Hi On Mon, Apr 25, 2011 at 7:05 PM, Daniel Kulp <[email protected]> wrote: > On Monday 25 April 2011 1:57:48 PM Benson Margulies wrote: >> Well, but why is the trailing > omitted? > > Looks like a bug: > > Line 159 of Attachment Util: > id = "Content-ID: <[email protected]"; > > Likely something is wrong around there. The best option is probably to move > the "decode" into the above "if" and set the id to just > [email protected] or something at that point. >
OK, I think that if should also be changed a bit, to move the check for '<' inside of the if block, because replacing Content-Ids which may not have angle brackets with default values seems not right > And "Content-ID: " definitely shouldn't be there. > Benson - can you please get the list of Attachments from MultipartBody and check individual Attachment headers ? How is Content-ID reported ? Or is it how you found about malformed (default) Content-IDs in the first place ? > Can you log that (and maybe fix it)? Benson created JIRA and I'll look into it. That should not affect multipart/form-data requests much as @Multipart and MultipartBody.getAttachmentObject will also check Content-Disposition, but has to be fixed anyway. Cheers, Sergey > > Dan > > > >> >> On Mon, Apr 25, 2011 at 1:53 PM, Sergey Beryozkin <[email protected]> > wrote: >> > Hi >> > >> > On Mon, Apr 25, 2011 at 6:47 PM, Benson Margulies <[email protected]> > wrote: >> >> I'm just posting a form from Chrome. Note the strange content-id, >> >> which is the same in all of my attachments. This is with 2.4.0. >> >> >> >> {Content-Disposition=[form-data; name="file"; filename="cmp.patch"], >> >> Content-Type=[application/octet-stream], Content-ID=[Content-ID: >> >> <[email protected]]} >> > >> > CXF probably defaults to <[email protected]> on the input ? >> > The format is a bit strange all right, but I think Content-Id is >> > actually OK, you can get the list of headers and Content-Id in >> > particular from Attachment objects and confirm >> > >> > Cheers, Sergey > > -- > Daniel Kulp > [email protected] > http://dankulp.com/blog > Talend - http://www.talend.com >
