdown_llm() -> None:
         test()
         print("OK")
     print("All tests passed!")
+
+
+def test_inline_image_in_heading_reduced_to_alt() -> None:
+    from markitdown.converters._html_converter import HtmlConverter
+
+    md = HtmlConverter().convert_string('<h1>Hello <img src="a.png" alt="World"></h1>').markdown
+    assert md.strip() == "# Hello World", repr(md)
+
+
+def test_inline_image_in_table_cell_reduced_to_alt() -> None:
+    from markitdown.converters._html_converter import HtmlConverter
+
+    html = (
+        "<table><tr><th>Pic</th><th>Name</th></tr>"
+        '<tr><td><img src="a.png" alt="AA"></td><td>bob</td></tr></table>'
+    )
+    md = HtmlConverter().convert_string(html).markdown
+    assert "| AA | bob |" in md, repr(md)
+    assert "![AA]" not in md, repr(md)
+
+
+def test_block_image_still_rendered() -> None:
+    from markitdown.converters._html_converter import HtmlConverter
+
+    md = HtmlConverter().convert_string('<p><img src="a.png" alt="x"></p>').markdown
+    assert "![x](a.png)" in md, repr(md)