URL record

A URL record is the structured internal representation of a URL, rather than its original text. The URL parser produces components such as scheme, host, port, path, query string and fragment identifier; URL serialization turns the record into a string. A component can be absent, and the representation of a path depends on the scheme.^[ref]

Parse https://example.org:8443/a?x=1#part. The host is example.org, the port is 8443, the path contains a, the query is x=1, and the fragment is part. The separators ? and # belong to the serialized syntax, rather than the stored query and fragment values. These fields let a program change one component without guessing its boundaries by cutting an arbitrary string.

The original spelling need not survive. Parsing https://EXAMPLE.org:443/a yields a lowercased host and no stored explicit port, because 443 is the scheme default. Therefore the record does not contain a faithful log of every character supplied by the user. If an application needs that log, it must retain the input separately.^[ref]