[tex-live] dviselect wishlist?
furue at hawaii.edu
Mon Nov 18 00:09:42 CET 2013
First of all, thank you all who responded.
> Is it really necessary to do such operations with DVI?
> It is easier with PDF, you can split a PDF file
> to one-page files by
It's not "necessary". But, my end product is a bunch of
EPS files, not PDF, and my LaTeX source includes
EPS figures via \includegraphics . So, dviselect
is the most natural tool to use. But, as I said in my
initial post, this is not a big deal. There are
If anybody is interested, what I'm doing right now
is to arrange, align, and combine EPS figures and
add labels to them (using the picture environment
of LaTeX) and produce a bunch of EPS files, each
including several EPS figures with labels added.
pdfLaTeX is much slower than LaTeX, especially when you
include a lot of figures, and also even the fastest PDF
viewer is much slower than xdvi, especially when you
run the viewer on an remote server through X over SSH.
> By the way, I think this bit:
> while true; do
> dviselect -i infile.dvi -o out$n.dvi $n || break
> could be simplified to:
> while dviselect -i infile.dvi -o out$n.dvi $n; do
Right! When I wrote my example for the posting, I based
it on my initial idea:
for n in `seq -w 1 99`; do
dviselect -i infile.dvi -o out$n.dvi $n || break
because I wanted "out01.dvi", "out02.dvi", etc., instead
of "out1.dvi", "out2.dvi", etc., and I didn't want to
add unnecessary details to my example: the seq command
isn't a standard Unix tool, and
while true; do
num=`printf "%2.2d" $n`
dviselect -i infile.dvi -o out$num.dvi $n || break
n=`expr $n + 1`
includes unnecessary details. This made me (at that time)
unable to see the simplification you showed.
More information about the tex-live