Lesson 02 · Host and path - Build an HTTP Client

Host and path

After the scheme comes the part that names the machine and the resource on it. Today you split the authority (the host) from the path, so the client knows both where to connect and what to ask for.

The goal

Parse the host and path out of the part of the URL after the scheme.

Start here - the target

TO DO

Scenario: Separating the host from the path

Giventhe URL string "http://example.com/a/b"

Whenit is parsed

Thenthe URL has host "example.com" and path "/a/b"

Andparsing "http://example.com" (no path) gives host "example.com" and path "/" (the default)

Background

After scheme:// comes the authority - for our purposes the host, the name of the machine to connect to - followed by the path, the resource to ask that machine for. In http://example.com/a/b, the host is example.com and the path is /a/b. The authority runs from just after :// up to the first /, ?, or #, whichever comes first; everything from that / onward is the path.

The one edge worth pinning now is the empty path. A URL like http://example.com names a host but no resource, and a request must still ask for something - so the path defaults to /, the root. Getting this default in early means every later lesson can assume the path is never empty.

Make it work

// after "scheme://" comes the authority, then the path.
// the authority ends at the first "/", "?", or "#" - or the
// end of the string. everything from that "/" onward is the path.
// when there is no path at all, the path defaults to "/".
func Parse(raw string) (*URL, error) {
  // strip "scheme://", then split authority from path on the first "/"
}

CheckpointDONE

A parsed URL now carries its host and a path that defaults to "/". Commit and stop here.