Euronews Articles Scraper

Fetches Euronews's most recent articles -- the full article, not just a summary. Two-stage: discovers the newest article URLs from Euronews's public sitemap, then reads each article's own embedded structured data for the actual content.

What you get

One SEARCH_SUMMARY row plus one ARTICLE row per article, most recently modified first:

• articleHeadline, articleDescription, articleBodyText (the full article, not a preview)
• articleDatePublished / articleDateModified (ISO 8601)
• articleAuthorNames, articleSections, articleImageUrl, articleUrl

Why this actor fetches one page per article

Unlike this NEWS family's other feed-based actors, Euronews's sitemap carries only a URL and a timestamp -- no title, no body. There is no single feed anywhere on the site that returns full articles in bulk; the content only exists on each article's own page (in a NewsArticle JSON-LD block). maxItems therefore fetches that many real pages, so it directly controls run time and is the actor's main cost knob -- default 10 keeps a default run fast; raise it for more coverage per run.

No keyword filter, on purpose

This actor does not offer a keywords input. Every other actor in this family that has one applies it to an already-fetched full result set; here that would only ever be able to search within the maxItems articles already fetched (not the whole month), which would look like a real search while quietly missing everything outside that small window. Rather than ship a filter with that footgun, maxItems simply returns the N most recent articles, unfiltered -- narrow further downstream if you need to.

Coverage: current month, most-recent-first

Euronews publishes one sitemap per calendar month (/sitemap/articles-<year>-<month>.xml, going back to 2008) -- this actor always uses the current month's file (picked by the highest year/month pair actually present in the sitemap index, not assumed from today's date) and returns the most recently modified articles within it first.

The month sitemap also lists /video/... news-bulletin pages. Those carry a VideoObject (no article body), not a NewsArticle, so they are skipped at discovery: maxItems counts real articles, and the SEARCH_SUMMARY row reports how many video URLs were skipped (videoUrlsSkipped).

No WAF, no auth; datacenter proxy by default

The sitemap index, a month sitemap, news.xml and article pages all answered clean on every TLS profile tested (chrome124, firefox133, safari17_0, chrome99_android), cold, no warmup (re-verified 2026-09-21). Because no gate was observed, the default proxyConfiguration is Apify's shared datacenter pool ({"useApifyProxy": true}), which is included in every plan; Residential remains selectable in the proxy editor but is not required and is billed per gigabyte. See CRAWLING_METHOD.md for the full robots.txt capture, including this site's unusual three-way split between a wildcard bot policy, a named-AI-bots allowlist, and a named GPTBot/CCBot/Google-Extended blocklist -- and why the wildcard group is the one that governs this actor.

Known limits

• One request per article means maxItems is capped (200) to keep a single run's cost bounded and predictable.
• No section/category filter -- Euronews sections vary by article and aren't a clean fixed enum on the sitemap side; articleSections is still returned per article for downstream filtering.
• Video bulletins (/video/... URLs) are not returned -- they have no article body. Their count is reported in videoUrlsSkipped.
• The sitemap location is upstream's to change: it moved once already (/sitemaps/en/ to /sitemap/, September 2026). The actor follows the Sitemap: directives' current target; if it moves again the run fails loudly with an ERROR row rather than returning an empty dataset.