Finding line numbers of blank lines

[steve]

Sometimes I have files which are double-spaced and I use

awk 'length > 2' file

to strip away the blank lines.

Now I have some blank lines within a file and I would like to know the
line numbers on which they occur. I thought that I could do something
like:

awk '{length < 5} print NR' file

However, that gives an error.

Any suggestions would be welcome.

Thanks, Steve

[steve]

Guess I should have continued googling first:

awk 'BEGIN {x=1}; {if (length < 3) print x; x++}' file

Steve

[Janis Papanagnou]

On 17.04.2013 23:48, steve wrote:
> Sometimes I have files which are double-spaced and I use
>
> awk 'length > 2' file
>
> to strip away the blank lines.

awk NF file

will strip away all empty (and blank) lines.

>
> Now I have some blank lines within a file and I would like to know the
> line numbers on which they occur. I thought that I could do something
> like:
>
> awk '{length < 5} print NR' file

You misplaced the brackets. Try

awk 'length < 5 {print NR}' file

But (as above) you may also want to try

awk '!NF {print NR}' file

(The variable NF provides the number of fields.)

Janis

[Janis Papanagnou]

On 17.04.2013 23:57, steve wrote:
> On Wed, 17 Apr 2013 21:48:13 +0000, steve wrote:
>
>> Sometimes I have files which are double-spaced and I use
>>
>> awk 'length > 2' file
>>
>> to strip away the blank lines.
>>
>> Now I have some blank lines within a file and I would like to know the
>> line numbers on which they occur. I thought that I could do something
>> like:
>>
>> awk '{length < 5} print NR' file
>>
>> However, that gives an error.
>>
>> Any suggestions would be welcome.
>>
>> Thanks, Steve
>
> Guess I should have continued googling first:

No, you rather should have avoided the web source where you found that code.

>
> awk 'BEGIN {x=1}; {if (length < 3) print x; x++}' file

That's a very clumsy way to do what you want. (See my other post.)

Janis

>
> Steve
>

[steve]

awk '!NF {print NR}' file worked properly. The 2nd suggestion didn't
return anything.

Thanks, Janis

[steve]

Correction:
awk 'length < 5 {print NR}' file worked properly.

awk '!NF {print NR}' file didn't return anything.

Steve

[Janis Papanagnou]

Which "2nd"? - You mean this one...? awk 'length < 5 {print NR}' file
Works for me.

If your awk's version doesn't support the length function without
arguments - but then your own try should as well not work - then try:
awk 'length($0) < 5 {print NR}' file
But I suggest not to compare to some arbitrary value 5 instead of any
real empty-test. But instead of "< 5" you could also just use
!length($0) {print NR}
Though, if you're operating in a WinDOS environment some spurious CR
(^M) might be considered a character and length return something >0.
(But I have no WinDOS to check that.)

Janis

>
> Thanks, Janis
>

[Janis Papanagnou]

Maybe spurious ^M in your file that are counted as part of field data?
Are the lines in your file probably CRLF terminated?

od -c file

will show you.

Janis

>
> Steve
>

[nag]

awk 'NF<1{print NR}'

[Janis Papanagnou]

On 18.04.2013 03:14, nag wrote:
> On Thursday, 18 April 2013 03:48:32 UTC+5:30, steve wrote:

[...]

>>
>>
>> awk '!NF {print NR}' file didn't return anything.
>>
>

> awk 'NF<1{print NR}'

NF can not become negative, so NF<1 is not different from NF<=0, NF==0,
or just !NF.

Janis

>