How to replace a string found with awk?

Question

I find the string I want to replace this way:

awk '/%block lattice_cart/{flag=1;next}/%endblock lattice_cart/{flag=0}flag {print $1}' pacet.cell | awk 'NR==1'

How can I replace the value found, for instance: "5 or any number in entry (1,1)" with "3"

Input:

%block lattice_cart
any_number   any_number  any_number
any_number   any_number  any_number
any_number   any_number  any_number
%endblock lattice_cart

Expected output:

%block lattice_cart
3.00000   any_number  any_number
any_number   any_number  any_number
any_number   any_number  any_number
%endblock lattice_cart

Post some sample text and show how you like to output to be. It looks like you get data between start/stop pattern, and the just select one line `'NR==1'`, strange. — Jotne, Jul 23 '19 at 11:45
Yes I want to choose the individual entries of the "matrix" or grid — Caterina, Jul 23 '19 at 11:48
The example is just for the first one, but I have to modify all the diagonal entries for some other value — Caterina, Jul 23 '19 at 11:49
This may help you https://stackoverflow.com/a/57078963/620097 . Good luck. — shellter, Jul 23 '19 at 13:38
wrt readability of your code, naming a flag variable `flag` is as useful as naming a numeric variable `number` instead of `count`, `sum`, `total`, `average`, `max`, or whatever it represents. Don't name variables based on the type of variable they are, name them based on what they represent. In this case at a minimum you could name it `found` or even better `inBlock` or similar. — Ed Morton, Jul 23 '19 at 14:25

RavinderSingh13 · Answer 1 · 2019-07-23T12:29:15.710

1

Could you please try following.

awk '$1==5{$1=sprintf("%.5f",3)} 1'  Input_file

EDIT: Generic one.

awk -v row=2 -v field=1 -v point="4" -v new_value=3 'FNR==row{if($field~/^[0-9]+/){$field=sprintf("%."point"f",new_value)}} 1'  Input_file

edited Jul 23 '19 at 12:29

answered Jul 23 '19 at 11:50

RavinderSingh13

130,504
14
57
93

This works, but I wanted to know what is ==5 doing? – Caterina Jul 23 '19 at 11:56
@Caterina, It is comparing if 1st field of any line is 5(as an value). – RavinderSingh13 Jul 23 '19 at 11:57
Ok then I guess this is not as general as I need it to be – Caterina Jul 23 '19 at 12:02
@Caterina, please provide example for generic one here. – RavinderSingh13 Jul 23 '19 at 12:12
I changed it. I have put "any_number" because I don't know those numbers beforehand – Caterina Jul 23 '19 at 12:14
@Caterina, please check my EDIT now and let me know? – RavinderSingh13 Jul 23 '19 at 12:22
Unfortunately the thing is that I don't know the old value. – Caterina Jul 23 '19 at 12:22
@Caterina, ok np, check my EDIT now? – RavinderSingh13 Jul 23 '19 at 12:23
It is printing 3.0000 at the top and then the whole file but without any modification. And the new edit doesn't work either :( – Caterina Jul 23 '19 at 12:27
@Caterina, I added 1 condition, check now? – RavinderSingh13 Jul 23 '19 at 12:31
It doesn't print the number at the top anymore but still doesn't work. Thanks for helping me btw :) – Caterina Jul 23 '19 at 12:33
@Caterina, It is working fine for me, change variables row, field etc as per need and see the magic :) – RavinderSingh13 Jul 23 '19 at 12:34
Ok so it works if I delete everything else in the file. Is it important to say that this are not the only numbers in the file? – Caterina Jul 23 '19 at 12:36
above and below I have more. That's why I tried to look only for those in a particular block as I specified in the question – Caterina Jul 23 '19 at 12:37
1

pluse-uno for tech support! (and a working solution, in progress!). – shellter Jul 23 '19 at 13:48

score 1 · Answer 2 · answered Jul 23 '19 at 11:52

1

This should give the output you posted from the input you posted.

awk '/%block lattice_cart/{f=1} f {gsub("5.","3.");print} /%endblock lattice_cart/{f=0}' file
%block lattice_cart
3.00000   0.00000   0.00000
0.00000   3.00000   0.00000
0.00000   0.00000   2.3.000
%endblock lattice_cart

It will only edit value between start/stop and print all lines including start/stop.

answered Jul 23 '19 at 11:52

Jotne

40,548
12
51
55

Yes but those values are arbitrary. I want this to be able to search for the first entry in general. Not only if it is equal to 5. – Caterina Jul 23 '19 at 11:55
@Caterina, you mean all first ids? Or some please be more clear on this one. – RavinderSingh13 Jul 23 '19 at 11:56
@RavinderSingh13 I mean I want to be able to change each value in this matrix regardless it's initial value. So for instance if I want to change entry (1,1) to 3 it shouldn't matter if it's initial value is 5.000 or 8.000 etc – Caterina Jul 23 '19 at 11:58

Ed Morton · Accepted Answer · 2019-07-23T15:10:50.107

$ cat tst.awk
/%block lattice_cart/ { inBlock=1 }
inBlock {
    if ( numLines++ == row ) {
        $col = val
    }
    block = block $0 ORS
    if ( /%endblock lattice_cart/ ) {
        printf "%s", block
        inBlock = numLines = 0
    }
}

$ awk -v row=1 -v col=1 -v val=3 -f tst.awk file
%block lattice_cart
3 any_number any_number
any_number   any_number  any_number
any_number   any_number  any_number
%endblock lattice_cart

$ awk -v row=3 -v col=2 -v val=27 -f tst.awk file
%block lattice_cart
any_number   any_number  any_number
any_number   any_number  any_number
any_number 27 any_number
%endblock lattice_cart

score 0 · Answer 4 · answered Jul 23 '19 at 13:05

Seems that

awk '/%block lattice_cart/{print;getline;$1="3"}1' pacet.cell

Does the job. And for the other entries I just add more getlines or change to the next column.

awk '/%block lattice_cart/{print;getline;print;getline;$2="3"}1' pacet.cell

It is not the cleanest way perhaps but it does the job.

How to replace a string found with awk?

4 Answers4