anynews/republisher
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Prefer content over description for item bodies

Browse source
This commit is contained in:
Abel Luck 2026-04-01 17:27:20 +02:00
parent 05ac6ce20d
commit cebf037753
2 changed files with 67 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

12
repub/spiders/rss_spider.py
View file

@ -281,6 +281,14 @@ class RssFeedSpider(BaseRssFeedSpider):
file_urls = []
audio_urls = []
video_urls = []
source_description_html = (
sanitize_html(entry.get("summary", "")) if "summary_detail" in entry else ""
)
has_content_html = any(
c.type == "text/html" and ((getattr(c, "value", "") or "").strip() != "")
for c in entry.get("content", [])
)
description_html = source_description_html if has_content_html else ""
def add_url(file_type, url):
if file_type == FileType.IMAGE:
@ -295,7 +303,7 @@ class RssFeedSpider(BaseRssFeedSpider):
item = E.item(
E.title(entry.get("title")),
E.link(entry.get("link")),
E.description(sanitize_html(entry.get("description", ""))),
E.description(description_html),
E.guid(
entry.get("id"),
{"isPermaLink": "true" if entry.guidislink else "false"},
@ -341,6 +349,8 @@ class RssFeedSpider(BaseRssFeedSpider):
image_urls.extend(urls[FileType.IMAGE])
video_urls.extend(urls[FileType.VIDEO])
audio_urls.extend(urls[FileType.AUDIO])
if not has_content_html and source_description_html.strip() != "":
item.append(CONTENT.encoded(CDATA(source_description_html)))
if isinstance(entry.get("media_content"), list):
for media in (

62
tests/test_feed_validation.py
View file

@ -437,10 +437,60 @@ def test_feed_generation_normalizes_dates_urls_and_xml_shapes() -> None:
assert "<" not in itunes_summary
assert ">" not in itunes_summary
assert "contenteditable=" not in xml
assert "mode=" not in xml
assert "querystring=" not in xml
assert (
f"https://mirror.example/feeds/demo/images/{local_image_path(source_image)}"
in xml
def test_item_body_uses_description_only_when_content_is_also_present() -> None:
xml, root = _serialize_feed(
feed_url="https://mirror.example",
feed_text="""<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
<title>Demo Feed</title>
<link>https://source.example/feed</link>
<description>Demo description</description>
<item>
<title>Description Only</title>
<link>https://source.example/description-only</link>
<description><![CDATA[<p mode="summary">Description body</p>]]></description>
<guid isPermaLink="false">entry-description-only</guid>
<pubDate>Tue, 31 Mar 2026 10:31:50 +0000</pubDate>
</item>
<item>
<title>Content Only</title>
<link>https://source.example/content-only</link>
<guid isPermaLink="false">entry-content-only</guid>
<pubDate>Tue, 31 Mar 2026 10:31:50 +0000</pubDate>
<content:encoded><![CDATA[<div mode="body">Content body</div>]]></content:encoded>
</item>
<item>
<title>Both Present</title>
<link>https://source.example/both-present</link>
<description><![CDATA[<p mode="summary">Summary body</p>]]></description>
<guid isPermaLink="false">entry-both-present</guid>
<pubDate>Tue, 31 Mar 2026 10:31:50 +0000</pubDate>
<content:encoded><![CDATA[<div mode="body">Full body</div>]]></content:encoded>
</item>
</channel>
</rss>
""",
)
items = root.findall("./channel/item")
assert len(items) == 3
description_only, content_only, both_present = items
assert description_only.findtext("description") in (None, "")
assert description_only.findtext("content:encoded", namespaces=nsmap) == (
"<p>Description body</p>"
)
assert content_only.findtext("description") in (None, "")
assert content_only.findtext("content:encoded", namespaces=nsmap) == (
"<div>Content body</div>"
)
assert both_present.findtext("description") == "<p>Summary body</p>"
assert both_present.findtext("content:encoded", namespaces=nsmap) == (
"<div>Full body</div>"
)