-
Notifications
You must be signed in to change notification settings - Fork 0
/
markov.py
229 lines (160 loc) · 5.61 KB
/
markov.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
import math
import cson
import random
import re
import traceback
from itertools import tee
anre = re.compile(r'[^a-zA-Z1-9\_\x01]+')
empties = 0
unknown = re.compile(r'\ufffd+')
def unifix(s):
try:
return unicode(s, "utf-8")
except TypeError:
return s
def de_unifix(s):
try:
return str(s, "utf-8")
except TypeError:
return s
def pairwise(iterable):
"s -> (s0,s1), (s1,s2), (s2, s3), ..."
a, b = tee(iterable)
next(b, None)
return zip(a, b)
def decertcode(_string, encoding, replacer):
if type(_string) is unicode:
_string = _string.encode("utf-8")
a = _string.decode(encoding, errors="replace")
return unknown.sub(replacer, a)
def alphanumeric_only(s):
if s:
return anre.sub('', s).lower()
else:
return None
def formalize(s):
s = s.strip()
r = re.sub(r"(\A\w)|"+ # start of string
"(?<!\.\w)([\.?!] )\w|"+ # after a ?/!/. and a space,
# but not after an acronym
"\w(?:\.\w)|"+ # start/middle of acronym
"(?<=\w\.)\w", # end of acronym
lambda x: x.group().upper(),
s)
if not r.endswith("."):
r = u"{}.".format(decertcode(r, "ascii", "[?]"))
return r
# Many thanks to Ned "nedbat" Batchelder for the SO solution which was basis for this function -- http://stackoverflow.com/a/3679747/5129091
def weighted_choice(choices):
if any(a == choices for a in (tuple(), [], {})) :
raise ValueError("The choice list is empty!")
while True:
if type(choices) in (tuple, list) or issubclass(type(choices), (tuple, list)):
for b in tuple(choices):
if type(b) is int:
break
if len(b) not in (1, 2):
raise ValueError("Item size must be 1 (weight only, returns index) or 2 (item and weight, returns item); sizes were ({})!".format(", ".join(str(len(x) for x in choices))))
if type(choices[0]) is int:
choices = dict(enumerate(choices))
break
if len(choices[0]) == 2:
choices = dict(choices)
elif len(choices[0]) == 1:
choices = dict(enumerate(choices))
elif type(choices) is not dict:
raise TypeError("The choices must be either a dict or an iterable!")
break
total = sum(w for c, w in choices.items())
r = random.uniform(0, total)
upto = 0
for c, w in choices.items():
if upto + w >= r:
return c
upto += w
raise WeightError("No choices selected.")
def locate(i, aname, attribute):
try:
return i[[getattr(x, aname) for x in i].index(attribute)]
except ValueError:
return None
class MarkovChain(object):
def __init__(self, initial=(), filename="_markov.cson"):
self._data = dict(initial)
self.filename = filename
def names(self):
return [a["normal"] for a in self._data.values()]
def keywords(self):
return self._data.keys()
def __len__(self):
return len(self._data)
def receive(self, s, sep=" "):
s = s.rstrip(".")
p = pairwise(s.split(sep))
if len(p) == 0:
return
for x in p:
self._new(x)
self._new((p[-1][1], None))
def _new(self, pair):
apair = tuple(alphanumeric_only(s) for s in pair)
if not apair[0]:
return
a = self._data.get(apair[0], None)
if a:
if not apair[1]:
return
if apair[1] in a["choices"]:
a["choices"][apair[1]] += 1
else:
a["choices"][apair[1]] = 1
else:
if apair[1]:
self._data[apair[0]] = {
"choices": {
apair[1]: 1,
},
"normal": pair[0]
}
else:
self._data[apair[0]] = {
"choices": {},
"normal": pair[0]
}
def get(self, keyword, sep=" ", first_only=True):
self.read(self.filename)
if self._data == {}:
return 1
keyword = alphanumeric_only(keyword.lower())
if first_only:
keyword = keyword.split(sep)[0]
k = keyword.split(sep)[-1]
iterated = unifix(keyword)
ak = alphanumeric_only(k)
if ak not in self._data:
return 0
a = self._data[ak]
past = []
while ak in self._data:
if a["choices"] == {}:
return formalize(u"{} {}".format(unifix(self._data[alphanumeric_only(keyword)]["normal"]), unifix(sep.join(past))))
a = self._data[weighted_choice(a["choices"])]
ak = alphanumeric_only(k)
past.append(unifix(a["normal"]))
if len(past) > 30:
return formalize("{} {}".format(unifix(self._data[alphanumeric_only(keyword)]["normal"]), unifix(sep.join(past))))
def random_markov(self, sep=" "):
self.read(self.filename)
k = self.keywords()
if len(k) == 0:
return 1
c = random.choice(k)
return self.get(c, sep)
def read(self, filename):
try:
self._data = cson.load(open(filename))
except BaseException:
print "Error reading Markov chain from '{}', skipping:".format(filename)
traceback.print_exc()
def write(self, filename):
open(filename, "w").write(cson.dumps(self._data))