find string in line

hi all!

I'm writing a bash script and I want to find how many times a specific string appears in every line of a file.
The string I'm looking for is <grow
The string might be a substring of another string such as:
old#<grown or young><growup
Is there a way to achieve that?

Something suitably cryptic...
$ perl -n -e '$n = ( $_ =~ s/grow//g );print "$n\n";'
old#<grown or young><growup
2
grow grow grow
3

Probably less bullet proof than Salem's, but less cryptic :)

grep -w grow /tmp/test |wc -w

where /tmp/test is:
grrow grow grow
old#<grown or young><growup

Quote:

---

Originally Posted by Salem

Something suitably cryptic...
$ perl -n -e '$n = ( $_ =~ s/grow//g );print "$n\n";'
old#<grown or young><growup
2
grow grow grow
3

---

Thank you for your answer.I have a few questions..
Is this suitable for a bash script?
Also I only want the substring <grow
If there's a mere grow without the < in the front I dont want to count it.

How does you script looks like at the moment? Did you try it?
Now if what you really want is to match the pattern /<grow/ then use this pattern instead of /grow/
If you use grep, you would have to remove the -w switch.

Quote:

---

Originally Posted by quo

Thank you for your answer.I have a few questions..
Is this suitable for a bash script?

---

It is in the sense that perl is standard most (but not all) places bash is standard. It isn't in the sense that some people might call it heavy-handed.

This is a fairly easy thing to do if you break the task into parts (which are hard to spot in one liners), and there are oodles of ways to do it. My advice is that you write a simple script to experiment with and start reading about the use of grep and sed. Might take you a bit longer than if you just cut n' paste someone else's code, but OTOH, next time you want to do something like this (it's also a fairly common kind of task) you'll have a better idea of how to go about it.

Quote:

---

Originally Posted by mariostg

How does you script looks like at the moment? Did you try it?
Now if what you really want is to match the pattern /<grow/ then use this pattern instead of /grow/
If you use grep, you would have to remove the -w switch.

---

It's just a test script so all it does is find the pattern /<grow/ in a file

Code:

---


#!/bin/bash
 
grep '\<grow' ./test | wc -w

---

for the following data(./test)

<grow fffffff uuuuuuuu eeeeeeee<growup
ddddddd iiiiiiii rrrrrrrr<growold <grow_hhhh grow
grow<s yyyyyy vvvvvyoung><growgrow ccccc bbbbbbbold#<growup

it should return 6 as the accepted are:


<grow fffffff uuuuuuuu eeeeeeee<growup
ddddddd iiiiiiii rrrrrrrr<growold <grow_hhhh grow
grow<s yyyyyy vvvvvyoung><growgrow ccccc bbbbbbbold#<growup

but it returns 14(the number of all the words)

Quote:

---

Originally Posted by quo

but it returns 14(the number of all the words)

---

Right, because:

Quote:

---

Originally Posted by man grep

grep searches the named input FILEs (or standard input if no files are named, or if a single hyphen-minus (-) is given as file name) for lines containing a match to the given PATTERN. By default, grep prints the matching lines.

---

If every line contains the pattern, grep will just output the entire file. Of course, you would notice this if you simply tried it on the command line ;). Fortunately:

Quote:

---

Originally Posted by man grep

-o, --only-matching
Print only the matched (non-empty) parts of a matching line, with each such part on a separate output line.

---

Which is why I suggested you read the manual, etc and do a little hands on experimenting. If you spend a little time learning to fish (anyone can do it), you'll save yourself time running around asking "Where are the fish?". Honestly. :cool:

Quote:

---

Originally Posted by MK27

Which is why I suggested you read the manual, etc and do a little hands on experimenting. If you spend a little time learning to fish (anyone can do it), you'll save yourself time running around asking "Where are the fish?". Honestly. :cool:

---

Yes you're right!Still trying to learn scripting.Thanks!:redface:

This can be done in "pure bash".

Code:

---

while read line ;do
  cnt=0
  while [ 1 ] ;do
    len=${#line}          # length of string
    line=${line/<grow//}  # substitute
    # if no change in length, break loop
    if [ ${#line} -eq $len ] ;then break ;fi
    let cnt++
  done
  let line_num++
  echo "Line $line_num: $cnt"
  let total=total+cnt
done
echo "Total: $total"

---