I have tried lots of softwares for converting pdf  containing Persian language to text, but non of them works.
By the way if you search google for pdf files that contain a specific Persian word it does not return accurate results as well, However I'm not sure about Arabic language. So I guess this is an open challenge at least for Persian language.

I have no idea as how to take of this conversion that you're looking for, but in general what you are trying to do is localization and not internationalization  - only for the title of the email sack

search google for OCR programs that support arabic  and PDF format as input.

There are two types of PDFs: application-generated and scanned PDFs
OCR software are for the second one (recognition of the text  contained in an image  and converts it to text).
But lots of pdf  files (at least academic ones) are just application generated, which is much easier to convert  to text compared to the scanned image.
But as for Persian I found no accurate software to do this. Acrobat Reader Professional, itself, has save as (doc, rtf, txt) feature, but it really does not work accurately for Persian language, even if you choose "select text" and copy the text, when you paste it in notepad it will ruin the characters.
Arabic characters are similar to Persian, and I guess if they knew how to extract Arabic, they would add Persian support.

www.proz.com/translation-articles/articles/128/

I have heard that OCR software for Arabic and Persian are expensive and inaccurate.

if you read again the original post you may notice that Ramya said pdf  has image.