Backward search in ...TeX generated PDF documents

[Oscar Chinellato <chinellato ma tum de>]

Dear all,

due to my job, I'm often confronted with the task of writing reports
and (scientific) articles, which is why I'm a frequent user of
(pdfla)tex in combination with evince as my pdf-viewer. It goes
without saying, that a backward search functionality would ease the
(tedious) task of editing a manuscript. Unfortunately, to the best of
my knowledge, yap is the only viewer which supports this functionality,
yet only for .dvi documents.

After having had a closer look at the source files, I added a few
lines in the file ev-view.c (after the "Oscar" tag) which, a
double-click being executed inside an opened document, retrieve
information like

file name
current page no.
x/y coordinates

and issue a shell command (call to a Python script named
pdfsync.py). The latter compares the gathered information with an
additional file generated by pdflatex (when using/including the
package pdfsync) and is then able to (roughly) determine the .tex
source file and the line number to which the double-click has
pointed. This, together with the emacsclient functionality, allows a
decent backward search behaviour which works like a charm.

I'm aware, that my hack, as it is now, can not go into the repository
since, so far, it is too simplistic and just restricted to
pdfsync/emacs (although an extension to further decent editors like
vim etc. seems straightforward). What should the next steps be?

Many thanks in advance!

Cheers,
Oscar