Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I have a PDF document which, for some reason, I can't copy and paste or search in. The PDF is a text-based and not image file. When I try to copy and paste the text into Microsoft Word or GNU Emacs, I get a lot of small boxes in place of the letters. When I try to search the text in Adobe Reader, I can't find words which I can see are there. The document doesn't seem to have any special protections applied to it. I've had PDFs once or twice before. I tried opening it in Google Docs but again, although it comes out as clear text, I cannot search it. Does this ring any bells with anyone?

I tried looking at the fonts of the PDF and it looks like this:

--font-65795-6-- (Embedded Subset)
Type: TrueType
Encoding: Built-in
Century (Embedded Subset)
Type: TrueType
Encoding: Built-in

followed by similar lines for Century, Helvetica, Symbol, Times-Roman, and Verdana.

share|improve this question
add comment

5 Answers

Is not there by any chance weird non-standard font used in that PDF ?

share|improve this answer
    
It may be, but I don't know how to check what font is being used. –  delete Jul 18 '10 at 8:40
1  
@Monkey Business: you can look at "Document Properties" to see which fonts are used. I bet your font uses some "custom encoding", and that's why your copy'n'paste doesn't work as expected. See also here for some hints about investigating a PDF: superuser.com/questions/163730/… and superuser.com/questions/161576/… –  Kurt Pfeifle Jul 18 '10 at 10:11
    
@pipitas: I tried that and it looks like Century, Helvetica, etc. See the edit. –  delete Jul 18 '10 at 11:23
1  
@Monkey Business: Try "pdfinfo.exe" and "pdffonts.exe" (see links from my comment above) to get a complete picture about your PDF. Amongst the XPDF tools linked to there, you'll also find "pdftotext.exe". Does this tool output any text for you? Which PDF producing software was creating the PDF (use "pdfinfo.exe")? –  Kurt Pfeifle Jul 18 '10 at 11:50
    
@pipitas - I meant weird in sense that "not present on 90-something % of reader platforms. You're absolutely right. –  Miro A. Jul 18 '10 at 18:18
show 1 more comment

This PDF probably contains its own font which is embedded into it. In this case, although the PDF will still display correctly, the correct text information is not always available and copying becomes impossible.

The fonts actually are all embedded, but in a way that all encoding information has been removed. This happens when a PDF that is still syntactically fully compliant with the PDF spec had important information about the meaning of the text in it thrown away during the process of making the PDF. It is very difficult to recover the encoding info, and sometimes the best option is to convert the pages to TIFF and then run OCR ...

You can try a PDF to Word Converter, such as AnyBizSoft or a website converter. After conversion, you can get whatever you want from the word or text file. Here is a step by step tutorial for AnyBizSoft.
(AnyBizSoft is recommended by many, but I have never used it personally.)

See also Best Free PDF Tools for more tools and converters.

share|improve this answer
add comment

With Adobe Acrobat Pro 9, I process the problematic unsearchable pdf due to custom font encoding as follows: All of these commands (1-4) are from File menu:

  1. Open the pdf
  2. List item
  3. Export -> image -> jpg
  4. Create PDF -> batch create multiple files
  5. Combine -> Merge files into single PDF

From Document Menu:

  1. scan OCR {this is to create a pdf image file searchable}.

But a 258 pages of pdf doc of 1457 KB size after converted using the above steps, it becomes a pdf file of 67565 KB. It becomes far bigger size! but it is now searchable.

share|improve this answer
add comment

Best solution yet! If you don't mind a little loss of quality...

1) Print your unsearchable pages to PDF using a rasterizing third-party PDF printer (Win2PDF worked for me). The end result is essentially a scan of the original PDF, stripped of font data.

2) Run Optical Character Recognition (OCR) text recognition tool from the Document (top bar) menu.

The end result is a searchable (albeit scruffy-looking) PDF. Something about OCR roughs up the characters in the file. But they are indeed searchable and copy/pasteable.

Enjoy.

share|improve this answer
add comment

Maybe try an online (or offline) OCR tool?

share|improve this answer
    
As I said it's text based not an image file, but I guess as a last resort I could print it to paper and OCR it back in again. –  delete Jul 18 '10 at 8:41
1  
@Monkey Business: you could use Acrobat{,Reader} to "Print as Image" with 600dpi (and print to file). That will be better quality than "print to paper" with additional scanning to get your image back for OCR. –  Kurt Pfeifle Jul 18 '10 at 10:07
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.