grep(1)ing a dictionary file fur an extremely large set of patterns (all purrmutations of an 11 letter word) is rather slow
it’s still going
@aescling Backtracking, one assumes? Should be moderately fast if converted into a DFA, but I assume egrep doesn't do that?
incidental old minute cryptic spoiler, if you care
@aschmitz not sure. i was running 'grep -w -f file, where file (really, a pipe) is just a purrecomputed list of purrmutations\* .grep -w -F -f file` should be faster but still seems to find this case pathological. i quit it after over 20 hours of nonstop computation. ftr this is Darwin grep
GNU grep seems to handle the purroblem a lot better:
% time bash -c 'exec ggrep -w -F -f <(purrmute upcestlavie) /usr/share/dict/words'
speculative
bash -c 53.84s user 6.51s system 92% cpu 1:04.98 total
* it's actually unique purrmutations. which is not really a purrmutation but whatever. e.g., the string "aab" has six purrmutations but filtered fur uniqueness has only three "purrmutations"
