[SOLVED] How can I extract strings based on character positions?

btacuso · 03-21-2010, 08:47 AM

I have a file with no regular way to grep or sed for any regular expression. The wordings, fields, spacings are random.

myfile:
asdjfvj jd sdfadjf dfafas asfafsaf
fdgjldkfg sdgfdg sfjhgfds h hgfh s sh fd
fj vv kzvh lkzv vzlb ffa.

I want to extract character positions 3 to 10 including spaces.

desired result:
djfvj j
gjldkfg
vv kz

Thank you very much.

troop · 03-21-2010, 09:03 AM

Code:

# cat myfile | awk '{print substr($0,3,7)}'
djfvj j
gjldkfg
 vv kzv

H_TeXMeX_H · 03-21-2010, 09:38 AM

Cut works too:

Code:

bash-3.1$ cut -c 3-9 test
djfvj j
gjldkfg
 vv kzv

Or

Code:

bash-3.1$ cut -c 3-10 test
djfvj jd
gjldkfg 
 vv kzvh

whichever characters you need, it looks like your example is wrong.

Tinkster · 03-21-2010, 01:14 PM

And while 'cut' is certainly the most sensible tool for this
job I'd like to add a sed example :}

Code:

$    echo "asdjfvj jd sdfadjf dfafas asfafsaf
> > fdgjldkfg sdgfdg sfjhgfds h hgfh s sh fd
> > fj vv kzvh lkzv vzlb ffa." | sed -r 's/^...(.{7}).*/\1/'
jfvj jd
dgjldkf
j vv kz

Cheers,
Tink

btacuso · 03-21-2010, 01:20 PM

Quote:

Originally Posted by H_TeXMeX_H

Cut works too:

Code:

bash-3.1$ cut -c 3-9 test
djfvj j
gjldkfg
 vv kzv

Or

Code:

bash-3.1$ cut -c 3-10 test
djfvj jd
gjldkfg 
 vv kzvh

whichever characters you need, it looks like your example is wrong.

I was about to put a note about it. The example is moved automatically leftwards when I submitted it, but all you guys gave me the answer. Thanks again. BTW how do I thank each of you using the option by this forum? I tried and still it showed I thanked nobody.

btacuso · 03-21-2010, 01:31 PM

Quote:

Originally Posted by troop

Code:

# cat myfile | awk '{print substr($0,3,7)}'
djfvj j
gjldkfg
 vv kzv

I tested all the posts. It all worked but this seems to be the one I have to use. Thanks again.

Tinkster · 03-21-2010, 01:38 PM

Quote:

Originally Posted by btacuso

I tested all the posts. It all worked but this seems to be the one I have to use. Thanks again.

I'm curious - WHY? :}

In terms of execution speeds/memory sizes:
cut > sed > awk

Why did you pick the slowest of the tools?

Cheers,
Tink

btacuso · 03-25-2010, 09:22 AM

Quote:

Originally Posted by Tinkster

I'm curious - WHY? :}

In terms of execution speeds/memory sizes:
cut > sed > awk

Why did you pick the slowest of the tools?

Cheers,
Tink

Sorry for not responding sooner. I picked the slowest one because it is in awk. I started my script in awk and I prefer not to mix other commands unless necessary. The speed is not noticeable though because I have only about 200-300 lines of input.

Tinkster · 03-25-2010, 12:31 PM

Fair enough, and no worries. Thanks for coming back!