I have a Rails route that takes stock ticker symbols as the :id
feeds/AMZN
will return a page for Amazon
feeds/AMZN.csv
will return a CSV representation of the same data.
But I also need to accomodate stocks like VIA.B (Viacom) so that both of these routes work:
feeds/VIA.B (html)
feeds/VIA.B.csv (csv)
Is this possible? How would I set the routing up?
I ran into this while patching the RubyGems API recently (trying to access the flickr.rb
using the API (/api/v1/gems/flickr.rb.json
) was not working).
The trick was to supply the route with a regexp to handle the :id
parameter, and then specify valid :format
. Keep in mind that the :id
regexp needs to be "lazy" (must end with a question mark), otherwise it will eat the .csv
and assume that it's part of the id. The following example would allow JSON, CSV, XML, and YAML formats for an id with a period in it:
resources :feeds, :id => /[A-Za-z0-9\.]+?/, :format => /json|csv|xml|yaml/
Old question, but I found a much simpler way that works with nested routes (I'm on Rails 3.2.4). This way allows all characters (including the dot) as opposed to the accepted answer which makes you specify the allowed charcters.
resources :feeds, :id => /([^\/])+?/
Note that I had found some other suggestions (e.g. here: http://coding-journal.com/rails-3-routing-parameters-with-dots/) of doing something like:
resources :feeds, :id => /.*/
but that didn't work for me with nested routes for some reason.
I ran into this as well, but in the reverse direction. (url_for()
produces "No route matches" only for IDs with . in them.)
I'm using match
instead of resources
to allow some name munging. If you're doing the same, this is what the fix looks like:
match "feeds/:id" => "stocks#feed", :constraints => {:id => /[\w.]+?/, :format => /html|csv/}