def recoursiveFind(self, sentenceDoc, subject, verb, root):
pred = verb.word
adv = findVerbModifier(sentenceDoc)
flag = True
if adv:
pred = pred + adv.orth_
flag = False
root = adv
for child in root.children:
if child.dep_ == "prep":
for proj in child.children:
if proj.dep_ == "pobj":
temp = self.nounArray.findWord(proj.orth_)
if flag:
newWord = datastructure.Word(child.orth_)
newWord.addType(child.pos_)
newWord.addUri(wordUri.findUri(newWord))
#newWord.addUri(newWord.word + "URI")
print(subject.uri, "- " + newWord.uri + " -",
temp.uri)
self.file.write(subject.uri + "; " + newWord.uri +
"; " + temp.uri + "\n")
else:
newWord = datastructure.Word(adv.orth_)
newWord.addType(adv.pos_)
newWord.addUri(wordUri.findUri(newWord))
#newWord.addUri(newWord.word + "URI")
print(subject.uri, "- " + newWord.uri + " -",
temp.uri)
self.file.write(subject.uri + "; " + newWord.uri +
"; " + temp.uri + "\n")
break
def findThird(self, sentenceDoc, subject, verb, children, flag):
for child in children:
if child.dep_ == "appos" or child.dep_ == "pobj":
temp = self.nounArray.findWord(child.orth_)
if temp is None:
w = datastructure.Word(child.orth_)
w.addType(child.pos_)
w.addUri(wordUri.findUri(w))
#w.addUri(w.word + "URI")
print(subject.uri, "- " + verb.uri + " -", w.uri)
self.writeOtter(subject.uri, verb.uri, w.uri)
else:
print(subject.uri, "- " + verb.uri + " -", temp.uri)
self.writeOtter(subject.uri, verb.uri, temp.uri)
#self.recoursiveFind(sentenceDoc, subject, verb, child)
if child.dep_ == "prep" or child.dep_ == "acomp":
if not flag:
verb = datastructure.Word(child.orth_)
verb.addType(child.pos_)
verb.addUri(wordUri.findUri(verb))
verbChildren = []
for ch in child.children:
verbChildren.append(ch)
self.findThird(sentenceDoc, subject, verb, verbChildren, True)
def findThird(self, sentenceDoc, subject, verb, children):
found = False
for child in children:
if child.dep_ == "attr":
temp = self.nounArray.findWord(child.orth_)
if temp is None:
w = datastructure.Word(child.orth_)
w.addType(child.pos_)
w.addUri(wordUri.findUri(child.lemma_))
#w.addUri(w.word + "URI")
print(subject.uri, "- " + verb.uri + " -", w.uri)
self.file.write(subject.uri + "; " + verb.uri + "; " +
w.uri + "\n")
else:
print(subject.uri, "- " + verb.uri + " -", temp.uri)
self.file.write(subject.uri + "; " + verb.uri + "; " +
temp.uri + "\n")
self.recoursiveFind(sentenceDoc, subject, verb, child)
found = True
if not found:
for word in sentenceDoc:
if word.dep_ == "ROOT":
verbDoc = word
break
self.recoursiveFind(sentenceDoc, subject, verb, verbDoc)
def findNouns(self):
nounArray = WordArray()
entities = list(self.doc.ents)
# print("There were {} entities found".format(len(entities)))
# print(entities)
for e in [
entity for entity in entities if entity.label_ not in [
'DATE', 'TIME', 'PERCENT', 'CARDINAL', 'MONEY', 'QUANTITY',
'ORDINAL', 'CARDINAL'
]
]:
temp = datastructure.Noun(e.orth_)
temp.addCategory(e.label_)
end = True
for w in self.doc:
if end:
string = e.orth_.split()
if w.orth_ in string:
temp.addType(w.pos_)
temp.addUri(wordUri.findUri(temp))
end = False
nounArray.addWord(temp)
for word in self.doc.noun_chunks:
temp = datastructure.Noun(word.orth_)
temp.addCategory("UNKNOWN")
temp.addType("NOUN")
if temp not in nounArray:
temp.addUri(wordUri.findUri(temp))
nounArray.addWord(temp)
print(temp)
for word in self.doc:
# print(nc.orth_)
if word.pos_ == 'NOUN' or word.pos_ == "PROPN":
temp = datastructure.Noun(word.orth_)
temp.addCategory("UNKNOWN")
temp.addType(word.pos_)
temp.addUri(wordUri.findUri(temp))
if temp not in nounArray:
nounArray.addWord(temp)
nounArray = purifyNounArray(nounArray)
nounArray = removeArticles(nounArray, self.doc)
return nounArray
def findVerbs(self):
verbsArray = WordArray()
temp = []
for word in self.doc:
if word.pos_ == 'VERB' and word.orth_ not in temp:
temp.append(word.orth_)
verb = datastructure.Verb(word.orth_)
verb.addRoot(word.lemma_)
verb.addType(word.pos_)
verb.addUri(wordUri.findUri(verb))
verbsArray.addWord(verb)
return verbsArray
def removeArticles(nounArray, doc):
removeList = ["det", "amod", "nummod"]
for word in nounArray:
if len(word.word.split()) > 1 and word.type != "PROPN":
for w in doc:
if w.orth_ in word.word.split() and w.dep_ in removeList:
temp = word.word.split()
string = ""
for t in temp:
if w.orth_ != t:
string = string + t + " "
word.word = string[:-1]
word.addUri(wordUri.findUri(word))
return nounArray