Comment by lqdc13

Edit: There was a bug. No it's actually slower when the locale is set to C.

I have 4434 files with about 100 words of text in them.

here is python func that searches utf8:

    def find_sp(paths):
        for path in paths:
            with open(path,'rb') as f:
                if b"snow" in f.read().lower():
                    yield True
                else:
                    yield False

    %timeit [o for o in find_sp(glob.glob('./*'))]
    10 loops, best of 3: 35.2 ms per loop

Here is the bash one:

    $>echo $LANG
    C
    $>time grep -Riq "snow" .

    real	0m0.015s
    user	0m0.005s
     sys	0m0.010s

However, if you have a lot of files and you are doing the searching for them in a not very smart way, it takes a lot longer.

    time for f in ./*txt; do grep -iq "snow" $f;done

    real	0m2.307s
    user	0m0.337s
     sys	0m2.113s