From: <jk...@ik...> - 2007-05-12 16:22:08
|
Jochen Küpper <jo...@fh...> writes: >> for the PostScript backend using (I think) dvips with psfrag, and >> neither of these is easily generalizable to work with svg or pdf. > > What's the problem of using dvi2pdf, dvipdfm, dvipdfmx, or soemthing > like that for PDF? > Looks quite similar to the dvips route for Postscript to me? The dvips route is not so simple. First matplotlib creates the PostScript file with all text strings replaced by tags, and this is fed to LaTeX using \includegraphics. The psfrag package replaces the tags with LaTeX constructs, and then the dvi file is converted using dvips into the final postscript file. There is no equivalent to psfrag that works with pdf (at least none that I know of). Probably you could create a pdf file without any texts and input that in your LaTeX file, and then -- with sufficient LaTeX-fu -- render the text strings at the correct positions, and run the LaTeX file through pdflatex, because pdftex includes a pdf parser so that you can do \includegraphics{foo.pdf}. So in that sense it may be generalizable, but it won't be a direct port of the ps backend. -- Jouni K. Seppänen http://www.iki.fi/jks |