Merged revisions 75339 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r75339 | georg.brandl | 2009-10-11 10:39:16 +0200 (So, 11 Okt 2009) | 1 line

  Update markov demo.
........
This commit is contained in:
Georg Brandl 2009-10-11 08:42:09 +00:00
parent e63514b0e3
commit 2cbd09ca7e
1 changed files with 34 additions and 30 deletions

View File

@ -5,11 +5,10 @@ class Markov:
self.histsize = histsize
self.choice = choice
self.trans = {}
def add(self, state, next):
if state not in self.trans:
self.trans[state] = [next]
else:
self.trans[state].append(next)
self.trans.setdefault(state, []).append(next)
def put(self, seq):
n = self.histsize
add = self.add
@ -17,26 +16,29 @@ class Markov:
for i in range(len(seq)):
add(seq[max(0, i-n):i], seq[i:i+1])
add(seq[len(seq)-n:], None)
def get(self):
choice = self.choice
trans = self.trans
n = self.histsize
seq = choice(trans[None])
while 1:
while True:
subseq = seq[max(0, len(seq)-n):]
options = trans[subseq]
next = choice(options)
if not next: break
seq = seq + next
if not next:
break
seq += next
return seq
def test():
import sys, string, random, getopt
import sys, random, getopt
args = sys.argv[1:]
try:
opts, args = getopt.getopt(args, '0123456789cdw')
opts, args = getopt.getopt(args, '0123456789cdwq')
except getopt.error:
print('Usage: markov [-#] [-cddqw] [file] ...')
print('Usage: %s [-#] [-cddqw] [file] ...' % sys.argv[0])
print('Options:')
print('-#: 1-digit history size (default 2)')
print('-c: characters (default)')
@ -49,16 +51,19 @@ def test():
print('exactly one space separating words.')
print('Output consists of paragraphs separated by blank')
print('lines, where lines are no longer than 72 characters.')
sys.exit(2)
histsize = 2
do_words = 0
do_words = False
debug = 1
for o, a in opts:
if '-0' <= o <= '-9': histsize = eval(o[1:])
if o == '-c': do_words = 0
if o == '-d': debug = debug + 1
if '-0' <= o <= '-9': histsize = int(o[1:])
if o == '-c': do_words = False
if o == '-d': debug += 1
if o == '-q': debug = 0
if o == '-w': do_words = 1
if not args: args = ['-']
if o == '-w': do_words = True
if not args:
args = ['-']
m = Markov(histsize, random.choice)
try:
for filename in args:
@ -72,13 +77,15 @@ def test():
if debug: print('processing', filename, '...')
text = f.read()
f.close()
paralist = string.splitfields(text, '\n\n')
paralist = text.split('\n\n')
for para in paralist:
if debug > 1: print('feeding ...')
words = string.split(para)
words = para.split()
if words:
if do_words: data = tuple(words)
else: data = string.joinfields(words, ' ')
if do_words:
data = tuple(words)
else:
data = ' '.join(words)
m.put(data)
except KeyboardInterrupt:
print('Interrupted -- continue with data read so far')
@ -86,16 +93,19 @@ def test():
print('No valid input files')
return
if debug: print('done.')
if debug > 1:
for key in m.trans.keys():
if key is None or len(key) < histsize:
print(repr(key), m.trans[key])
if histsize == 0: print(repr(''), m.trans[''])
print()
while 1:
while True:
data = m.get()
if do_words: words = data
else: words = string.split(data)
if do_words:
words = data
else:
words = data.split()
n = 0
limit = 72
for w in words:
@ -103,15 +113,9 @@ def test():
print()
n = 0
print(w, end=' ')
n = n + len(w) + 1
n += len(w) + 1
print()
print()
def tuple(list):
if len(list) == 0: return ()
if len(list) == 1: return (list[0],)
i = len(list)//2
return tuple(list[:i]) + tuple(list[i:])
if __name__ == "__main__":
test()