Search Results

What's your system? What's your shell? This will be a lot harder in some shells than others, most can't directly turn a character into an integer value like that.

If perl is available I'd...

oops wrong post--snipped

This splits every sentence.
awk -v RS="." '!/^\n$/ { print "<p>" $0 ".</p>" }' < input > output

This splits every 5 sentences.

awk -v RS="." '!/^\n$/ { if(!((++N)%5)) printf("</p>\n<p>");...

And why are you trying to get all the lines from your while loop into newvar1?

What are you actually trying to do? What is the function of that awk, there?

I am not diverting you or trying to...

And why are you trying to get all the lines from your while loop into newvar1?

What are you actually trying to do?

There may be more direct ways, is what I am thinking, if only I knew what you...

It's because of the pipe.

The while loop executes inside a subshell, and doesn't affect variables in the shell outside it.

What are you actually trying to do?

An alternative:
awk '{gsub(/.{100}/,"&^");n=split($0,a,"^");for(i=1;i<=n;i++) print a[i]}' inputfile

Assuming that "^" character does not appear in the input records.

If that's what you want, why not use the fold command? fold -w 100 -s

You could try this:

awk 'gsub(/^.*<div class="t0">/,"") && gsub(/<\/div>.*$/,"")'

Hi,

Try gawk instead of awk.
OS X has some different construct in awk. Just make try with gawk.
Cheers,
Ranga:-)

You have to use gensub for grouping.
This will works fine with GNU awk.
What is your OS?

Hi,

Try this one,

awk '{print gensub(/.*<div class="t0">(.*)<\/div>.*/,"\\1","g",$0);}' file

Cheers,
Ranga:-)

counts=${#myline}
if [ $counts -gt 20 ]
then
echo "yes"
else
echo "NO"
fi

That if is outside the scope of the awk script...
and I feel all of this can be done by a single awk command. But, I do not clearly know your intention/purpose.

That regular expression won't work as desired even in GNU sed. *? in that regular expression is intended to be a non-greedy match. GNU sed regular expressions (basic and extended) do not support that...

Hi,

Try this one,

echo ${myline} | perl -ne '$t=$_;$c=$t=~s/[^ ]\s+[^ ]//g;print "Words:$c\nLetters:",length($_),"\n";'

Cheers,
Ranga:-)

${#myline} will give you the length of myline.

wc can count letters, words, and lines.

Regards,
Alister