How to delete specific pattern in a file with SED?

UNIX HP-UX
1

I have one file which is having content as following...

0513468211,,,,20091208,084005,5,,2,3699310,
0206554475,,,,20090327,123634,85,,2,15615533
0206554475,,,,20090327,134431,554,,2,7246177
0103000300,,,,20090523,115501,89,,2,3869929
0736454328,,,,20091208,084005,75,,2,3699546
18656400,,,,20091208,084005,42,,2,370012
0306937734,,,,20090519,134145,411,,2,4573207,,

The second filed is the date field. I don't want the records less than may month of 2009 i.e record should less than 200905.

I am able to print this with

sed '/20091208/d' <filename>

Please suggest me way to remove record less than may month with SED.

2

Why must it be sed?...

Anyway:

sed -n '/^[^,]*,,,,20090[1-4].*/p' infile
0206554475,,,,20090327,123634,85,,2,15615533
0206554475,,,,20090327,134431,554,,2,7246177
# or
awk -F"," '$5 < 20090500 {print}' infile
0206554475,,,,20090327,123634,85,,2,15615533
0206554475,,,,20090327,134431,554,,2,7246177
3

Hi Zaxxon,
thanks for the instant response...
am able to print those record with awk, but now how do i delete it?
Once again thanks in advance

4

Ah, I had misread sorry:

awk -F"," '$5 >= 20090500 {print}' infile

Or do you mean to instantly edit the file too?

5

Yes , i want to remove all the line from that file whose date is less than 200905 i.e may 2009.
Your first query prints the result, but to delete those lines from file please suggest me way.

6

Is redirecting the output from my last code to a new file and mv'ing it to original file sufficient?

7

Hi Zaxxon,

Done ...thank you vary much for such kind support...

God bless u...Cheers...

8

Using the same example as above, is there a way to just simply delete the date field only, and not the whole line from the existing file?