Heiko Oberdiek wrote:
400K files don't need to be read as a whole. An intelligent implementation would parse the necessary parts. With the information of \pdffilesize files can also read from behind and parse pdf or dvi files, for example. But it is easy to add another option for the read direction for parameter "whence" of fseek (SEEK_SET, SEEK_CUR, SEEK_END).
that's indeed the point i wanted to make: that kind of parsing only makes sense if one can also seek; better make a full implementation now then wait for new features later.
Currently the file is not remembered, thus "current" has no meaning. Keyword "offset" can be given an option "end":
\pdffiledump [offset [end] <int>] [length <int>] <general text>
"offset 10 length 2" means byte 10 and 11, "offset end 10 length 2" means byte 10 and 11 before end of file.
ah, i see, interesting Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------