Unix - the number of columns in the file

Question

For a file with data similar to this (i.e. store.dat file)

sid|storeNo|latitude|longitude 2|1|-28.03720000|153.42921670 9|2|-33.85090000|151.03274200 

What will be the command to output the number of column names?

i.e. In the above example, this will be 4. (the number of channel characters + 1 in the first line)

I thought something like:

 awk '{ FS = "|" } ; { print NF}' stores.dat 

but it returns all rows instead of the first, and for the first row it returns 1 instead of 4

Answer 1

 awk -F'|' '{print NF; exit}' stores.dat 

Just close right after the first line.

Answer 2

This is a workaround (for me: I don't use awk very often):

Display the first line of the file containing the data, replace all channels with new lines, and then count the lines:

 $ head -1 stores.dat | tr '|' '\n' | wc -l 

Answer 3

If you do not use spaces, you should use | wc -w | wc -w on the first line.

wc is a "Word Count" that simply counts the words in the input file. If you send only one row, it will tell you the number of columns.

Answer 4

You can try

cat FILE | awk '{print NF}'

Answer 5

If you have python installed, you can try:

 python -c 'import sys;f=open(sys.argv[1]);print len(f.readline().split("|"))' \ stores.dat 

Answer 6

This is usually used to count the number of fields:

 head -n 1 file.name | awk -F'|' '{print NF; exit}' 

Answer 7

Perl's solution is similar to Mat awk's:

 perl -F'\|' -lane 'print $#F+1; exit' stores.dat 

I tested this in a file with 1,000,000 columns.

---

If the field separator is a space (one or more spaces or tabs) instead of a channel:

 perl -lane 'print $#F+1; exit' stores.dat 

Answer 8

Based on Cat Kerr answer. This team is working on solaris

 awk '{print NF; exit}' stores.dat 

Answer 9

you can try:

 head -1 stores.dat | grep -o \| | wc -l 

Answer 10

select any line in the file (in the example below, this is the second line) and count the number of columns where the separator is a space:

 sed -n 2p text_file.dat | tr ' ' '\n' | wc -l