Re: PDF to TIFF hash - Avoid date/time reference

On 1 December 2015 at 07:44, Giacomo Boccardo <[email protected]> wrote:
> Hi,
>     I'm using Ghostscript to extract pages' images from a PDF using a
> command like:
>
> gs -dQUIET -dNOPAUSE -dBATCH -dSAFER -dFirstPage=1 -dLastPage=1
> -dBandHeight=100 -dNOPAUSE -sDEVICE=tiff12nc -r200
> -dNumRenderingThreads=2 -dBandBufferSpace=500000000
> -dBufferSpace=750000000 -sBandListStorage=memory -sOutputFile=page1.tiff
> input.pdf
>
> For caching purposes I need that the hash of each tiff must be the same
> if it's produced from the same page. Unfortunately, I noticed that a
> date/time reference is added at the "beginning" (prolog?) of each file.
> That's the only thing that alters the hash: if you produce images in the
> same second, the hash is the same.
>
> So, is it possible to avoid to add that time reference?
> If it's not possible, is there a way to test if two tiffs have the same
> "appearance", without considering the metadata?

You cannot disable that, to the best of my knowledge. I'm not even
sure it's something we have control over - I'd have to look at the
source, and I don't have time right now.

libtiff comes with a few tools that exercise the library, one being
tiffcmp which compares the image samples byte-for-byte. That may
satisfy your requirements.

On Linux distros, they're usually packaged under a name like "libtiff-tools".

> I suppose the last chance is to change the source code and recompile it :)

When I have some time, I'll check the gs source, and see if it is
something we control, or if it's something libtiff does automagically
- I certainly won't be changing the libtiff code!

Chris