type, but didn't try the reset attribute. Interestingly, when viewing the
returned data without the CFCONTENT tag, there isn't any extra whitespace
that I can detect (I used a raw telnet connection so I could see
everything), so who knows that that actually did, but it worked.
Cheers,
barneyb
> -----Original Message-----
> From: Dave Carabetta [mailto:[EMAIL PROTECTED]
> Sent: Friday, February 13, 2004 5:39 AM
> To: CF-Talk
> Subject: Re: Custom WSDL
>
> > This is undoubtedly a total noob question, but not in my
> general realm of
> > operations, so I'm totally clueless.
> >
> > I'm writing custom WSDL for a couple web services that
> return complex
> types
> > and need to be consumable by .NET clients. So far, so
> good. As long as
> the
> > wsdl file is a static file on the server or generated by the cfc
> > (mycfc.cfc?wsdl), the .NET `wsdl` tool will happily digest
> it and spit out
> > the proxy class, and the .NET app will read the data in
> (though untyped).
> > As soon as I make the wsdl dynamic (passing the static file
> through CF),
> > everything goes awry.
> >
> > I can happily add any extension to the file (other than cfm
> and cfc, of
> > course), and everything stays fine, so it's not the file name that's
> > screwing things up. Whether it matters or not, IE's usual XML
> > highlight/expand/contract thing mirrors the issue. It
> displays the pretty
> > output for the static files or the ?wsdl output, but just a
> blank screen
> > (with the XML available from view source) when it's
> dynamically generated,
> > so something's definitely different, but I have no idea what it is.
> >
> > The reason I need to dynamically generate is because the
> same WSDL needs
> to
> > work on multiple domains (so I'm using a variable to
> populate the address
> > within the <wsdl:service> block). If anyone's got any
> tricks that attack
> > from this angle, I'd be just as interested.
> >
>
> Are you absolutely sure there is zero whitespace in the dynamically
> generated WSDL file? Have you tried serving the file with
> cfcontent to see
> if that makes a difference? (I only say that because
> cfcontent's "reset"
> attribute is pretty good, though not perfect, at getting rid
> of whitespace
> up to the point of the cfcontent tag call.)
>
> Regards,
> Dave.
>
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]
