[Biopython-dev] [Bug 1939] Doc/Makefile does not build pdf, html, txt files completely correctly

[bugzilla-daemon at portal.open-bio.org]

http://bugzilla.open-bio.org/show_bug.cgi?id=1939





------- Comment #7 from bill at barnard-engineering.com  2006-01-28 17:56 -------
In the original report:
> > Finally the output files (html, pdf, txt) are currently
> > checked into CVS; they should be removed from CVS and
> > generated by the make/build process instead.
> 
> Having them in CVS can be very helpful for those just
> browsing the web interface to see what's changed recently.

Agreed. Maybe it's best to leave them in CVS. Whoever checks in new doc source
files will need to remember to run the make to update the outputs though.

> 
> In comment 1 Bill wrote:
> > I added a one line mod to setup.py to call
> > the Doc make [ os.system('make -C Doc') ].
> 
> Will this work on the Windows platform?  I have never tried to build BioPython
> on Windows, but I am fairly sure that make isn't present by default.

Good point. The system call returns a zero on my linux box. If make is not
found in the path, as would be normal on a windows box, then the os.system call
simply returns the exit status for the command and the code continues running.
The Doc make calls the pdflatex, hevea, & hacha programs which are unlikely to
be on a windows box either.

In any case, if the Doc build programs are not present in the path then some
error messages are written to stdout and setup.py continues to run. I think the
worst case on Windows would be that the Doc build output files might be out of
date with respect to the .tex source files.

So I'd leave the new line in setup.py.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.