Help with a sed command

[jamefarm]

Hello, I have a text file that is tab delimited with 6 fields, what I need to do is replace the final field. Here's a sample of the text file:



NAVI Navisite -0.08 -0.08 -0.09 11-Mar
NEN New England Realty null null null 11-Mar AMC
NPSP NPS Pharmaceuticals -0.13 -0.14 -0.18 11-Mar 4:30 PM


I need to replace the date field with a variable, but lets just use 2010-03-11 for now. The beginning of the field will begin with either 1 or 2 numbers followed by a dash etc... e.g. 9-Mar, or 11-Mar AMC

Here's what I tried, though it didn't work.


sed 's/[0-9]+(?=-.)/2010-03-11/g'

[PmDematagoda]

This seems to work:-

sed 's/[0-9]*-[A-Z][a-z][a-z]/11-03-2010/'

[jamefarm]

Just about, but I need it to remove anything after that as well, such as the AMC or the time so that the last field is only '2010-03-11'

[PmDematagoda]

Then this may do the trick:-

sed 's/[0-9][0-9]*-[A-Z a-z 0-9 :]*/11-03-2010/'

Edit:- Whoops, didn't see that the time was also there at times.

[jamefarm]

Works like a charm! Thanks a bunch PmDematagoda!

[DaithiF]

another couple of variations:

~/tmp$ sed 's/[^\t]*$/11-03-2010/' testfile
NAVI Navisite -0.08 -0.08 -0.09 11-03-2010
NEN New England Realty null null null 11-03-2010
NPSP NPS Pharmaceuticals -0.13 -0.14 -0.18 11-03-2010
~/tmp$ cut -f1-5 testfile | sed 's/$/\t11-03-010/'
NAVI Navisite -0.08 -0.08 -0.09 11-03-010
NEN New England Realty null null null 11-03-010
NPSP NPS Pharmaceuticals -0.13 -0.14 -0.18 11-03-010