Python tool for converting files and office documents to Markdown.
fix(rss): support namespace-prefixed Atom feeds (#2429)
Xing Zhengcb785cb2f721
diff --git a/packages/markitdown/src/markitdown/converters/_rss_converter.py b/packages/markitdown/src/markitdown/converters/_rss_converter.pyindex 6b7b1201..0dff4992 100644--- a/packages/markitdown/src/markitdown/converters/_rss_converter.py+++ b/packages/markitdown/src/markitdown/converters/_rss_converter.py@@ -29,6 +29,7 @@ CANDIDATE_FILE_EXTENSIONS = [".xml",]+ATOM_NAMESPACE = "http://www.w3.org/2005/Atom"XHTML_NAMESPACE = "http://www.w3.org/1999/xhtml"@@ -109,9 +110,9 @@ class RssConverter(DocumentConverter):def _feed_type(self, doc: Any) -> str | None:if doc.getElementsByTagName("rss"):return "rss"- elif doc.getElementsByTagName("feed"):- root = doc.getElementsByTagName("feed")[0]- if root.getElementsByTagName("entry"):+ root = doc.documentElement+ if root.localName == "feed" and root.namespaceURI in (None, ATOM_NAMESPACE):+ if root.getElementsByTagNameNS(root.namespaceURI, "entry"):# An Atom feed must have a root element of <feed> and at least one <entry>return "atom"return None@@ -143,10 +144,10 @@ class RssConverter(DocumentConverter):Returns None if the feed type is not recognized or something goes wrong."""- root = doc.getElementsByTagName("feed")[0]+ root = doc.documentElementtitle = self._get_data_by_tag_name(root, "title")subtitle = self._get_data_by_tag_name(root, "subtitle")- entries = root.getElementsByTagName("entry")+ entries = root.getElementsByTagNameNS(root.namespaceURI, "entry")md_text = f"# {title}\n"if subtitle:md_text += f"{subtitle}\n"@@ -182,7 +183,7 @@ class RssConverter(DocumentConverter):Values flagged as markup are converted by ``_parse_content``; plain textis returned verbatim for the caller to emit as-is."""- nodes = entry.getElementsByTagName(tag_name)+ nodes = entry.getElementsByTagNameNS(entry.namespaceURI, tag_name)if not nodes:return None, True@@ -311,7 +312,10 @@ class RssConverter(DocumentConverter):"""Get data from first child element with the given tag name.Returns None when no such element is found."""- nodes = element.getElementsByTagName(tag_name)+ if element.namespaceURI == ATOM_NAMESPACE:+ nodes = element.getElementsByTagNameNS(ATOM_NAMESPACE, tag_name)+ else:+ nodes = element.getElementsByTagName(tag_name)if not nodes:return Nonefc = nodes[0].firstChilddiff --git a/packages/markitdown/tests/test_rss_converter.py b/packages/markitdown/tests/test_rss_converter.pyindex 19669cc7..ddc690e4 100644--- a/packages/markitdown/tests/test_rss_converter.py+++ b/packages/markitdown/tests/test_rss_converter.py@@ -1,9 +1,99 @@import io-from markitdown import StreamInfo+import pytest++from markitdown import MarkItDown, StreamInfofrom markitdown.converters import RssConverter+ "root_prefix, child_prefix",+ [("", ""), ("a:", "a:"), ("a:", "b:"), ("a:", ""), ("", "a:")],+)+def test_atom_namespace_prefixes(root_prefix: str, child_prefix: str) -> None:+ # Fully prefixed feeds do not need a default namespace declaration.+ default_namespace = (+ "" if root_prefix and child_prefix else "http://www.w3.org/2005/Atom"+ )+ feed = f"""<?xml version="1.0" encoding="utf-8"?>+<{root_prefix}feed xmlns="{default_namespace}"+ xmlns:a="http://www.w3.org/2005/Atom"+ xmlns:b="http://www.w3.org/2005/Atom">+ <{child_prefix}title>Example feed</{child_prefix}title>+ <{child_prefix}subtitle>Feed subtitle</{child_prefix}subtitle>+ <{child_prefix}id>urn:example:feed</{child_prefix}id>+ <{child_prefix}updated>2026-01-01T00:00:00Z</{child_prefix}updated>+ <{child_prefix}author><{child_prefix}name>Example author</{child_prefix}name></{child_prefix}author>+ <{child_prefix}entry>+ <{child_prefix}title>Example entry</{child_prefix}title>+ <{child_prefix}id>urn:example:entry</{child_prefix}id>+ <{child_prefix}updated>2026-01-02T00:00:00Z</{child_prefix}updated>+ <{child_prefix}summary type="html"><p>An <em>entry summary</em>.</p></{child_prefix}summary>+ <{child_prefix}content type="xhtml">+ <div xmlns="http://www.w3.org/1999/xhtml">+ <p>Read the <strong>important details</strong>.</p>+ </div>+ </{child_prefix}content>+ </{child_prefix}entry>+</{root_prefix}feed>+""".encode(+ "utf-8"+ )+ stream_info = StreamInfo(mimetype="application/xml", extension=".xml")+ converter = RssConverter()+ stream = io.BytesIO(feed)++ assert converter.accepts(stream, stream_info)+ assert stream.tell() == 0+ result = converter.convert(stream, stream_info)++ assert result.title == "Example feed"+ assert result.markdown.startswith("# Example feed\nFeed subtitle\n")+ assert "## Example entry\nUpdated on: 2026-01-02T00:00:00Z" in result.markdown+ assert "An *entry summary*." in result.markdown+ assert "Read the **important details**." in result.markdown++ # The public API must recognize the feed instead of falling back to raw XML.+ converted = MarkItDown().convert_stream(io.BytesIO(feed), stream_info=stream_info)+ assert converted.title == result.title+ assert converted.markdown == result.markdown.strip()+++def test_atom_without_namespace_is_still_supported() -> None:+ feed = b"""<feed>+ <title>Example feed</title>+ <entry><title>Example entry</title><content>Entry content.</content></entry>+</feed>"""+ converter = RssConverter()+ stream_info = StreamInfo(extension=".xml")++ assert converter.accepts(io.BytesIO(feed), stream_info)+ result = converter.convert(io.BytesIO(feed), stream_info)++ assert result.title == "Example feed"+ assert "## Example entry" in result.markdown+ assert "Entry content." in result.markdown+++def test_atom_ignores_elements_from_other_namespaces() -> None:+ feed = b"""<a:feed xmlns:a="http://www.w3.org/2005/Atom" xmlns="urn:example:other">+ <title>Unrelated title</title>+ <a:title>Example feed</a:title>+ <entry><title>Unrelated entry</title><content>Unrelated content.</content></entry>+ <a:entry>+ <a:title>Example entry</a:title>+ <content>Unrelated content.</content>+ <a:content>Entry content.</a:content>+ </a:entry>+</a:feed>"""+ result = RssConverter().convert(io.BytesIO(feed), StreamInfo(extension=".atom"))++ assert result.title == "Example feed"+ assert "## Example entry" in result.markdown+ assert "Entry content." in result.markdown+ assert "Unrelated" not in result.markdown++def test_atom_xhtml_content_is_preserved() -> None:feed = b"""<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom">