Regular expression weirdness: egrep vs awk

Question

Having used scripting before on other UNIX flavours I ran into the following question recently (Mac OS X 10.4.10 but also earlier versions):
Consider two files which I created for illustration purposes
a) text file t1.txt
b) awk script t1.awk
and run the commands egrep and awk against t1.txt
/Users/me/bin>cat t1.txt
1
2
12
1
2
34
3
4
/Users/me/bin>egrep '^[0-9]{2}$' t1.txt
12
34
/Users/me/bin>cat t1.awk
/^[0-9]$/ { printf("Number: %i has one digit\n",$1) }
/^[0-9]{2}$/ { printf("Number: %i has two digits\n",$1) }
/Users/me/bin>awk -f t1.awk < t1.txt
Number: 1 has one digit
Number: 2 has one digit
Number: 1 has one digit
Number: 2 has one digit
Number: 3 has one digit
Number: 4 has one digit
/Users/me/bin>
Same regular expression, egrep acts as expected, awk does not. I am obviously too dumb to see what I am doing wrong, anybody any idea?
Thanks
DocJoc

Answer 1

Ok, I checked this out a little further following the suggestions given. I find
1) awk is installed, nawk is not
2) I got gawk and
a) it generates the very same error if called without addtl parameters
b) it runs as expected if adding option "--posix" as suggested by LittleSaint
Well, looks like a bug to me because according to the man pages awk should already show the expected behaviour. Anyway, using gawk solves the problem here.
Thanks for the answers!