Added support for isolating sitemap URLs from the array and adding them to the sitemap index.

[RAMOhio]

Had a case where it was necessary to generate sitemaps using another method as well as use this library (long story)

Modified the code so that if a URL contains "sitemap" and ends with ".xml" it'll assume it's a sitemap, remove it from the URLs, and add it to the sitemap index.

Scenarios:

• There are under 50,000 URLs and at least one of the URLs is a sitemap: It'll generate the sitemap index containing the sitemap(s) found in the URLs and sitemap-0.xml will contain the URLs (with the sitemap URLs removed).
• There are over 50,000 URLs and at least one of the URLs is a sitemap: It'll generate the sitemap index and sitemap-#.xml routes like normal (with the sitemap URLs removed), but all of the sitemaps found in the URLs will be added to the sitemap index.
• There are only sitemaps in the URLs: It'll create the sitemap index with the sitemaps and nothing else.
• There are no sitemaps in the URLs: It'll behave the same as the package has always behaved.

Notes:

• Implementation could be improved but it works.
• If someone uses sitemap-#.xml in their URLs and the package ends up generating that same path, I'm not sure what the behavior would be, I haven't tested that. For my extra sitemaps I use a different route structure than the library uses (recommended approach).
• Sitemaps in the URLs must be on the base domain (didn't do any parsing to see if it's external)

[Kikobeats]

Hey @RAMOhio, I will be happy to merge this if you add a test 🙂