couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jason Smith (Commented) (JIRA)" <>
Subject [jira] [Commented] (COUCHDB-1416) the requested_path that is passed to a show is wrong on a vhost with a path
Date Thu, 23 Feb 2012 11:53:48 GMT


Jason Smith commented on COUCHDB-1416:

Ryan, I believe there is still a bug here, triggered with just a vhost/rewrite combo. Would
it be possible for you to modify the test to strictly use hostnames with no paths as the vhost

Jan, two things:

1. Your description is counter to the actual CouchDB code. Paths are valid vhost keys.
and are distinct vhosts in 1.2.x.

2. In any case, the bug in this ticket is that the path a browser sends to couch is somehow
forgotten by the time it hits a _show function as req.requested_path. Ryan might fill in the
details but AFAIK, this will fail.

[vhosts] = /db/_design/ddoc/_rewrite

"_rewrites": [{"from":"show", "to":"_show/show_doc"}]

function(doc, req) {
  // show_doc
  return JSON.stringify(req.requested_path)

GET /foo

The response should be "/foo" but it is not.
> the requested_path that is passed to a show is wrong on a vhost with a path 
> ----------------------------------------------------------------------------
>                 Key: COUCHDB-1416
>                 URL:
>             Project: CouchDB
>          Issue Type: Bug
>          Components: HTTP Interface
>    Affects Versions: 1.2
>            Reporter: Ryan Ramage
>            Priority: Minor
>         Attachments: A_0001-Testing-requested_path-for-various-combinations-of-r.patch,
A_0002-Compatibility-with-the-CLI-test-runner.patch, A_0003-Store-the-entire-requested-path-in-x-couchdb-vhost-f.patch,
> In a show or list, it is impossible to construct a full url that an end user could use
to re-request the resource, given the various combinations of vhosts and rewrites. 
> The major one is if the vhost contains a path component, this path information is not
passed to the show at all. 
> I have created three tests that highlight the condition, currently failing for one test,
with the two passing to prevent regressions.
> The commit can be found here:
> I have talked this over with JasonSmith and bennoitc on #couchdb and they asked me to
write the tests and raise the jira. 

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


View raw message