/python/engine/PinYin/tools/filter.awk

#!/usr/bin/gawk -f
{
_len = length ($1)
_freq = int ($3)
if ((_len == 2 && _freq > 20000) ||
    (_len == 3 && _freq > 100000) ||
    (_len == 4 && _freq > 500000))
	print $0
}