Shenon-Fano coding script

Author: Flyer-DM

Shenon-Fano coding/Shenon-Fano coding.py

@@ -0,0 +1,103 @@
+from collections import Counter
+from fractions import Fraction
+from math import log
+
+
+def get_letter_chances(message: str) -> tuple:
+    letters = dict(Counter(message))
+    message_length = len(message)
+    letters = dict(sorted(letters.items(), key=lambda x: x[1], reverse=True))
+    letter_chances = {i: Fraction(letters[i], message_length) for i in letters}
+    entropy = -(sum(map(lambda x: letter_chances[x] * log(letter_chances[x]), letter_chances)))
+    entropy = round(entropy, 1)
+    letter_chances = dict(sorted(letter_chances.items(), key=lambda x: x[1], reverse=True))
+    return letter_chances, entropy
+
+
+Code = [] 
+
+
+def splitter(group: list, code=None, where=None) -> None:
+    half = sum([i[1] for i in group]) / 2
+    left, right = [], []
+    sum_right = Fraction(0, 1)
+    for i in range(len(group)):
+        if sum_right + group[i][1] >= half:
+            left.append(group[i])
+            right.extend(group[i+1:])
+            break
+        else:
+            left.append(group[i])
+            sum_right += group[i][1]
+
+    if len(left) == 1 and code is None and where is None:
+        Code.append((left[0], '0'))
+        if len(right) == 1:
+            Code.append((right[0], '1'))
+        else:
+            splitter(right, '1', 'r')
+    if len(left) > 1 and code is None and where is None:
+        splitter(left, '0', 'l')
+        if len(right) == 1:
+            Code.append((right[0], '1'))
+        else:
+            splitter(right, '1', 'r')
+    if len(left) > 1 and where == 'l':
+        c = code + '0'
+        splitter(left, c, 'l')
+        c = code + '1'
+        if len(right) == 1:
+            Code.append((right[0], c))
+        else:
+            splitter(right, c, 'l')
+    if len(left) == 1 and where == 'l':
+        c = code + '0'
+        Code.append((left[0], c))
+        c = code + '1'
+        if len(right) == 1:
+            Code.append((right[0], c))
+        else:
+            splitter(right, c, 'l')
+
+    if len(right) > 1 and where == 'r':
+        c = code + '1'
+        splitter(right, c, 'r')
+        c = code + '0'
+        if len(left) == 1:
+            Code.append((left[0], c))
+        else:
+            splitter(left, c, 'r')
+    if len(right) == 1 and where == 'r':
+        c = code + '1'
+        Code.append((right[0], c))
+        c = code + '0'
+        if len(left) == 1:
+            Code.append((left[0], c))
+        else:
+            splitter(left, c, 'r')
+
+
+def encode(letter_chances: dict, entropy: float) -> dict:
+    group = list(letter_chances.items())
+    splitter(group)
+    encoded_chances = Code.copy()
+    encoded_chances.sort(key=lambda x: x[0][1], reverse=True)
+    lengths = Counter([len(i[1]) for i in encoded_chances])
+    lengths_sum = len(encoded_chances)
+    math_expect = sum(map(lambda x: x * Fraction(lengths[x], lengths_sum), lengths))
+    math_expect = round(float(math_expect), 3)
+    code_economy = round(entropy / math_expect, 3)
+    beauty_encoded_chances = {i[0][0]: i[1] for i in encoded_chances}
+    print("Математическое ожидание длины закодированной буквы равно", math_expect)
+    print("Экономность кода равна", code_economy)
+    return beauty_encoded_chances
+    
+
+def get_codes(codes: dict) -> None:
+    for symbol in codes:
+        print("|Символ {0:1} закодирован как {1:10}|".format(symbol, codes[symbol]))
+    
+
+
+test0, test1 = get_letter_chances('дверь_коридор_лестница')
+get_codes(encode(test0, test1))