X-Git-Url: https://git.njae.me.uk/?a=blobdiff_plain;f=lettercount.py;fp=lettercount.py;h=0000000000000000000000000000000000000000;hb=311b300d197536622980f7a837294d8245e326b4;hp=4a7082d1068669762d1c8526c761382d07ed6182;hpb=d7224fba67d9f99c01bd78ef669c96189686e4c2;p=cipher-tools.git diff --git a/lettercount.py b/lettercount.py deleted file mode 100644 index 4a7082d..0000000 --- a/lettercount.py +++ /dev/null @@ -1,21 +0,0 @@ -import collections -import string - -def sanitise(text): - return [l.lower() for l in text if l in string.ascii_letters] - -corpora = ['shakespeare.txt', 'sherlock-holmes.txt', 'war-and-peace.txt'] -counts = collections.defaultdict(int) - -for corpus in corpora: - text = sanitise(open(corpus, 'r').read()) - for letter in text: - counts[letter] += 1 - -sorted_letters = sorted(counts, key=counts.get, reverse=True) - -with open('count_1l.txt', 'w') as f: - for l in sorted_letters: - f.write("{0}\t{1}\n".format(l, counts[l])) - - \ No newline at end of file