On 14 August 2013 10:49, Miguel de Benito Delgado < m.debenito.delg...@gmail.com> wrote:
> > Joris asked me to convert it to C++, and it stopped there. >> > > You wrote it in scheme, I guess? Did you send the patch to savannah? > no, in bash.... it was called ps_to_eps, and looked like this: --- #!/bin/bash # 1 2 wh 3 4 wh 5 out 6 in CRC=crc32 CACHE=~/.TeXmacs/system/cache/images echo crc32 $6 CRC=`$CRC $6 | sed 's/\(.\)/\\1\//g'` echo $CRC mkdir -p $CACHE/$CRC img=$CACHE/$CRC/img_$1x$2_$3x$4.png EPSFILE=`mktemp -t ps_to_png_tmp.XXXX` if ! [ -a $img ] ; then ps2epsi $6 $EPSFILE # This is only needed because at some point # There was a bug that didn't create correct eps files, and now # I have some .tm files with wrong ps in them. gs -dQUIET -dNOPAUSE -dBATCH -dSAFER -sDEVICE=png16m -dGraphicsAlphaBits=4 -dEPSCrop -g$1"x"$2 -r$3"x"$4 -sOutputFile=$img $EPSFILE rm $EPSFILE fi ln -s $img $5 -- Primitive, I know.... but it worked pretty well. For sessions (like R...) filename+width+height won't be good... they are all the same dimensions, and probably have similar names. I see that we have MD5 routines in dpxcrypt in Pdf. I could use that. > Should I restore that? (And if I add the CRC caching to the C++ source, >> where's the best place to put it?) >> > > I guess you could start looking around in scalable_image and > cached_load_picture. > > sure, I'll look. > Would this CRC caching also cache scaled versions of the images? That is: > would we still suffer the performance hit of conversion each time the > canvas scrolls? If yes, could you add width+height to the index of the > cache? > > The scheme above caches all scales of images. One needs to think about when/how to clean the cache. Michael
_______________________________________________ Texmacs-dev mailing list Texmacs-dev@gnu.org https://lists.gnu.org/mailman/listinfo/texmacs-dev