Re: Sed advice needed

[Tyler Smith <tyler.smith@mail.mcgill.ca>]

On 2007-05-31, Piers Kittel <debian@biased.org> wrote:
>
> "181","1324.014027","111.111.111.111","111.111.111.111","RTP","Payload t 
> ype=ITU-T H.261, SSRC=2008229573, Seq=54520, Time=1725612773, Mark"
> "185","1324.078941","111.111.111.111","111.111.111.111","RTP","Payload t 
> ype=ITU-T H.261, SSRC=2008229573, Seq=54521, Time=1725616276"
>
> I need to convert the above to the below:
>
> "181","1324.014027","111.111.111.111","111.111.111.111","RTP","54520"
> "185","1324.078941","111.111.111.111","111.111.111.111","RTP","54521"
>
> What's the best way to do this?  I've been reading the man pages of  
> sed, cut and awk but I can't quite figure out how to do this.  Any  
> ideas?
>

If I understand what you want, you need the fields 1-5 and 8 from a
comma-delimited record. At least, that's what cut will see - the
quotes around Payload...Mark will get ignored if you tell cut to use a
, as delimiter:

cut -d, -f1,2,3,4,5,8

Then you just need to get rid of the Seq= line, which sed will do
with:

sed s/Seq=//g

So put them together:

cut -d, -f1,2,3,4,5,8 input.file | sed s/Seq=//g

That looks almost right. I'm missing the quotes around the last field.
Two more sed commands fix this:

cut -d, -f1,2,3,4,5,8 input.file | 
    sed -e s/Seq=//g -e 's/, /,"/g' -e 's/$/"/g'

That seems to do the job with your example text.

HTH,

Tyler