What the extractor reads
The input is XML from one urlset or sitemapindex. Namespace-prefixed files are supported. The parser reads only a direct loc child of each direct url or sitemap entry, so image locations and nested markup do not enter the result.
What the result contains
For a urlset, each row is a page URL with optional lastmod. For a sitemapindex, each row is the address of a child sitemap. The tool labels these results separately and does not pretend that child sitemap addresses are page URLs.
Safety and limits
The file stays in the browser. Inputs are limited to 5 MB and 10,000 entries. XML with DOCTYPE or ENTITY is rejected. URL text is never inserted as HTML or executed.