Add Markov chain

markov.py

@@ -0,0 +1,55 @@
+#!/usr/bin/env python3
+from sys import argv
+from string import ascii_letters
+from random import choice
+
+
+def get_word_list(text):
+    # Define a set of allowed letters
+    allowed_letters = ascii_letters + "áàéèíìóòúù" + "ÁÀÉÈÍÌÓÒÚÙ" + "'."
+    # Keep only the allowed replacing the others with a space
+    text = "".join(c.lower()
+                   if c in allowed_letters else " "
+                   for c in text)
+    # Split by words, using space as separator
+    return text.split()
+
+
+def get_associations(word_list):
+    # Build a dictionary where each word is a key, and the values are the list
+    # of the words that follow the key one
+    # ['a', 'b', 'a',  'c']  became {'a': ['b', 'c'], 'b': ['a']}
+    associations = {}
+    for i, word in enumerate(word_list[:-1]):
+        if word not in associations:
+            associations[word] = []
+        associations[word].append(word_list[i+1])
+    return associations
+
+
+def generate(text, max_words=100):
+    word_list = get_word_list(text)
+    associations = get_associations(word_list)
+    out_list = []
+    current_word = choice(word_list)
+
+    for i in range(max_words):
+        out_list.append(current_word)
+        choice_list = associations.get(current_word, word_list)
+        if current_word.endswith("."):
+            break
+        current_word = choice(choice_list)
+
+    out_string = " ".join(out_list)
+    out_string = out_string.replace(" .", ".")
+    out_string = out_string[0].upper() + out_string[1:]
+
+    return out_string
+
+
+if __name__ == '__main__':
+    if argv[1:]:
+        with open(argv[1]) as f:
+            print(generate(f.read()))
+    else:
+        print('Usage:', argv[0], 'something_antani.txt')