Pages in output PDF not guaranteed to be in correct order on every platform

[pjh1974]

In Linux

glob.glob("%s*.jpg" % basename)

will return the files in the natural filesystem order, rather than name order. This means that the pages in the output PDF file aren't always in the correct order. Might I suggest that the code (in pypdfocr_pdf.py) be changed to:

sorted(glob.glob("%s*.jpg" % basename))

to resolve this issue.

BTW great work - this was just what I was looking for.

[virantha]

Ah ok, thanks! Had no idea about that subtlety. I'll put this in for a patch release this weekend.
Glad the utility is helpful!

[virantha]

Fixed, used a natural-sort to ensure everything is sorted even when page count > 9.

[pjh1974]

Thanks, good spot on the page count >9 think - I feel a little silly about that one now!