Improve pdfiototext text extraction

[kleuter]

Trying to understand how to pdfiototext.c works. The code seem to output too many extra unnecessary spaces for this PDF.

weird_spaces.pdf

[michaelrsweet]

Keep in mind that the pdfiototext program is just a proof of concept/example. If you need text output there are far better options available...

In this case, the spaces are placed between fragments in the content stream. This particular file does a lot of kerning and pdfiototext is not smart enough to merge adjacent fragments.