Existing mime sniffer wasn't great at distinguishing between plaintext file types, so added a custom extension based mapping for common web formats that may be expected to be used with this.