How does your smart OSeditor handle this

the Black Sun and said:

This is fairly typical for multicolumn formats. This is why most people who have brains say, "Is this going to be read on a computer? If so, don't make it multicolumn!" Editors like vim and emacs have "visual block" selecting, where you can select an arbitrary-sized vertical block of text. That's ESC, Ctrl-V in vim (followed by y or d or whatever), check your manual for emacs. If you have a situation like this, that might not help:

column 1 text column 2 text column 1 text column 2 text column 1 text column 2 text that is really long and overflows

...but using or omitting the -layout, -raw, and-or -htmlmeta options to pdftotext may create a better-formatted file. Maybe. Using pdftotext often doesn't work well because many fine layout adjustments that can be done in commercial programs don't *have* an equivalent in text.

The structure that you want may not be there in the original PDF. It's possible that the original was encoded as (pseudocode) "(column 1 line1) horiz space (column 2 line1) vert space (column 1 line2) horiz space (column 2 line2)..." or something like that. I've seen some truly bizarre things like that in PDFs, though those were usually produced from OCRed images and so have... more problems than usual.

Also, don't forget that PDF is practically a read-only format, and PDFs produced by idiots may have "the copying text from this document is Not Allowed!!1!" bit set. No, pdftotext won't work on those, although for some of them, you can use xdpf's "print to file" to turn it into PostScript and potentially recover text data with a lot of annoying editing slogging. Don't forget the "large block of text that was encoded as an image instead of text" thing that crops up in PDFs at the least convenient times either. HTH anyway,

-- Matt GThere is no Darkness in Eternity-But only Light too dim for us to see Brainbench MVP for Linux Admin mail: TRAP + SPAN don't belong ----------------------------- penguins, is Tux." --MegaHAL