Awk or Bash, Replacing string from specific column and line number to foreward

Question

I want to substituted capital letter B with C in column 5 and from line 6 to the end of the file, need to keep the spaces as it is from my original input file as it is.

ATOM   1939  HG2 PRO A 125      35.681  32.906  38.437  1.00 43.59           H  
ATOM   1940  HG3 PRO A 125      34.593  33.765  37.652  1.00 41.79           H  
ATOM   1941  HD2 PRO A 125      37.364  34.075  37.624  1.00 43.38           H  
ATOM   1942  HD3 PRO A 125      36.333  34.312  36.415  1.00 41.29           H  
TER   
ATOM   1944  N   MET B  11      16.583  29.975  -4.306  1.00 51.32           N  
ATOM   1945  CA  MET B  11      15.542  30.263  -3.327  1.00 39.92           C  
ATOM   1946  C   MET B  11      16.146  30.366  -1.933  1.00 32.50           C

I have read:

and my attempt is: awk 'NR == 6 && $ == 5, { sub(" B ", " C ") }'

`NR == 6` should be `NR >= 6`. What do you expect `$ == 5` to mean? `$` has to be followed by a number to refer to the contents of that column. — Barmar, Aug 17 '22 at 02:05

RavinderSingh13 · Accepted Answer · 2022-08-17T02:26:26.813

4

This simple awk should help you in same. Written and tested in GNU awk.

awk '
FNR>=6 && match($0,/^(\S+[[:space:]]+)(\S+[[:space:]]+)(\S+[[:space:]]+)(\S+[[:space:]]+)(\S+)(.*)$/,arr) && arr[5]=="B"{
  $0=arr[1] arr[2] arr[3] arr[4] "C" arr[6]
}
1
'  Input_file

Using match function here to keep your spaces as it is even after substitution.

edited Aug 17 '22 at 02:26

answered Aug 17 '22 at 02:08

RavinderSingh13

130,504
14
57
93

Thank you. But, how can I specify the condition with column 5? – Another.Chemist Aug 17 '22 at 02:09
@Another.Chemist, yes please we need to give it in `sub` function itself. – RavinderSingh13 Aug 17 '22 at 02:10
@Another.Chemist, Please try my updated code once, this will keep the spaces as it is, please do test it and let me know how it goes. – RavinderSingh13 Aug 17 '22 at 02:22
I have tested and it promps me: `awk: line 2: syntax error at or near ,` – Another.Chemist Aug 17 '22 at 02:24
@Another.Chemist, It works fine for me and tested it successfully. Could you please make sure its properly copied, I re-tested it now and it works fine. – RavinderSingh13 Aug 17 '22 at 02:25
@Another.Chemist, also this requires GNU `awk`, so you can check your `awk`'s version by doing a:`awk --version` also. – RavinderSingh13 Aug 17 '22 at 02:29
it promps me `awk: not an option: --version` – Another.Chemist Aug 17 '22 at 02:31
@Another.Chemist, ohhhh may be you are using old `awk`. Do a `which awk` once and let me know. Also on another note: If possible can you update awk to latest if yes then do it to GNU `awk`. Let me know. – RavinderSingh13 Aug 17 '22 at 02:33
I have the last version of `mawk` – Another.Chemist Aug 17 '22 at 02:36
@Another.Chemist, oh ok, not sure if `mawk` supports match funcition's 3rd argument or not. Is it Sun OS you are having? – RavinderSingh13 Aug 17 '22 at 02:37
1

I installed `gawk` and it worked ! Thank you !!! – Another.Chemist Aug 17 '22 at 02:40
1

@Another.Chemist, I have updated your question a bit for keeping the spaces same thing, to make sure all are on same page, cheers. – RavinderSingh13 Aug 17 '22 at 02:41
1

I like how you write _This simple awk_ ... – kvantour Aug 17 '22 at 07:44
@kvantour, initially it was simple(this one is also simple only :) ) but then to preserve spaces/tabs etc edited it, that was kept there as it is. – RavinderSingh13 Aug 17 '22 at 07:46
@RavinderSingh13 : since you're using a `gawk`-specific solution, why not group all of `$1-$4` into 1 array cell, `$5` standalone, then rest of the line in 3rd cell ? – RARE Kpop Manifesto Aug 19 '22 at 14:50

score 3 · Answer 2 · answered Aug 17 '22 at 06:39

You can use any awk and preserve the format by using a sub() on the entire record. This prevents awk from recalculating the fields. For instance, you can do:

awk 'NR > 5 { sub(/MET B/,"MET C") }1' file

Here awk will replace the first occurrence of "MET B" with "MET C" beginning with the 6th record leaving the spacing alone.

Output

ATOM   1939  HG2 PRO A 125      35.681  32.906  38.437  1.00 43.59           H
ATOM   1940  HG3 PRO A 125      34.593  33.765  37.652  1.00 41.79           H
ATOM   1941  HD2 PRO A 125      37.364  34.075  37.624  1.00 43.38           H
ATOM   1942  HD3 PRO A 125      36.333  34.312  36.415  1.00 41.29           H
TER
ATOM   1944  N   MET C  11      16.583  29.975  -4.306  1.00 51.32           N
ATOM   1945  CA  MET C  11      15.542  30.263  -3.327  1.00 39.92           C
ATOM   1946  C   MET C  11      16.146  30.366  -1.933  1.00 32.50           C

score 1 · Answer 3 · answered Aug 17 '22 at 02:09

1

Use NR >= 6 to update all lines starting with line 6. Your code does it on line 6 only.

Put $5 in the sub() call to make the substitution only in that field.

awk 'NR >= 6 {sub("B", "C", $5)} 1'

I hope you don't care about the number of spaces between the columns. The line that are updated will have multiple spaces collapsed into one, so

ATOM   1944  N   MET B  11      16.583  29.975  -4.306  1.00 51.32           N

becomes

ATOM 1944 N MET C 11 16.583 29.975 -4.306 1.00 51.32 N

answered Aug 17 '22 at 02:09

Barmar

741,623
53
500
612

Thank you. But the format is broken – Another.Chemist Aug 17 '22 at 02:13
How the format can be conserved? – Another.Chemist Aug 17 '22 at 02:18
Instead of just printing the line with `1`, you can write code that prints the formatted line with `printf()`. I'll leave that as an exercise for the reader, it's not complicated. – Barmar Aug 17 '22 at 02:22

score 0 · Answer 4 · answered Aug 17 '22 at 12:55

If you can make use of gawk, you could use split which will have a seps array that stores the separators.

You can split on the field separator FS, and then use the number returned by split to loop through all the fields.

When encountering field 5 and the value is B, then change it to C

awk 'NR > 5 {
  nr = split($0, a, FS, seps)
  for (i = 1; i <= nr; ++i) {
    if (i == 5 && a[i] == "B") a[i] = "C" 
    printf "%s%s", a[i], seps[i]
  }
  printf "\n"
  next
}1' file

Output

ATOM   1939  HG2 PRO A 125      35.681  32.906  38.437  1.00 43.59           H
ATOM   1940  HG3 PRO A 125      34.593  33.765  37.652  1.00 41.79           H
ATOM   1941  HD2 PRO A 125      37.364  34.075  37.624  1.00 43.38           H
ATOM   1942  HD3 PRO A 125      36.333  34.312  36.415  1.00 41.29           H
TER
ATOM   1944  N   MET C  11      16.583  29.975  -4.306  1.00 51.32           N
ATOM   1945  CA  MET C  11      15.542  30.263  -3.327  1.00 39.92           C
ATOM   1946  C   MET C  11      16.146  30.366  -1.933  1.00 32.50           C

score 0 · Answer 5 · answered Aug 19 '22 at 14:45

here's a way to preserve all the spaces and tabs you like without using vendor-proprietary solutions :

 mawk 'BEGIN {   _ = length(FS="[ \t]+") } NR<_ || NF<_ ||
              $!NF = sprintf("%.*sC%s", (__ = index($!_,
                             $_))- !!_, $!_, substr($!_, ++__))'

ATOM   1939  HG2 PRO A 125      35.681  32.906  38.437  1.00 43.59           H  
ATOM   1940  HG3 PRO A 125      34.593  33.765  37.652  1.00 41.79           H  
ATOM   1941  HD2 PRO A 125      37.364  34.075  37.624  1.00 43.38           H  
ATOM   1942  HD3 PRO A 125      36.333  34.312  36.415  1.00 41.29           H  
TER   
ATOM   1944  N   MET C  11      16.583  29.975  -4.306  1.00 51.32           N  
ATOM   1945  CA  MET C  11      15.542  30.263  -3.327  1.00 39.92           C  
ATOM   1946  C   MET C  11      16.146  30.366  -1.933  1.00 32.50           C

Awk or Bash, Replacing string from specific column and line number to foreward

5 Answers5