def initialise(wcfg, wfsa, root, goal, intersection):
"""
Calculate a first derivation based on a simpler (thus smaller/faster) version of the grammar
Thereby determining the initial conditions.
Only applicable with the 'milos' grammar format, i.e. non-terminals have the form: '[P1234*2_1]'
"""
smaller = WCFG([])
logging.debug('Creating a smaller grammar for initial conditions...')
for line in wcfg:
if 0 < permutation_length(line.lhs) <= 2:
smaller.add(line)
elif line.lhs == root or line.lhs == '[UNK]':
smaller.add(line)
if intersection == 'nederhof':
init_parser = Nederhof(smaller, wfsa)
elif intersection == 'earley':
init_parser = Earley(smaller, wfsa)
else:
raise NotImplementedError('I do not know this algorithm: %s' %
intersection)
logging.debug('Init Parsing...')
init_forest = init_parser.do(root, goal)
if not init_forest:
print 'NO PARSE FOUND'
return {}
else:
logging.debug('Forest: rules=%d', len(init_forest))
logging.debug('Init Topsorting...')
# sort the forest
sorted_nodes = top_sort(init_forest)
# calculate the inside weight of the sorted forest
logging.debug('Init Inside...')
init_inside_prob = inside(init_forest, sorted_nodes)
logging.debug('Init Sampling...')
gen_sampling = GeneralisedSampling(init_forest, init_inside_prob)
init_d = gen_sampling.sample(goal)
return get_conditions(init_d)
def initialise(wcfg, wfsa, root, goal, intersection):
"""
Calculate a first derivation based on a simpler (thus smaller/faster) version of the grammar
Thereby determining the initial conditions.
Only applicable with the 'milos' grammar format, i.e. non-terminals have the form: '[P1234*2_1]'
"""
smaller = WCFG([])
logging.debug('Creating a smaller grammar for initial conditions...')
for line in wcfg:
if 0 < permutation_length(line.lhs) <= 2:
smaller.add(line)
elif line.lhs == root or line.lhs == '[UNK]':
smaller.add(line)
if intersection == 'nederhof':
init_parser = Nederhof(smaller, wfsa)
elif intersection == 'earley':
init_parser = Earley(smaller, wfsa)
else:
raise NotImplementedError('I do not know this algorithm: %s' % intersection)
logging.debug('Init Parsing...')
init_forest = init_parser.do(root, goal)
if not init_forest:
print 'NO PARSE FOUND'
return {}
else:
logging.debug('Forest: rules=%d', len(init_forest))
logging.debug('Init Topsorting...')
# sort the forest
sorted_nodes = top_sort(init_forest)
# calculate the inside weight of the sorted forest
logging.debug('Init Inside...')
init_inside_prob = inside(init_forest, sorted_nodes)
logging.debug('Init Sampling...')
gen_sampling = GeneralisedSampling(init_forest, init_inside_prob)
init_d = gen_sampling.sample(goal)
return get_conditions(init_d)
def make_grammar(fsa):
cfg = WCFG()
cfg.add(Rule('[S]', ['[X]'], 0.0))
cfg.add(Rule('[X]', ['[X]', '[X]'], 0.0))
for word in fsa.itersymbols():
cfg.add(Rule('[X]', [word], 0.0))
return cfg
def make_grammar(fsa):
cfg = WCFG()
cfg.add(Rule('[S]', ['[X]'], 0.0))
cfg.add(Rule('[X]', ['[X]', '[X]'], 0.0))
for word in fsa.itersymbols():
cfg.add(Rule('[X]', [word], 0.0))
return cfg