Hi,

I have a sequence which has 30000 strings which looks like this

>string2991 234445
>string224 470561
>string121 675386
>string4098 177229
>string8049 255838
>string8 672382
>string1115 578415

>string8 672382
>string121 675386
>string224 470561
>string1115 578415
>string2991 234445
>string4098 177229
>string8049 255838

try:

awk '{a=$0;sub("^[^0-9]*","",a);b[a$1]=a":::"$0} END{for (i in b) print b[i]}' infile | sort -n | sed 's/.*::://'

sed 's/^\([^0-9]*\)\([0-9]*\)/\2 :::\1\2/' infile | sort -n | sed 's/.*::://'

It worked!!

---------- Post updated at 02:55 PM ---------- Previous update was at 02:54 PM ----------

Thanks very much!! It worked!!!

also, another awk example without the need to pipe to sort and sed:

awk '{a=$1;sub("^[^0-9]*","",a);b[a]=$0; m?(0):(m=a); if (a<m) m=a; if (a>x) x=a;}
END{for (i=m; i<=x; i++) {if (b[i])print b[i]}}' infile

Hi rdrtx1,
At least on OS X this doesn't work. Since a is created by a call to sub(), it is being treated as a string instead of as a number in the comparisons that set the minimum and maximum values for the range to be printed. I believe that the script is working on OS X as specified by the standards. I haven't tried it on any other system to see if awk behaves differently on other systems.

If you change m?(0):(m=a) in your script to a += 0 (which converts a from a string to a number), it works as expected.

How about:

sort -k1.8n infile

# cat infile
>string2991 234445
>string224 470561
>string121 675386
>string4098 177229
>string8049 255838
>string8 672382
>string1115 578415

# sort -k1.8n infile
>string8 672382
>string121 675386
>string224 470561
>string1115 578415
>string2991 234445
>string4098 177229
>string8049 255838