When scraping web content, one usually hopes that the content is laid out logically, and that proper or at least consistent web annotation exists. This means well-nested HTML, appropriate use of tags, descriptive CSS classes and unique CSS IDs. Ideally it also means that a given CSS selector will yield a consistent datatype, also. ..well that's Sqrape. In fact, see examples/tweetgrab.go for the above as a CLI tool.