[
https://issues.apache.org/jira/browse/ASYNCWEB-12?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12592296#action_12592296
]
Julien Vermillard commented on ASYNCWEB-12:
-------------------------------------------
To make things clear : you parse this fast-cgi reply message with the
HttpRequestDecoder ?
Is it a usual http server behaviour ? Perhaps lighthttpd/apache parse it only
when receiving from fast cgi backend ?
> AsyncWeb does not understand the Status-Header e.g. sent by FastCGI backends.
> -----------------------------------------------------------------------------
>
> Key: ASYNCWEB-12
> URL: https://issues.apache.org/jira/browse/ASYNCWEB-12
> Project: Asyncweb
> Issue Type: Bug
> Components: Common
> Environment: AsyncWeb 0.9.0 CVS used in AsyncFCGI
> Reporter: Daniel Wirtz
> Priority: Trivial
>
> When a Status header is provided, AsyncWeb does not transform it into the
> equivalent HTTPStatus in a HttpResponse.
> A response from a FastCGI backend could look like this:
> Content-Type: text/html\r\n
> Status: 404\r\n
> \r\n
> File not found!
> or
> Content-Type: text/html\r\n
> Status: 404 Some Custom Message\r\n
> \r\n
> File not found!
> in PHP this looks like:
> header("404: Some Custom Message"); or
> header("HTTP/1.1 404 Not Found"); etc.
> -> both produce the Status-Header.
> As far as I know the webserver (this would be asyncweb in this case) converts
> the status header into the correct http status (e.g. "HTTP/1.1 404 Not
> Found"). I'm using PHP a lot, the status header is the default behaviour and
> Apache / Lighttpd handle it this way (even with mod_php instead of FastCGI).
> I don't want to preparse the response because of unneccessary double checking
> instead of simply feeding it into a HttpResponseDecoder. However, I think the
> status header is not part of the http specification but since it's used a
> lot, it would make sence.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
