You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
84 lines
2.7 KiB
84 lines
2.7 KiB
import sys
|
|
import dill
|
|
dill.load_session('voilastate.db')
|
|
inpu = sys.argv[1]
|
|
|
|
try:
|
|
outsentences, punctuations = oi.CellInputText2SplitsentencesWithspacy(inpu)
|
|
except:
|
|
out = 'Bitte auf die Zeichensetzung achten! '
|
|
print(out)
|
|
exit()
|
|
|
|
outsentences = sy.replaceOpposites(outsentences)
|
|
insentences = outsentences
|
|
outsentences, punctuations = sent_seg.putAppendixesIntoOwnSentences(insentences, punctuations)
|
|
|
|
insentences = outsentences
|
|
|
|
sentsegformattedsentences = []
|
|
for sentence in insentences:
|
|
|
|
sentsegformattedsentences.append([sentence])
|
|
|
|
#print(sentsegformattedsentences)
|
|
sentences = sent_seg.CommaSentenceOrNot(sentsegformattedsentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetUtteranceNumber(sentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetQuestionOrNot(sentences)
|
|
#print(sentences)
|
|
sentences, punctuations = sent_seg.AndOrSolver(sentences, punctuations)
|
|
#print(len(punctuations))
|
|
#print('sentences after AndOrSolver', sentences, punctuations)
|
|
sentences = sent_seg.CommaSentenceOrNot(sentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetUtteranceNumber(sentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetQuestionOrNot(sentences)
|
|
#print(sentences)
|
|
sentences, punctuations = sent_seg.SplitSentencesIntoHauptNebenTuple(sentences, punctuations)
|
|
#print(len(punctuations))
|
|
#print('sentences after split haupt neben', sentences)
|
|
sentences = sent_seg.CommaSentenceOrNot(sentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetUtteranceNumber(sentences)
|
|
#print(sentences)
|
|
sentences = sent_seg.GetQuestionOrNot(sentences)
|
|
#print(sentences)
|
|
outsentences, punctuations = sent_seg.SplitCommatas(sentences, punctuations)
|
|
|
|
insentences = outsentences
|
|
outsentences = solSh.ExplainShortsInSentencesWithBrackets(insentences)
|
|
insentences = outsentences
|
|
outsentences = p2a.replacePassivForms(insentences)
|
|
insentences = outsentences
|
|
outsentences = gs.ReplaceGenitivWithDativ(insentences)
|
|
|
|
insentences = []
|
|
for sentence in outsentences:
|
|
insentences.append([word.replace('ö','oe').replace('ü','ue').replace('ä','ae').replace('ß','ss') for word in sentence])
|
|
|
|
|
|
outsentences = cs.replaceConjunctives(insentences)
|
|
insentences = outsentences
|
|
outsentences, punctuations = fwb.fremdEintragAppend(insentences, punctuations)
|
|
insentences = outsentences
|
|
outsentences, punctuations = medi.Medioreplace(insentences, punctuations)
|
|
|
|
out = ''
|
|
#print('just before looking if the lenght is the same')
|
|
#print(len(outsentences), len(punctuations))
|
|
if len(outsentences) == len(punctuations):
|
|
#print('the length is the same')
|
|
for n in range(len(outsentences)):
|
|
outsentence = outsentences[n]
|
|
outsentence[0] = outsentence[0].title()
|
|
out += ' '.join(outsentence) + punctuations[n] + '\n'
|
|
|
|
|
|
print(out)
|
|
|
|
|
|
|
|
|