On Tuesday 11 November 2003 14:48, Andrew Gaffney wrote: > Christopher Stanton wrote: > > RedHat Linux 9 > > Perl v5.8.0 built for i386-linux-thread-multi > > > > I am trying to parse a mjpeg stream out of an html response. The http > > server is using server push to push the stream of jpegs to the client. I > > have written a test client and am able to receive the stream but am > > having trouble figuring out what Perl libraries I need to use to separate > > and save the individual jpegs as actual jpegs and not MIME encoded data. > > > > The Content Type is "multipart/x-mixed-replace; boundary=--myboundary". > > "--myboundry" is the flag used to delimit the individual data fields. > > > > This is a stream of JPEGs so, the server will continue streaming as long > > as the connection is open. I am using the Net::HTTP library since I have > > to parse as it arrives rather than wait for the whole page to be > > downloaded (since it can't be). > > > > The stream's format: > > #------------------------------------------------------------------------ > >--------- HTTP/1.0 200 OK > > Connection: Close > > Server: Camd > > Content-Type: multipart/x-mixed-replace; boundary=--myboundary > > > > --myboundary > > Content-Type: image/jpeg > > > > <SOME JPEG IMAGE> > > > > --myboundary > > Content-Type: image/jpeg > > > > <SOME JPEG IMAGE> > > > > --myboundary > > Content-Type: image/jpeg > > > > <SOME JPEG IMAGE> > > > > --myboundary > > . > > . > > . > > #------------------------------------------------------------------------ > >--------- > > > > I am able to break all of the pieces out of the original stream and am > > currently just saving them to a file: > > chunk-1.mime > > #------------------------------------------------------------------------ > >--------- > > > > Content-Type: image/jpeg > > > > <SOME JPEG IMAGE> > > > > #------------------------------------------------------------------------ > >--------- > > > > Can anyone point me in the direction of the libraries or tools I need to > > use to export the encoded <SOME JPEG IMAGE> to a raw JPEG so I can save > > it to a file? > > Can't you just strip out "Content-Type: image/jpeg\n\n"?
I was under the impression that the image was mime encoded. The chunk is being striped down to: #------------------------------------------------------------------------ Content-Type: image/jpeg^M ^M <SOME JPEG IMAGE>^M ^M #------------------------------------------------------------------------ ^M => \r\n I have tried stripping the leading "\n", "Content-Type: image/jpeg\r\n", "\r\n" line, "\r\n" after the image data, and the trailing "\r\n" line but the resulting file does not seem to be a valid jpeg image. any ideas? thanks, Christopher -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
