2010-03-11 18:53:45 -04:00
|
|
|
#! /usr/bin/env python3
|
1993-12-14 06:08:02 -04:00
|
|
|
|
|
|
|
class Markov:
|
2004-07-18 02:56:09 -03:00
|
|
|
def __init__(self, histsize, choice):
|
|
|
|
self.histsize = histsize
|
|
|
|
self.choice = choice
|
|
|
|
self.trans = {}
|
2009-10-11 05:42:09 -03:00
|
|
|
|
2004-07-18 02:56:09 -03:00
|
|
|
def add(self, state, next):
|
2009-10-11 05:42:09 -03:00
|
|
|
self.trans.setdefault(state, []).append(next)
|
|
|
|
|
2004-07-18 02:56:09 -03:00
|
|
|
def put(self, seq):
|
|
|
|
n = self.histsize
|
|
|
|
add = self.add
|
|
|
|
add(None, seq[:0])
|
|
|
|
for i in range(len(seq)):
|
|
|
|
add(seq[max(0, i-n):i], seq[i:i+1])
|
|
|
|
add(seq[len(seq)-n:], None)
|
2009-10-11 05:42:09 -03:00
|
|
|
|
2004-07-18 02:56:09 -03:00
|
|
|
def get(self):
|
|
|
|
choice = self.choice
|
|
|
|
trans = self.trans
|
|
|
|
n = self.histsize
|
|
|
|
seq = choice(trans[None])
|
2009-10-11 05:42:09 -03:00
|
|
|
while True:
|
2004-07-18 02:56:09 -03:00
|
|
|
subseq = seq[max(0, len(seq)-n):]
|
|
|
|
options = trans[subseq]
|
|
|
|
next = choice(options)
|
2009-10-11 05:42:09 -03:00
|
|
|
if not next:
|
|
|
|
break
|
|
|
|
seq += next
|
2004-07-18 02:56:09 -03:00
|
|
|
return seq
|
1993-12-14 06:08:02 -04:00
|
|
|
|
2009-10-11 05:42:09 -03:00
|
|
|
|
1993-12-14 06:08:02 -04:00
|
|
|
def test():
|
2009-10-11 05:42:09 -03:00
|
|
|
import sys, random, getopt
|
2004-07-18 02:56:09 -03:00
|
|
|
args = sys.argv[1:]
|
|
|
|
try:
|
2009-10-11 05:42:09 -03:00
|
|
|
opts, args = getopt.getopt(args, '0123456789cdwq')
|
2004-07-18 02:56:09 -03:00
|
|
|
except getopt.error:
|
2009-10-11 05:42:09 -03:00
|
|
|
print('Usage: %s [-#] [-cddqw] [file] ...' % sys.argv[0])
|
2007-07-17 17:59:35 -03:00
|
|
|
print('Options:')
|
|
|
|
print('-#: 1-digit history size (default 2)')
|
|
|
|
print('-c: characters (default)')
|
|
|
|
print('-w: words')
|
|
|
|
print('-d: more debugging output')
|
|
|
|
print('-q: no debugging output')
|
|
|
|
print('Input files (default stdin) are split in paragraphs')
|
|
|
|
print('separated blank lines and each paragraph is split')
|
|
|
|
print('in words by whitespace, then reconcatenated with')
|
|
|
|
print('exactly one space separating words.')
|
|
|
|
print('Output consists of paragraphs separated by blank')
|
|
|
|
print('lines, where lines are no longer than 72 characters.')
|
2009-10-11 05:42:09 -03:00
|
|
|
sys.exit(2)
|
2004-07-18 02:56:09 -03:00
|
|
|
histsize = 2
|
2009-10-11 05:42:09 -03:00
|
|
|
do_words = False
|
2004-07-18 02:56:09 -03:00
|
|
|
debug = 1
|
|
|
|
for o, a in opts:
|
2009-10-11 05:42:09 -03:00
|
|
|
if '-0' <= o <= '-9': histsize = int(o[1:])
|
|
|
|
if o == '-c': do_words = False
|
|
|
|
if o == '-d': debug += 1
|
2004-07-18 02:56:09 -03:00
|
|
|
if o == '-q': debug = 0
|
2009-10-11 05:42:09 -03:00
|
|
|
if o == '-w': do_words = True
|
|
|
|
if not args:
|
|
|
|
args = ['-']
|
|
|
|
|
2004-07-18 02:56:09 -03:00
|
|
|
m = Markov(histsize, random.choice)
|
|
|
|
try:
|
|
|
|
for filename in args:
|
|
|
|
if filename == '-':
|
|
|
|
f = sys.stdin
|
|
|
|
if f.isatty():
|
2007-07-17 17:59:35 -03:00
|
|
|
print('Sorry, need stdin from file')
|
2004-07-18 02:56:09 -03:00
|
|
|
continue
|
|
|
|
else:
|
|
|
|
f = open(filename, 'r')
|
2007-07-17 17:59:35 -03:00
|
|
|
if debug: print('processing', filename, '...')
|
2004-07-18 02:56:09 -03:00
|
|
|
text = f.read()
|
|
|
|
f.close()
|
2009-10-11 05:42:09 -03:00
|
|
|
paralist = text.split('\n\n')
|
2004-07-18 02:56:09 -03:00
|
|
|
for para in paralist:
|
2007-07-17 17:59:35 -03:00
|
|
|
if debug > 1: print('feeding ...')
|
2009-10-11 05:42:09 -03:00
|
|
|
words = para.split()
|
2004-07-18 02:56:09 -03:00
|
|
|
if words:
|
2009-10-11 05:42:09 -03:00
|
|
|
if do_words:
|
|
|
|
data = tuple(words)
|
|
|
|
else:
|
|
|
|
data = ' '.join(words)
|
2004-07-18 02:56:09 -03:00
|
|
|
m.put(data)
|
|
|
|
except KeyboardInterrupt:
|
2007-07-17 17:59:35 -03:00
|
|
|
print('Interrupted -- continue with data read so far')
|
2004-07-18 02:56:09 -03:00
|
|
|
if not m.trans:
|
2007-07-17 17:59:35 -03:00
|
|
|
print('No valid input files')
|
2004-07-18 02:56:09 -03:00
|
|
|
return
|
2007-07-17 17:59:35 -03:00
|
|
|
if debug: print('done.')
|
2009-10-11 05:42:09 -03:00
|
|
|
|
2004-07-18 02:56:09 -03:00
|
|
|
if debug > 1:
|
2007-08-06 18:07:53 -03:00
|
|
|
for key in m.trans.keys():
|
2004-07-18 02:56:09 -03:00
|
|
|
if key is None or len(key) < histsize:
|
2007-07-17 17:59:35 -03:00
|
|
|
print(repr(key), m.trans[key])
|
|
|
|
if histsize == 0: print(repr(''), m.trans[''])
|
|
|
|
print()
|
2009-10-11 05:42:09 -03:00
|
|
|
while True:
|
2004-07-18 02:56:09 -03:00
|
|
|
data = m.get()
|
2009-10-11 05:42:09 -03:00
|
|
|
if do_words:
|
|
|
|
words = data
|
|
|
|
else:
|
|
|
|
words = data.split()
|
2004-07-18 02:56:09 -03:00
|
|
|
n = 0
|
|
|
|
limit = 72
|
|
|
|
for w in words:
|
|
|
|
if n + len(w) > limit:
|
2007-07-17 17:59:35 -03:00
|
|
|
print()
|
2004-07-18 02:56:09 -03:00
|
|
|
n = 0
|
2007-07-17 17:59:35 -03:00
|
|
|
print(w, end=' ')
|
2009-10-11 05:42:09 -03:00
|
|
|
n += len(w) + 1
|
2007-07-17 17:59:35 -03:00
|
|
|
print()
|
|
|
|
print()
|
1993-12-14 06:08:02 -04:00
|
|
|
|
2004-09-11 13:34:35 -03:00
|
|
|
if __name__ == "__main__":
|
|
|
|
test()
|