Identifying image-only vs. image+text PDF files

[Zamdrist]

I have a large collection of PDFs in a large directory/subdirectory structure, we are talking 366K PDFs.

I need to find some way of identifying which are image only from all others.

I've tried various things, I have both Windows & Linux at my disposal. I was trying to parse the PDFs source and search for "Font" Or "Annot" which I believe would identify something other than image-only types of PDFs. Baring that I'm hoping someone knows of a better way or if there is a tool out there already.

Any help would be greatly appreciated. Thanks.

[Zamdrist]

Does anyone have any thoughts on this? The need is urgent and would great appreciate any feedback. Thank you.

[Leonard_Rosenthol@adobeforums.com]

There are a number of 3rd party solutions that offer this functionality for both Windows and Linux. A web search should turn them up.

If you want to use Acrobat, you would need to write a custom plugin OR you could try to use either IAC or JavaScript to do text extraction (though that's not 100% reliable since you could have a map w/o text).

Leonard

[Zamdrist]

Thank you Leonard for your reply. Despite my reasonably good Googling skills, I've not been able to find anything in the sorts of a plug-in or third party tool.

Ironically I keep finding my own posted question, for which no one seems to have an answer for.

Keep in mind I need a tool which will search and return ALL PDFs in a large directory/sub-directory structure and their path names.

Not looking for a tool that will just let me know if the open PDF is image only or image+text, which is quite easy to ascertain.

Thank you.

[Leonard_Rosenthol@adobeforums.com]

I don't now of a company with an "off the shelf" directory walker, etc. but a variety of companies have the necessary component to do the checking on a single PDF (which you could then connect up to the folder walker).

Check with companies such as Apago ([url]http://www.apago.com[/url]), Traction Software ([url]http://www.traction-software.co.uk/[/url]) and Glyph and Cog ([url]http://www.glyphandcog.com[/url]).

Leonard

[Carsten Hammer]

You could have a look at the adobe ifilter plugin for windows. I allows
to index pdfs. The windows platform sdk contains example binaries how to
use this windows api. It is very easy to examine pages of pdf files
using the samples in the sdk. I guess a pdf page containing only images
gives back no text.

Something like


filtdump testfile.pdf


should extract text after you installed the ifilter plugin. filtdump is
part of the platform sdk.

[url]http://www.adobe.com/support/downloads/detail.jsp?ftpID=2611[/url]

[url]http://msdn.microsoft.com/library/default.asp?url=/library/en-us/indexsrv/html/ixrefint_9sfm.asp[/url]



Zamdrist schrieb:

> Thank you Leonard for your reply. Despite my reasonably good Googling skills, I've not been able to find anything in the sorts of a plug-in or third party tool.
>
> Ironically I keep finding my own posted question, for which no one seems to have an answer for.
>
> Keep in mind I need a tool which will search and return ALL PDFs in a large directory/sub-directory structure and their path names.
>
> Not looking for a tool that will just let me know if the open PDF is image only or image+text, which is quite easy to ascertain.
>
> Thank you.

[Carsten_Hammer@adobeforums.com]

You could have a look at the adobe ifilter plugin for windows. I allows to index pdfs. The windows platform sdk contains example binaries how to use this windows api. It is very easy to examine pages of pdf files using the samples in the sdk. I guess a pdf page containing only images gives back no text.

Something like

filtdump testfile.pdf

should extract text after you installed the ifilter plugin. filtdump is part of the platform sdk.

<http://www.adobe.com/support/downloads/detail.jsp?ftpID=2611>

<http://msdn.microsoft.com/library/default.asp?url=/library/en-us/indexsrv/html/ixrefint_9sfm.asp>