find not working as expected

luusac · 08-27-2008, 03:32 PM

Hi,
I am trying to find (on a NAS) the string "Hesketh" anywhere in under root (recursive) but only in .doc or .Doc files.

I found this snippet on google:

find . -name *.doc | grep -lir "Hesketh" *

it has been chugging away for hours and so far has come up with:

EMAIL/LU/FOLD391O.PMM
EMAIL/LU/FOLO77O5.PMM

These are email folders, but the file extension is .PMM and not .doc

What is going on here - is it the fault of the trailing * ? But why, because the input of grep should be the output of find right? And find is only finding files of .doc type, right ?

Can anybody help?
thanks
lu

matthewg42 · 08-27-2008, 03:40 PM

Find produces a list of files. If you pipe that list into grep, it will identify file names which match the pattern passed to grep as an argument.

If you want to search inside the files, use xargs to pass the listed file names to grep as extra arguments. This way grep will search inside these files.

If you use the -print0 option to find and the -0 option to xargs, file names will be delimited with the ASCII NUL character, avoiding problems with file names containing whitespace.

i.e.

Code:

find . -name '*.doc' -type f -print0 | xargs -0 grep -li "Hesketh"

The "-type f" tells find to only locate files (not directories). I also quoted the pattern '*.doc' to prevent it being pre-expanded by the shell in the case where there are one or more files matching that glob pattern in the present working directory.

H_TeXMeX_H · 08-27-2008, 03:52 PM

I would do:

Code:

find . -iname "*hesketh*.doc"

luusac · 08-27-2008, 04:07 PM

many thanks to you both, I will try these.
lu

matthewg42 · 08-27-2008, 06:36 PM

Quote:

Originally Posted by H_TeXMeX_H

I would do:

Code:

find . -iname "*hesketh*.doc"

But that just searches the file names, not the contents. Luusac please clarify which you want - files with "hesketh" in the filename, or files with the string "hesketh" contained within the file.

colucix · 08-28-2008, 02:56 AM

Code:

find . -name *.doc | grep -lir "Hesketh" *

Moreover, you obtained PPM files because you added an extra * as argument to the grep command, so that it did not parsed the piped output but all the items in the current directory.

I would add a little tip to the suggestion by matthewg42. If you want to look for either .doc or .Doc files you can use -iname to do a case insensitive search:

Code:

find . -iname '*.doc' -type f -print0 | xargs -0 grep -li "Hesketh"

H_TeXMeX_H · 08-28-2008, 03:14 AM

Quote:

Originally Posted by matthewg42

But that just searches the file names, not the contents. Luusac please clarify which you want - files with "hesketh" in the filename, or files with the string "hesketh" contained within the file.

Oh, I see, if he wants that then I agree with colucix on this solution:

Code:

find . -iname '*.doc' -type f -print0 | xargs -0 grep -li "Hesketh"

luusac · 08-28-2008, 11:00 AM

Quote:

Originally Posted by matthewg42

please clarify which you want - files with "hesketh" in the filename, or files with the string "hesketh" contained within the file.

files with extension matching .doc (case insensitive) which contain the string "Hesketh" (which can also be case insensitive) within the file (not filename)
lu

colucix · 08-28-2008, 11:07 AM

Quote:

Originally Posted by luusac

files with extension matching .doc (case insensitive) which contain the string "Hesketh" (which can also be case insensitive) within the file (not filename)
lu

So you should already have got the solution.