awk script

[Zafar A. Mohammed]

Hi,
I have a file and each line is greater than 3000 bytes. I have to search the word and count and print it. But the problem with awk I cannot do becuase of its 3000 bytes restriction.
Is someone have any alternative inputs.
Thanks
Zafar

[A. Clay Stephenson]

Hi:

Your best bet is to download GNU's versions of awk - 'gawk'. It is available from any of the HP-UX Porting Centres and has dynamically allocated maximum record size that can be set from the command line.

Clay

[Zafar A. Mohammed]

If was thinking of gawk; but I am not sure about that. If you the link for gawk download, I'll appreciate it.

[harry d brown jr]

THis site shows a benchmark of quite a few programs, and you can download the source code.

http://www.bagley.org/~doug/shootout/bench/wordfreq/

[Wodisch]

Hello Zafar,

well, you can still substitute something in those long
lines with a NEWLINE using something like "tr(1)" and
then pipe that to "awk(1)" to find and count it:

tr "\040\t" "\n" YOUR-FILE-HERE | awk ...

This example would split your file at every space and
tab into a new line, hence getting around the line length
restriction...

Of course, you max add more "split-characters" into the
first parameter of "tr(1)"!

HTH,
Wodisch

[Patrick Wallek]

Here is a link to download gawk:

http://hpux.connect.org.uk/hppd/hpux/Gnu/gawk-3.1.0/

[R. Allan Hicks]

I may be missing the point of the question....

Is all you want the word count? If so, will wc work?

[harry d brown jr]

I think Zafar is looking for something that will find a word and count the number of times it shows up.