Hi Tim:
I think that this is the same problem.
It is caused by the way that Hobo uses the request path. I made a change in
hobo_rapid_helper.rb to resolve the issue.
Here is my pull request with the commit info:
https://github.com/Hobo/hobo/pull/194/commits/b4376d5b62eef3a41e77d632429527a3d04d9371
Ed
On Wednesday, August 2, 2017 at 9:35:27 PM UTC-6, Tim Griffin wrote:
>
> Hi Ed;
>
> I sent this to Ignacio and then found your semi-related post, and figured
> the more eyes the better. I also have an app hosted in a non-root location
> and (after considerable debugging) discovered a problem in
> hobo/hobo_route_helper.rb:76 (in Hobo 2.2.6). I wondering if you also think
> this to be a problem, and related to your issue?
>
> Here's the line from hobo_route_helper:
>
> if Rails.application.config.action_controller.relative_url_root
> base_url =
> url.gsub(/^#{Rails.application.config.action_controller.relative_url_root}/,
> "")
> end
>
> In my app, I've set relative_url_root to "/app/lts-test" and and my
> default_url_options are also set as follows:
>
> { :host => 'example.com',
> :protocol => "https",
> :port => nil,
> :only_path => false,
> :script_name => '/app/lts-test' }
>
> Notice that my host is a full URL (which I normal, I guess): example.com
>
> When Hobo calls into its route_helper, though, the regex is set to look
> for my relative_url_root at the *beginning* of the string ("gsub(/^
> #{Rails...."), which fails because at this point my url is actually this:
>
> https://example.com/app/lts-test/users/1/edit
>
> This means, the "/app/lts-test" part isn't being removed, causing other
> failures such as this:
>
> recognize_path has failed: No route matches "
> https://example.com/app/lts-test/user/1/edit"
>
> I tried switching my default_url_options to be :only_path => true which
> partly solves the problem, but then causes trouble with other routes that
> expect a full path, even ones that within Hobo that I've not overridden.
>
> So, does this strike you as an error? Why would the relative_url_root be
> searched for only at the *beginning* of the string? It seems to make more
> sense to remove it from anywhere in the entire url string.
>
> Your thoughts?
> Tim
>
>
>
>
>
>
>
>
>
> On Thursday, June 15, 2017 at 9:57:48 PM UTC-4, Ed Gomolka wrote:
>>
>> I'm not sure if this is related, but I submitted a pull request for a
>> problem that I was having when I tried to locate my Hobo application in a
>> non-root location. Hobo has a variable called page_path returns the full
>> path. It messes things up if you are setting RAILS_RELATIVE_URL_ROOT.
>>
>> Here is the pull request: https://github.com/Hobo/hobo/pull/194
>>
>>
>> On Tuesday, June 13, 2017 at 9:54:19 AM UTC-6, kevinpfromnm wrote:
>>>
>>> Subject pretty much sums it up. I have part updates working in dev, but
>>> production is on passenger. To confirm that it's the relative url path
>>> breaking it, I configured dev similarly. It looks like the recognized_path
>>> call fails to account for the relative path
>>>
>>
--
You received this message because you are subscribed to the Google Groups "Hobo
Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/hobousers.
For more options, visit https://groups.google.com/d/optout.