def chartParser():
"""
线图句法分析
"""
from nltk.grammar import CFG
from nltk.parse.chart import ChartParser, BU_LC_STRATEGY
# BNF格式文法 开始符号：S 终结符号：单词
grammar = CFG.fromstring("""
S -> T1 T4
T1 -> NNP VBZ
T2 -> DT NN
T3 ->IN NNP
T4 -> T3 | T2 T3
NNP -> 'Tajmahal' | 'Agra' | 'Bangalore' | 'Karnataka'
VBZ -> 'is'
IN -> 'in' | 'of'
DT -> 'the'
NN -> 'capital'
""")
cp = ChartParser(grammar, BU_LC_STRATEGY, trace=True)
# trace=True可以看见分析过程
# strategy=BU_LC_STRATEGY是默认的，不写好像也行
sentence = 'Bangalore is the capital of Karnataka'
tokens = sentence.split()
chart = cp.chart_parse(tokens) # 对单词列表分析，并存到chart对象
parses = list(chart.parses(grammar.start())) # 将chart取到的所有分析树赋给parses
print('Total Edges:', len(chart.edges())) # 输出chart对象所有边的数量
for tree in parses:
print(tree)
tree.draw()
def __init__(self, **parser_args):
"""
Constructor. Initializes a MbmaParser.
Args:
- parser_args: needs a keyword grammar which is of
type :class:`ContextFreeGrammar`
"""
ChartParser.__init__(self, [], BU_RHR_STRATEGY, **parser_args)
def generate_context_free_grammar_novel_text(
self, number_of_words_in_sentence=0, number_of_sentences_per_record=0, number_of_records=0
):
"""
This method utilizes NLTK's Context Free Grammar parser objects to parse an available .*cfg file and generate
novel text from it.
@param number_of_words_in_sentence: An indicator as to the number of words to generate in each novel sentence.
@type number_of_words_in_sentence: int
@param number_of_sentences_per_record: An indicator as to the number of sentences per record to generate.
@type number_of_sentences_per_record: int
@param number_of_records: An indicator as to the total number of records to generate.
@type number_of_records: int
@return: str
"""
words = []
punct_selector = [". ", "! ", "? "]
punctuation_stop_symbols = dict((ord(char), None) for char in string.punctuation)
parser = None
grammar = None
try:
if isinstance(self._corpus, CFG):
_grammar = self._corpus
if _grammar is not None:
parser = ChartParser(_grammar)
grammar = parser.grammar
elif isinstance(self._corpus, FeatureGrammar):
_grammar = self._corpus
if _grammar is not None:
parser = FeatureChartParser(_grammar)
grammar = parser.grammar()
elif isinstance(self._corpus, PCFG):
_grammar = self._corpus
if _grammar is not None:
parser = InsideChartParser(_grammar)
grammar = parser.grammar()
else:
grammar = CFG.fromstring(self._corpus)
if grammar is not None:
for _ in range(number_of_records):
novel_sentence = []
for _ in range(number_of_sentences_per_record):
sentence = " ".join(
[
sent
for _, sent in enumerate(generate_text(grammar, depth=2, n=number_of_words_in_sentence))
]
)
sentence = sentence.translate(punctuation_stop_symbols) + random.choice(punct_selector)
sentence = sentence[0:].capitalize()
novel_sentence.append(sentence)
words.append("".join(novel_sentence))
except Exception, error:
self.logger.error(
"TextGenerator.generate_context_free_grammar_novel_text: Error occurred - {0}".format(str(error))
)
def __init__(self, grammar,
strategy=BU_LC_FEATURE_STRATEGY,
trace_chart_width=20,
chart_class=FeatureChart,
**parser_args):
ChartParser.__init__(self, grammar,
strategy=strategy,
trace_chart_width=trace_chart_width,
chart_class=chart_class,
**parser_args)
def __init__(self, grammar,
strategy=BU_LC_FEATURE_STRATEGY,
trace_chart_width=20,
chart_class=FeatureChart,
**parser_args):
ChartParser.__init__(self, grammar,
strategy=strategy,
trace_chart_width=trace_chart_width,
chart_class=chart_class,
**parser_args)
def generate_context_free_grammar_novel_text(
self, corpus, number_of_words_in_sentence,
number_of_sentences_per_record, number_of_records):
'''
This method utilizes NLTK's Context Free Grammar parser objects to
parse an available .*cfg file and generate novel text from it.
Params:
-------
- number_of_words_in_sentence (int): An indicator as to the number of
words to generate in each novel sentence.
- number_of_sentences_per_record (int): An indicator as to the number
of sentences per record to generate.
- number_of_records (int): An indicator as to the total number of
records to generate.
Returns: str
'''
words = []
punct_selector = ['. ', '! ', '? ']
punctuation_stop_symbols = dict(
(ord(char), None) for char in string.punctuation)
parser = None
grammar = None
try:
if isinstance(corpus, CFG):
_grammar = corpus
if _grammar is not None:
parser = ChartParser(_grammar)
grammar = parser.grammar
elif isinstance(corpus, FeatureGrammar):
_grammar = corpus
if _grammar is not None:
parser = FeatureChartParser(_grammar)
grammar = parser.grammar()
elif isinstance(corpus, PCFG):
_grammar = corpus
if _grammar is not None:
parser = InsideChartParser(_grammar)
grammar = parser.grammar()
else:
grammar = CFG.fromstring(corpus)
if grammar is not None:
for _ in range(number_of_records):
novel_sentence = []
for _ in range(number_of_sentences_per_record):
sentence = ' '.join([
sent for _, sent in enumerate(
generate_text(grammar,
depth=2,
n=number_of_words_in_sentence))
])
sentence = sentence.translate(
punctuation_stop_symbols) + random.choice(
punct_selector)
sentence = sentence[0:].capitalize()
novel_sentence.append(sentence)
words.append(''.join(novel_sentence))
except Exception, error:
logging.error('TextGenerator: Error occurred - {0}'.format(
str(error)))
from nltk.grammar import CFG
from nltk.parse.chart import ChartParser, BU_LC_STRATEGY
grammar = CFG.fromstring("""
S -> T1 T4
T1 -> NNP VBZ
T2 -> DT NN
T3 -> IN NNP
T4 -> T3 | T2 T3
NNP -> 'Tajmahal' | 'Agra' | 'Bangalore' | 'Karnataka'
VBZ -> 'is'
IN -> 'in' | 'of'
DT -> 'the'
NN -> 'capital'
""")
cp = ChartParser(grammar, BU_LC_STRATEGY, trace=True)
sentence = "Bangalore is the capital of Karnataka"
tokens = sentence.split()
chart = cp.chart_parse(tokens)
parses = list(chart.parses(grammar.start()))
print("Total Edges :", len(chart.edges()))
for tree in parses:
print(tree)
tree.draw()