View Single Post
  #11  
Old Aug 24, 2012, 08:40 AM
LiquidAcid LiquidAcid is offline
Trusted Editor
 
Join Date: May 2008
Posts: 1,644
Default

Quote:
Originally Posted by Efendija View Post
^ Vastly hmmm... It depends. Open the pdf with any pdf reader, use the rectangle selection area tool of greenshot and you have an image ready for uploading. Practically two steps. I won't argue if the other methods work better for other people, but this one is quick and easy for me.
It is _not_ quick. You have to do this selecting and copying for _every_ area you want to copy. Furthermore it involved a needless decompression step. Most of these booklet PDFs will contain DCT-encoded image data anyway, which you would then decompress to RGB and then recompress to JPEG in an additional step. And you have to do this all manually.

With pdfimages it's a one-liner:
pdfimages -j input.pdf outputprefix

Add a resizing step with e.g. ImageMagick and it's two lines. You can even get "the full package" for content which is not only consisting of encapsulated bitmap data: pdftoppm
And even here it's a maximum of two lines.

Your approach is kinda like copy&pasting every single trackname manually to your media player when tagging an album: Select trackname, ctrl+c, switch to media player, select the track, open properties box, ctrl+v, switch back to browser.
I really, _really_, REALLY hope nobody is actually doing this in practice. because it is:
(1) Slow
(2) error-prone
(3) tedious

EDIT: And probably most important:
(4) Totally and utterly defeats the purpose of a computer, which was invented to do the "dirty" work for you.
Reply With Quote