From 4e09d776f5368f432971429f9868052f55151555 Mon Sep 17 00:00:00 2001 From: Anthony Ramine Date: Mon, 16 May 2011 11:36:28 +0200 Subject: Document the path info feature --- README.md | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'README.md') diff --git a/README.md b/README.md index 5c7df2d..133af3b 100644 --- a/README.md +++ b/README.md @@ -124,6 +124,16 @@ you accept anything in that position. For example if you have both "dev-extend.eu" and "dev-extend.fr" domains, you can use the match spec `[<<"dev-extend">>, '_']` to match any top level extension. +Finally, you can also match multiple leading segments of the domain name and +multiple trailing segments of the request path using the atom `'...'` (the atom +ellipsis) respectively as the first host token or the last path token. For +example, host rule `['...', <<"dev-extend">>, <<"eu">>]` can match both +"cowboy.bugs.dev-extend.eu" and "dev-extend.eu" and path rule +`[<<"projects">>, '...']` can math both "/projects" and +"/projects/cowboy/issues/42". The host leading segments and the path trailing +segments can later be retrieved through `cowboy_http_req:host_info/1` and +`cowboy_http_req:path_info/1`. + Any other atom used as a token will bind the value to this atom when matching. To follow on our hostnames example, `[<<"dev-extend">>, ext]` would bind the values `<<"eu">>` and `<<"fr">>` to the ext atom, that you -- cgit v1.2.3