nixCraft

#1 (**permalink**) 09-08-2005, 02:54 AM

Is it possible to find a file I have written in Open office wordprocessor+ I have Mandrake Linux 10.0

The following does find files only written in the editor vi.
$find ./ -name '' name of the file'' -print

#2 (**permalink**) 09-08-2005, 03:38 AM

i opened my word processor and created a file tt. on saving it by default it came with a .abw extension.
then i used

Code:

find / -name tt* -type f -name

i got many files.........i scrolled up to find my file in /home/user/tt.abw

then i changed the command to

Code:

find /home -name tt* -type f -name

it just came up with two file including the tt.abw
does this help u?

monk · #3 (**permalink**) 09-08-2005, 06:13 PM

Well kavi is right, however you should try following too:
Open Office file extension is SXW so to find out all .sxw file command would be:

Code:

find / -iname "*.sxw"

Hey kavi, you are doing good job why don't u create account?

#4 (**permalink**) 09-09-2005, 02:00 AM

Thanks everybody

It worked fine.

I want to use the ' grep ' command to fine a string in one of the files written in Open Office.
It worked fine on all the files I wrote using vi editor.

Look at the following file.

/home/ka/Documents/Cisco questions and answers.sxw

I want to use the grep command to find a word like 'for' or ' of ' in the above file. Is it possible?

It worked fine for the files written in vi editor.

Your help is deeply appreciated.

sweta · #5 (**permalink**) 09-09-2005, 06:11 PM

Well I think grep works with text file only and not binary data files. May be try out desktop search provided by KDE or Beagle desktop Searching tool.

#6 (**permalink**) 09-09-2005, 10:40 PM

Thanks sweta for the reply.

Even the files which has the 'html' endings could be traced by grep family of commands. Look at the following:

[ka@c83-250-89-72 ka]$ fgrep title myfile1.html
<title>My homepage</title>
[ka@c83-250-89-72 ka]$

[ I looked for the word 'title' in a file called ' myfile.html ' .]

I just want to know the problem here. Why can't it penetrate a letter written in the wordprocessor, OpenOffice?

monk · #7 (**permalink**) 09-10-2005, 02:58 AM

Quote:

Originally Posted by Gorky

I just want to know the problem here. Why can't it penetrate a letter written in the wordprocessor, OpenOffice?

Short answer grep can't find the text in binary format. OpenOffice saves file in binary format for optimization and saving of disk space. You can save as your files as html or even text format so that you can do a search using grep.

Long answer ... Okay let us try to understand OpenOffice file format. First it uses XML to store everything and your file divided into different files. But you always see one file called foo.sxw. This file is basically nothing but zip file containing all other .XML files. So just use unzip command to open them, for example if your file is foo.sxw:

Code:

mkdir /tmp/test
cp foo.sxw /tmp/test
cd /tmp/test
unzip foo.sxw

And you should 3-4 different xml file. Next content.xml contains your actual data written in text format as well as XML codes. So just grep into it:

grep -iE “text-to-search” content.xml

In other words you can find what you wanna if you automate this procedure (unzip file and grep it etc) via shell or perl script you can do the job.

kavi · #8 (**permalink**) 09-10-2005, 07:03 AM

Quote:

Originally Posted by monk

Well kavi is right, however you should try following too:
Open Office file extension is SXW so to find out all .sxw file command would be:

Code:

find / -iname "*.sxw"

Hey kavi, you are doing good job why don't u create account?

monk, i did create an account. hope u recognized me. I am the guest with 'content of the file' problem.
it was just ur inspiration that i learnt a bit of how things happen with shell. thanx man.....