Possibly - it isn't clear from their post why SRW doesn't fit their
purposes.
Matthew
> -----Original Message-----
> From: Z39.50 Next-Generation Initiative [mailto:[log in to unmask]]
> On Behalf Of Mike Taylor
> Sent: 07 July 2005 14:50
> To: [log in to unmask]
> Subject: [[log in to unmask]: [Yazlist] large HTTP GET's]
>
> Here is another use-case where someone needs SRU/POST.
>
> ------- Start of forwarded message -------
> Envelope-to: [log in to unmask]
> Delivery-date: Thu, 07 Jul 2005 15:39:25 +0200
> Date: Thu, 07 Jul 2005 15:38:59 +0200
> From: Ko van der Sloot <[log in to unmask]>
> X-Accept-Language: en-us, en
> To: yaz <[log in to unmask]>
> X-Enigmail-Version: 0.90.0.0
> X-Enigmail-Supports: pgp-inline, pgp-mime
> Content-Type: text/plain; charset=ISO-8859-1
> X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at uvt.nl
> Subject: [Yazlist] large HTTP GET's
> Reply-To: "Discussion on the YAZ Z39.50 toolkit"
> <[log in to unmask]>
> Sender: [log in to unmask]
> X-Spam-Checker-Version: SpamAssassin 3.0.0 (2004-09-13) on
> bagel.indexdata.dk
> X-Spam-Status: No, score=-2.6 required=5.0 tests=BAYES_00
> autolearn=ham
> version=3.0.0
> X-Spam-Level:
>
> Hello,
>
> We encountered the following problem(s):
> We have an SRU/SRW service build, based on the Generic File Server.
> One of our applications tries to send very large (>16.000
> bytes) GET requests to the service.
> This fails with an HTTP 500 message.
>
> The reason is that in comstack.c there is a 8192 byte limit
> set to HTTP headers. (in cs_complete_auto) I can see some
> reason behind limitting. But why not some more?
> According to RFC 2616 3.2.1 URI's can be of unbound length.
> Which is maybe a bit to much ;) But anyawy: If the limit is
> exceeded, an 414 (Request-URI Too Long) should be returned.
> Lifting the limit solves our problem.
>
> Then we also tried converting GET to POST.
> This fails because the yaz_sru_decode() function (in
> srwutil.c) only supports GET.
> It was not that difficult to add some code to handle POST,
> but i'm not so sure that it is an optimal/general/conforming solution.
>
> So what to do now?
> We could propose a patch to raise the HTTP header limit to "unbounded"
> (say 100K)
> We could propose a patch to add POST support to SRU. (We have
> a working
> hack)
> We could do both.
>
> any comments?
>
> Ko vd Sloot
> Tilburg University
>
>
> _______________________________________________
> Yazlist mailing list
> [log in to unmask]
> http://www.indexdata.dk/mailman/listinfo/yazlist
> ------- End of forwarded message -------
>
|