awk 循环打印错误的次数

Question

I have a file like this (space tab):

Agent 299301 1
Person 259672 2
Place 208239 3
Location 208239 4
PopulatedPlace 156701 5
Region 153246 6
AdministrativeRegion 153246 7
Work 96536 8
Agent 299301 1
Person 259672 2

I want to return a different number for each row as many times as it appears in the second column

For example: first, return number 1 299301 times, then return 2 259672 times, then return 3 208239 times

For that I am using this awk command:

cat file | awk -F ' ' '{for (i=1; i<=$2; i++) print NR}'  > output

It seems to work well with small numbers on second column but in this file sample, I don't know why is returning each number ( 1 in this case) the incorrect number of times:

It's returning number 1 558973 times instead of 299301 times

But it returns the correct number in the rest of lines of the file (numbers 2, 3, 4... )

And if I add more lines to the file, it also returns the wrong number of times with numbers 2, 3, 4 until number 9 , but then it also works well with numbers 10, 11, 12, 13...

So I don't know why is this happening, hope you can help me with this.

Thanks in advance.

Answer 1

It's not your script that's wrong, it's how you're trying to validate it's output. You're piping the output to grep '1' | wc -l grep '1' | wc -l or similar and so are counting the number of 1 s ( 299301 ) + the number of 10 s ( 259672 ) and so getting the total 558973 .

$ awk '{for (i=1; i<=$2; i++) print NR}' file | grep '1' | wc -l
558973

$ awk '{for (i=1; i<=$2; i++) print NR}' file | grep '^1$' | wc -l
299301

By the way, cat file and -F ' ' aren't doing any real harm but they also do nothing useful in this case, just use awk '{for (i=1; i<=$2; i++) print NR}' file instead as I did above.

Answer 2

When fed the input presented in the question, the program presented in the question outputs '1' exactly 299301 times for me, as you expected. I am inclined to suppose that @choroba's comment on the question is indicative of the nature of a different program with which you observed instead 558973 lines of '1'. That is, the one in this, or something substantially equivalent:

cat file | awk -F ' ' '{for (i=1; i<=$2; i++) print $3}'  > output

The difference is that one prints NR , the input line number, whereas the other prints $3 , the third field read from the input line.