Geert Josten wrote: Hi,
> Can you give some kind of test case? I thought MarkLogic should > be able to handle multipart requests properly.. Sure. Let's consider the example 5.2 in the RFC 2387 for instance, at http://tools.ietf.org/html/rfc2387#section-5.2. If you post it to the following query (using your favorite HTTP tool): xquery version "1.0-ml"; <results> { for $b in xdmp:get-request-body() return <result> { (: Transform binary() to xs:base64Binary :) (: To avoid type error, cannot add binary() to element :) if ( $b instance of document-node() and $b/node() instance of binary() ) then xs:base64Binary($b/node()) else $b } </result> } </results> then I get the following result (only one body): <results><result>LS1leGFtcGxlLTIKQ29udGV [...]</result></results> I've truncated it here, but when I base64-decode the returned result, I get back the entire multipart request entity content (AKA its "body"). Which means that in that case, the multipart is not decoded, it is returned to the query as a big binary chunck. So I wonder if there is another API call to get the decoded multipart content (in the above example as three different contents, one text and 2 binary pictures), or if there is a library to rather decode the big binary returned by xdmp:get-request-body()? Regards, -- Florent Georges http://fgeorges.org/ http://h2oconsulting.be/ _______________________________________________ General mailing list [email protected] http://developer.marklogic.com/mailman/listinfo/general
