Substring Match in shell

[mediator]

Hi,

 

  I'm trying to create a script which includes substring match. Consider the following test cases.

 

  List of strings - 12345678, 657895

 

 Now I need to see if substrings - 1234,12,657,65,6578 are in this above strings.

 

My problem is that the substrings - 1234,12 will both be found in 12345678. But I need to go for 1234 if it exists, rather than 12. Similarly, if 6578 exists in 657895, then it should be selected and not 65 or 657.

 

I know I can create sub-cases inside an 'if statement'. But I'm looking for an easier work around.

 

Thanx :)

[snaphat (Myles Landwehr) Member]

Use regex and jut put the longest match first (Longest|Middle|Shortest)

[mediator]

Can you please give some examples with bash shell?

[snaphat (Myles Landwehr) Member]

Something like this will return the longer match. If flip the order you'll see that it will match "te" instead of "test"

echo "test" | perl -ne 'print $1 if /(test|te)/'

Note: I'm using perl because the OR operator wasn't working with grep.

 

EDIT:
 

Here is a slightly better example with your input that will return all matches line by line instead of just the first match:

echo "12345678, 657895, 1234" | perl -nE 'say for /(6578|65|1234|12)/g'

[mediator]

Thanx a ton snapchat. First example works. The other example you gave did not work. I'm testing this on AIX 6.1 were the default shell is ksh. With second example it says -E is an unrecognised switch.

[snaphat (Myles Landwehr) Member]

Here's a slightly longer form for the second example (without using the say module -- i assume you have an older or different perl):

echo "12345678, 657895, 1234" | perl -ne 'while(/(6578|65|1234|12)/g){print "$&\n"}'

P.S.: my nick is snaphat ;-)

 

EDIT:  here is a slightly better revision:

echo "12345678, 657895, 1234" | perl -ne 'print "$&\n" while /(6578|65|1234|12)/g'

[mediator]

Thanx a lot Snaphat. I have tested your latest examples and they work. You're a life saver! :)

[mediator]

Can you please also help me with this e.g I have a file where every position of a each line represents some meaning. I need to find a string from certain position e.g 10 characters from 5th position. Let assume the complete string (one line of a file) is the below one where blank spaces are also characters. I need to see the 10  characters starting from 5th position. How can I do this?

 

abcdefgh      12345678      porp

[snaphat (Myles Landwehr) Member]

^ dunno if the spacing is correct in the following, but you are looking for something like this:

echo "abcdefgh      12345678      porp" | perl -ne 'print substr($_,5,10)'

[mediator]

Thanx a ton again Snaphat! It works. :)