# Rails 2 route oddity **URL:** \ **Category:** Rails\ **Created:** [January 6, 2008, 7:09pm UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178 "2008-01-06T19:09:19Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 6, 2008, 7:09pm UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/1 "2008-01-06T19:09:19Z")
I have this in my routes.db map.connect ‘posts/tags/:id/:chain’, :requirements =\> { :controller =\> ‘posts’, :action =\> ‘tags’, :chain =\> /chain/ }, :chain =\> nil given this path /posts/tags/onion+funny/chain params becomes chain: chain action: tags id: onion+funny controller: posts params[:id] is NOT ‘onion funny’ as I expect. It is under rails 1.2.5. ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 6, 2008, 8:04pm UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/2 "2008-01-06T20:04:24Z")
Why should it be ‘onion funny’? :id is ‘onion+funny’, that’s what the url says. If you wanna have additional logic, I would put it in the controller. ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 7, 2008, 2:15am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/3 "2008-01-07T02:15:05Z")
It could be “onion funny” if you did: params[:id].split(“+”).join(" ") but what if someone puts a + in their tag? On Jan 7, 2008 5:33 AM, Kitto [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > =\> /chain/ }, > > id: onion+funny > > controller: posts > > > > params[:id] is NOT ‘onion funny’ as I expect. It is under rails 1.2.5. – Ryan B. > **[frozenplague.net](http://www.frozenplague.net)** > > This domain may be for sale! Feel free to add me to MSN and/or GTalk as this email. ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 7, 2008, 2:29am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/4 "2008-01-07T02:29:06Z")
yes it’s easily handled in the controller. I just thought it odd that it worked under rails 1.2.5 but not 2.0.2 I don’t allow + in tags. I also use space as tag separator, since I figure I display tag cloud using space as separator, so it’s less confusing to the user if I require them to enter tags that way. To have compound word tags, I ask them to use - , e.g., pop-culture. (this makes for a nicer url /posts/tags/pop-culture instead of /posts/ tags/pop%20culture.) Now I wonder if the default comma as tag separator isn’t the better way to go. ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 15, 2008, 12:41am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/5 "2008-01-15T00:41:47Z")
I’m running into this exact same problem. On rails 1.2.6, by default my paths that included a name with spaces used + instead of the space. So we have URLs like /lists/Best+Movies When I tried upgrading to 2.0, link\_to now generates this /lists/Best Movies So now all my old URLs are broken. Any ideas? I’m surprised more people aren’t feeling this pain. Ben ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 7, 2008, 2:40am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/6 "2008-01-07T02:40:14Z")
Or you could just do the cool thing and use commas to seperate the tags? What if someone wants to put a space in their tags? “awesome post, large rant, big dog” as examples. On Jan 7, 2008 11:58 AM, rubynuby [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > separator isn’t the better way to go. > > > – Ryan B. > **[frozenplague.net](http://www.frozenplague.net)** > > This domain may be for sale! Feel free to add me to MSN and/or GTalk as this email. ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 15, 2008, 1:02am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/7 "2008-01-15T01:02:51Z")
Oh, I’m no routing expert by any means, so there’s definitely a chance that something weird in my setup is causing this (especially of others aren’t have this problem). I looked over my routes.rb and don’t _think_ I’m doing anything weird, but I’ve certainly been wrong before. Any and all ideas about the source of the problem or possible workarounds would be much appreciated. Thanks! Ben ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 15, 2008, 4:37am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/8 "2008-01-15T04:37:19Z")
OK, after digging around a little bit in actionpack-2.0.2/lib/ action\_controller/routing.rb, it looks like Rails has switched from using CGI.escape to URI.escape. And you can clearly see the difference in the console > > CGI.escape(’ ') > > =\> “+” > > > URI.escape(’ ') > > =\> “%20” So, my question becomes - does anyone know if this change in behavior (in the case of spaces) was intentional? Are reason why the old behavior is not supported? And any ideas on workarounds (realistically, changing all our links isn’t a great option because we’ll break existing incoming links to our app). Thanks! Ben ---
**Author:** ![rubynuby](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/ba9def/32.png) [@rubynuby](https://www.ruby-forum.com/u/rubynuby)\ **Post date:** [January 15, 2008, 4:58am UTC](https://www.ruby-forum.com/t/rails-2-route-oddity/125178/9 "2008-01-15T04:58:22Z")
OK, I did a little more investigation. It looks like this was intended. Here is a snippet of actionpack/test/controller/ routing\_test.rb * * * ``` # No + to space in URI escaping, only for query params. results = rs.recognize_path "/file/hello+world/how+are+you%3F" assert results, "Recognition should have succeeded" assert_equal ['hello+world', 'how+are+you?'], results[:path] # Use %20 for space instead. results = rs.recognize_path "/file/hello%20world/how%20are%20you ``` ## %3F" assert results, “Recognition should have succeeded” assert\_equal [‘hello world’, ‘how are you?’], results[:path] Still, this is an unfortunate situation. I would think this would break quite a few Rails apps when they upgrade to 2.0. Also, %20 isn’t a particularly pretty separator if you are trying to make your URLs easier to read. Does anyone know why this was changed? Or workaround for making spaces escape to something else? Thanks Ben