Check if the file has given number of columns [duplicate]

问题

This question already has answers here:

Loop through file to count field number (3 answers)

Closed last year.

I have a few csv files that I need processing if the number of columns is 4 througout the file

file_1 :

1,aa,bb,cc
2,dd,ee,ff
3,gg,hh,ii

file_2 :

1,xx
2,yy
3,zz
4,xy

I'm using awk for this

for file in file_1 file_2 ... file_n
do
    if [[ `awk -F"," 'NF==4' $file` ]]
    then
        #process
    else
        continue
    fi
done

I'm looking for a solution that

• Doesn't involve a subshell
• Checks for number of columns throughout the file
• Performs better than the above code (if possible)

回答1:

You may use:

for f in file_1 file_2 file_n; do
   if awk -F, 'NF!=4{exit 1}' "$f"; then
      echo "good $f"
   else
      echo "bad $f"
   fi
done

This awk exits with 1 as soon it finds a record with no of columns not equal to 4.

回答2:

Solution 1st: If you want to check that any line is NOT having 4 fields and move to next file then following may help you on same.

awk -F, 'NF!=4{nextfile} 1' File_*

Solution 2nd: In case you want to skip only those lines which are NOT having 4 fields then following may help you on same.

awk -F, 'NF!=4{next} 1' File_*

来源：https://stackoverflow.com/questions/49130135/check-if-the-file-has-given-number-of-columns