#!/usr/bin/env python3

# http://inamidst.com/voynich/stacks
# Created by Sean B. Palmer

import voynich

with open("voynich101_frequent.txt", "rb") as f:
    # Change to top48
    top48 = f.read(48).decode("iso-8859-1")
top48 = set([c for c in top48])

lengths = [1, 2, 3, 4, 5, 6, 7, 8, 9] # [10, 11, 12]
stacks = {g: {length: [0] * length for length in lengths} for g in top48}

# lengths = set()

distribution = [0] * 12

for word in voynich.words():
    if len(word) < 3:
        continue
    word = word[1:-1]
    length = len(word)
    distribution[length - 1] += 1
    if length > 9:
        continue
    # lengths.add(length)
    for (i, char) in enumerate(word):
        if char in top48:
            stacks[char][length][i] += 1

for glyph in top48:
    total = sum(sum(row) for row in stacks[glyph].values())
    for length in lengths:
        row = stacks[glyph][length]
        for p in range(length):
            # Convert to percentage
            if total is 0:
                row[p] = 0
            else:
                row[p] = row[p] / total * 100
        stacks[glyph][length] = row

print("glyph\tlength\tposition\tfrequency")
for glyph in top48:
    for length in lengths:
        for p in range(length):
            f = stacks[glyph][length][p]
            print("%s\t%s\t%s\t%s" % (ord(glyph), length, p + 1, f))