def __init__(self, options):
self.sentence_frame = []
self.options = options
if self.options.engine == 'relex':
self.codec = relex.relex()
self.relex = relex_analyze()
self.rule_engine = rule_engine()
self.helper = sentence_helper()
self.concepts = Concepts(self.options)
class sentence:
""" the main processing class for all of Godel """
def __init__(self, options):
self.sentence_frame = []
self.options = options
if self.options.engine == 'relex':
self.codec = relex.relex()
self.relex = relex_analyze()
self.rule_engine = rule_engine()
self.helper = sentence_helper()
self.concepts = Concepts(self.options)
def process(self, Sentence):
if not Sentence or len(Sentence) < 2:
return
word_set = []
self.hg = HG()
tokenized = self.helper.tokenize(Sentence)
tagged = self.helper.tag(tokenized)
## create a container for each word
for word in tokenized:
currentWord = universal_word(word)
self.helper.universalWord(currentWord)
word_set.append(currentWord)
universalSentence = universal_sentence(word_set)
## put the tokens and the sentence as a whole in the
## container.
universalSentence.whole_sentence = Sentence
universalSentence.tokenized = tokenized
processed = self.codec.process(Sentence)
parsed = self.codec.parse_output(processed)
## tie the original sentence with the parsed one
to_analyze = (Sentence, parsed)
universalSentence.analysis = to_analyze
features = self.relex.parse_features(to_analyze)
if features:
universalSentence.features = features
## hypergraph work
self.hg.sentenceToHG(universalSentence)
if features:
self.hg.features_in(universalSentence)
hg = self.hg.get_hypergraph()
self.hg.analysis()
universalSentence.hypergraph = hg
## then init the rule engine
self.rule_engine.initialize(universalSentence)
self.concepts.deriveConcepts(universalSentence)
debug(universalSentence)