Calculating CRC in awk

Has anyone implemented the POSIX 1003.2 compiliant CRC algorithm (as output by cksum) in awk/gawk? I'm needing to do a checksum on an in memory string (not the whole file) and shelling out to call cksum is slow and expensive.

My overall need is to generate a numerical checksum that fits within 10 digits or less. Other hash/CRC functions could work too, anyone have any thing handy?

A Google search and a scan of awk.info turned up nothing interesting.

EDIT:
I ended up using the external cksum command, but caching the results into an awk associative array. Performance was good enough and I didn't need to reinvent the wheel.

Solution

Since cksum uses a large table, it's probably impractical to re-implement it in AWK. You might be able to calculate it on the fly without using a table, but that's likely to be slower than calling cksum.

References:

Translating it from C to AWK should be fairly trivial, however, if someone were so inclined.

By the way, gawk has coprocesses:

gawk 'BEGIN {
    cmd="cksum"
    print "hello" |& cmd
    close(cmd, "to")
    while (cmd |& getline a > 0)
        print a
    close(cmd)
    }'