Is it possible to search multiple pdf files using the 'grep' command. It doesn't seem to work, how do people search content on multiple pdf files?
I was just looking for a simple search capability here
Samuel
2010-04-22 06:18:21
to note: in the future simple search is more of a superuser question, imo. I answered these because these are solutions you might use to do it programmatically.
xenoterracide
2010-04-22 08:50:50
+2
A:
Pdf is a binary format, that's why searching it with grep is not that helpful. You can search the strings is a pdf with grep like this:
ls dir_with_pdfs/*.pdf|xargs strings|grep "keyword"
Or you can use the pdf2text command on pdf's and then search result with grep.
Bozhidar Batsov
2010-04-22 05:48:54
Sorry, that is just nonsense! PDF normally uses compressed objects and even if the objects were uncompressed, the text is only partly written in cleartext inside the pdf.
Patrick
2010-04-22 12:12:03
+1
A:
Well, PDF is a binary format, and grep can search binary files as if they were text
grep -a
or you can just use pdftotext (which comes with xpdf) like this:
pdftotext whee.pdf | grep pattern
wsh
2010-04-22 05:49:43
I am able to get this command working only if a pass a "-" after the file name to be searched. i.e.pdftotext whee.pdf - | grep pattern
Samuel
2010-04-22 06:19:53
Oh, weird...the - means stdout (which is what you need the text to be passed to for the pipe to work properly), which in my shell you don't need to specify afaik.
wsh
2010-04-22 06:24:27
+1
A:
You don't mention which OS you're using, but under Mac OS X you can use mdfind from the command line:
mdfind -onlyin search/directory/path "kind:pdf search text"
Coxy
2010-04-22 07:00:49
+1
A:
PDF is a binary dump of objects used to display the pages. There may be some meta data you can grep but the actual page text is in a Postscript stream and may be encoded in a variety of ways. Its also not guaranteed to be in any order. You need to think of PDF as more like a Vector image file than a text file.
There is a short article explaining text in PDFs in more detail at http://pdf.jpedal.org/java-pdf-blog/bid/27187/Understanding-the-PDF-file-format-text-streams
mark stephens
2010-04-22 08:08:20