Newbie: awk and the Field Separator Variable

[goldenradium2001]

Any help will be appreciated. I'm learning awk right now and wanted to clear up a few things.

I have a file that looks like this (the <TAB> actually indicates that I'm using Tab spaces):

one<TAB>two<TAB><TAB>three

I issue the following command to put a dash between these columns:

nawk '{print $1&quot;-&quot;$2&quot;-&quot;$3&quot;-&quot;$4}' FS=&quot;\t&quot; filename

The output that I get is:

one-two--three

My question is: Why did I have to specify four fields in order to get this output? Is the space between the two dashes -- actually a field?

When I specify only three fields using the following command:

nawk '{print $1&quot;-&quot;$2&quot;-&quot;$3}' FS=&quot;\t&quot; filename

I get the following output:

one-two-

Is this proof that there exists a field between the two -- dashes?

I know I can use:

nawk '{print $1&quot;-&quot;$2&quot;-&quot;$3}' FS=&quot;\t+&quot; filename

to get around this dilemma of having two dashes but the reason why I'm writing is because I want to know WHY the space between the two dashes -- is considered a field?

Thanks for enlightening me!!! Any guesses are welcome!

 

[goldenradium2001]

All right, I'm going to answer my own question. Yes, if there two instances of a field separator then there really is a field between the two field separator however it's an empty string.

Hope this helps someone.

 

[dchoulette]

The field separator is a regular expression. It could match several characters. So if you want all consecutive <TAB>s to count as only separator, this should work:

FS="\t+"