Shell Programming and Scripting

I need to write a script that reads through an input .txt file and replaces the end value with the end value of the next line for lines that have distance <=4000. The first label line is not actually in the input. In the below example, 3217 is the distance from the end of the first line to the start of the second line. 14021 is the distance from the previous line (not included) to the start of the first line. So once the script finds a distance <=4000, it needs to replace the end of the previous line with the end of the current line.

Any help would be greatly appreciated! Thanks!

INPUT:

chrm start end block length distance
chr7 27398704 27399096 ENm010Block536 392 14021
chr7 27402314 27402466 ENm010Block537 152 3217
chr7 27412536 27412726 ENm010Block538 190 10069
chr7 27416032 27416424 ENm010Block539 392 3305
chr7 27420022 27420972 ENm010Block540 950 3597

Desired OUTPUT:

chr7 27398704 27402466
chr7 27412536 27420972

If I understand correctly the output should be:

Code:

chr7 27398704 27402466
chr7 27412536 27416424

Am I missing something?

No, actually I meant to put the original
27420972

because the next distance is <=4000 as well, so those two would get merged as well. See, when you have several distances<=4000 consecutively, you continue to merge them, until the distance is no longer <=4000.

Could you post a bigger input sample with the desired input?

Sure! This may take me a little while since I'm doing it manually, but it should be up in about 15 minutes. Thanks for your interest!

Hm,
with this code (use nawk or /usr/xpg4/bin/awk on Solaris):

Code:

awk 'END { print _, __ } 
1 == NR || $NF >= 1000 {
  if (c) print _, __ 
  _ = $1 FS $2
  c = 1
  }  
{ __ = $3 }' file

I get this output:

Code:

chr7 27104483 27105154
chr7 27106872 27110789
chr7 27111956 27112830
chr7 27114388 27125180
chr7 27126966 27131260
chr7 27135440 27137796

Do you realy want to treat the last line as in the example output?

It makes the code a bit ugly:

Code:

awk 'END { print _, ___ RS ____ } 
1 == NR || $NF >= 1000 {
  if (c) print _, __ 
  _ = $1 FS $2
  c = 1 
  }  
{ ___ = __
  __ = $3 
  ____ = $1 FS $2 FS $3 }
' file

You may need to check if the last line has $NF >= 1000,
if that matters, I should add more code.

And ..., don't blame me for choosing such variable names,
if you don't like them, just change them