A multilingual news corpus built from Common Crawl CC-News, indexed and queriable in milliseconds, cleaned an enriched with language and topic id