def __init__(self):
self.sentences = []
self.matchesCrafted = []
self.matchesRegex = []
self.matchesNeural = []
self.title = ""
self.textLoaded = False
self.Neural = False
self.st = SplitText(self)
self.gt = GetText()
self.ch = Chunking()
self.ch.addTT(self)
self.nameListNeural = [
"Value", "Pattern Grammar", "Pattern Sentence", "Article Grammar",
"Article Sentence", "Compared"
]
self.notFile = "C:\\Users\\trace\\projects\\python\\masters\\informationPull\\pdata\\notlist.txt"
self.resultFolder = 'C:\\Users\\trace\\projects\\python\\masters\\informationPull\\pdata\\results'
self.notList = []
self.patterns = False
self.load()
def main(dramaname, autor, act):
u"""Main zum Ausführen des Programms."""
gd = GetData()
gt = GetText()
gs = GetSentiment()
gm = GraphMalen()
if dramaname:
dramaname = dramaname
else:
dramaname = click.prompt('Gib den Namen eines Dramas ein')
if autor:
autor = autor
else:
autor = click.prompt('Gib den Nachnamen des Autors ein')
draname = gd.eingabe_drama(dramaname, autor)
tei = gd.get_tei(draname)
csv_drama = gd.get_csv(draname)
replik = gt.create_replik_dict(csv_drama)
soup = gt.stir_the_soup(tei)
if act:
pass
else:
print("Das ausgewählte Drama hat {} Akte".format(
gt.how_many_acts(soup)))
act = click.prompt(
'Gib den Akt ein, den du analysieren willst (falls du das Netzwerk für das gesamte Drama haben möchtest, wähle 0)'
)
which_act = int(act) - 1
if which_act == -1:
total = gt.drama_total(soup)
replik = gt.which_type(total, replik)
else:
act = gt.drama_act(soup, which_act)
replik = gt.which_type(act, replik)
replik = gs.get_sentis(replik)
all_in_all = gs.average_senti(replik)
nodes = gm.get_nodes(csv_drama)
edges = gm.get_edges(all_in_all)
labels_edges = gm.get_labels(edges)
graph = gm.graph(edges, nodes)
gm.malen(graph, labels_edges, draname, which_act + 1)
os.system('clear')
menu()
def screenShot():
"""用于截图并保存"""
print('请按F1开始截图')
if keyboard.wait(hotkey='f1') == None:
print('复制剪切板的图片，请按Ctrl+b，不复制继续截图')
if keyboard.wait(hotkey='Ctrl+b') == None:
sleep(0.02) # 防止获取的是上一张截图
# 复制剪贴板里面的图片
im = ImageGrab.grabclipboard()
im.save('Picture.png')
if __name__ == '__main__':
baiduapi = BaiDuAPI('password.ini')
for _ in range(sys.maxsize):
screenShot()
texts = baiduapi.picture2Text('Picture.png')
print(texts)
GetText.setText(texts) # 剪贴板剪贴
sleep(0.02)
GetText.getText()
print('退出请按Ctrl+x')
if keyboard.wait(hotkey='Ctrl+x') == None:
name = input('请输入保存图像识别文字文件名:')
f = open(name + '.txt', 'w')
f.write(texts)
f.close()
break
#获取剪切板上的图片信息并保存到本地
def screenShot():
#QQ截图按住Ctrl+A+alt截图，enter完成截图
if keyboard.wait(hotkey='ctrl+alt+a')==None:
if keyboard.wait(hotkey='enter')==None:
sleep(0.01)
#获取剪切板的图像内容
im=ImageGrab.grabclipboard()
#判断im的类型是否为图片
if isinstance(im,Image.Image):
im.save('imageGrab.png')
else:
print('重新截图')
else:
print('请按Ctrl+A+Alt来截图识别文字')
if __name__=='__main__':
baiduapi=BaiDuAPI(r'C:\Users\Administrator\.spyder-py3\screenshot\password.ini')
#baiduapi=BaiDuAPI
#maxsize for循环是为了可以循环截图，一直不结束程序
for _ in range(sys.maxsize):
screenShot()
text=baiduapi.picture2Text(r'C:\Users\Administrator\.spyder-py3\screenshot\imageGrab.png')
print(text)
GetText.setText(text)
GetText.getText()
# 同上，这里是截图开始
if keyboard.wait(hotkey='Enter') is None:
# 因为获取剪切板的图片太快了，所以要停顿一小段时间
sleep(0.01)
# 复制剪切板中的图片
im = ImageGrab.grabclipboard()
sleep(0.01)
im.save('Picture.png')
if __name__ == "__main__":
# 为了可以多次截图而不是截一次图就结束程序
baidu_api = BaiDuAPI()
for i in range(sys.maxsize):
if i == 0:
print("注意事项：请使用QQ的默认截图按键（Ctrl+alt+a）,截图完成后请敲击回车键完成截图!")
print("截图完成后，你截取图片中的文字自动复制，您可自由粘贴！")
print("现在开始截取你需要的图片吧！\n")
else:
print("您已成功截图，请截取下一张图！\n")
screenShot()
res = baidu_api.pictureText()
print('截取的内容：' + res)
GetText.setText(res)
sleep(0.01)
GetText.getText()
class TextObject:
def __init__(self):
self.sentences = []
self.matchesCrafted = []
self.matchesRegex = []
self.matchesNeural = []
self.title = ""
self.textLoaded = False
self.Neural = False
self.st = SplitText(self)
self.gt = GetText()
self.ch = Chunking()
self.ch.addTT(self)
self.nameListNeural = [
"Value", "Pattern Grammar", "Pattern Sentence", "Article Grammar",
"Article Sentence", "Compared"
]
self.notFile = "C:\\Users\\trace\\projects\\python\\masters\\informationPull\\pdata\\notlist.txt"
self.resultFolder = 'C:\\Users\\trace\\projects\\python\\masters\\informationPull\\pdata\\results'
self.notList = []
self.patterns = False
self.load()
def loadPatterns(self):
self.patterns = self.st.getPatternObject()
def load(self):
'Load troublesome word file. "not" etc.'
f = open(self.notFile, "r", encoding="utf8")
ns = f.readlines()
for x in ns:
self.notList.append(x.strip())
self.loadPatterns()
def reset(self):
self.sentences = []
self.matchesCrafted = []
self.matchesRegex = []
self.matchesNeural = []
#Pattern Sentence
def returnPOSList(self, sentence):
return self.ch.returnPOSList(sentence)
def getGraph(self, sentence):
#return self.ch.treeHTML(sentence)
return self.ch.displacyService(sentence)
def sentenceTokenDisplay(self, sentence):
return self.st.sentenceTokenDisplay(sentence)
def sentenceTokenDisplayList(self, sentenceList):
ls = []
for x in sentenceList:
ls.append(self.st.sentenceTokenDisplay(x))
return ls
def getArticle(self, name):
'Given an article title, return list of sentences from article'
article = self.gt.getArticle(name)
if article != False:
if len(article) > 0:
title = article[0]
text = article[1]
sentences = self.st.textToSentenceList(text)
self.saveSentences(title, sentences)
return [title, sentences]
return False
def splitSentencesInSectionList(self, articleSections):
for x in articleSections:
tx = x["Text"]
sentences = self.st.textToSentenceList(tx)
txList = []
for y in sentences:
txList.append(y)
x["Text"] = txList
return articleSections
def getArticleSectionList(self, name):
articleSections = self.gt.getArticleSectionList(name)
if articleSections != False:
if len(articleSections) > 0:
title = articleSections[0]
textDict = articleSections[1]
sentences = self.splitSentencesInSectionList(textDict)
#sentences = textDict
self.saveSentences(title, sentences)
return [title, sentences]
return False
def isTroubleSome(self, sentence):
checkSentence = sentence.lower()
for x in self.notList:
if x in checkSentence:
return True
return False
def removeTroublesomeSentencesRegex(self, matchList, nameToCheck):
for x in matchList:
if (len(x) > 0) and (x[0] != False):
for y in x:
if self.isTroubleSome(y[nameToCheck]):
x.remove(y)
return matchList
def removeTroublesomeSentencesNeural(self, matchList, nameToCheck):
'Given a list of dicts, removes entries where any word in "nameToCheck" matches ban list'
for x in matchList:
if self.isTroubleSome(x[nameToCheck]):
matchList.remove(x)
return matchList
def findRegexMatches(self):
print('Finding Regex Matches')
articleList = self.getTextOnly(self.sentences)
patternList = self.getPatterns()
matches = self.ch.compareRegexAll(articleList, patternList)
writeList = []
for x in matches:
if x:
writeList.append(x[0])
self.writeMatched(writeList, '-regex')
#rt = self.st.findRegexMatches(inputText)
#rtNoTrouble = self.removeTroublesomeSentencesRegex(rt,self.st.nameListRegex[3])
#self.writeMatched(rtNoTrouble,'-regex')
#return rtNoTrouble
return matches
def getTextOnly(self, textDict):
textList = []
for x in textDict:
for y in x['Text']:
textList.append(y)
return textList
def textToPOSTokens(self, text):
return 0
def constructReducedSentence(self, index, articleSentence, fullSentence,
shortSentence):
indexList = index.split(" ")
indexs = []
for x in indexList:
try:
indexs.append(int(x))
except:
print(index, articleSentence, fullSentence, shortSentence)
return self.ch.constructReducedSentence(indexs, articleSentence,
fullSentence, shortSentence)
def getColumn(self, lst, col):
return [val[col] for val in lst]
def getPatterns(self):
return self.patterns
def findExactStructureMatches(self):
text = self.getTextOnly(self.sentences)
tks = self.sentenceTokenDisplayList(text)
patterns = self.getPatterns()
POSList = []
matchList = []
if not patterns:
return False
for x in text:
for y in patterns:
match = self.ch.exactMatch(x, y['ShortSentence'])
if match:
reducedSen = self.constructReducedSentence(
y['ShortIndex'], x, y['FullSentence'],
y['ShortSentence'])
resultDict = {
'Article Sentence': x,
'Pattern': y['ShortSentence'],
'Match': 'Exact',
'Reduced Sentence': reducedSen
}
matchList.append(resultDict)
#for x in patterns:
# POSList.append([x['ShortPOS'],x['ShortSentence']])
#for x in tks:
# xList = self.getColumn(x,1)
# for y in POSList:
# yList = y[0].split(' ')
# compareNum = self.listCompareValue(xList,yList)
# if (compareNum == 1.0):
# textSen = ' '.join(self.getColumn(x,0))
# savedSen = y[1]
# resultDict = {
# 'Article Sentence':textSen,
# 'Pattern':savedSen,
# 'Match':'Exact',
# }
# matchList.append(resultDict)
return matchList
def treeMatchFunct(self, x, patterns, matchList):
return False
#added = False
#for idy,y in enumerate(patterns):
# if not added:
# if 'ShortSentence' in y:
# if self.ch.compareTree(x,y['ShortSentence']):
# match = self.ch.exactMatch(x,y['ShortSentence'])
# added = True
#reducedSen = ''
# resultDict = {
# 'Article Sentence':x,
# 'Short Sentence':y['ShortSentence'],
# }
# if match:
# resultDict['Exact'] = 'Exact'
# if (y['FullSentence'] is not y['ShortSentence']) and match:
# reducedSen = self.constructReducedSentence(y['ShortIndex'],x,y['FullSentence'],y['ShortSentence'])
# resultDict['Reduced Sentence'] = reducedSen
# matchList.append(resultDict)
def findTreeMatches(self):
matches = self.ch.compareTreeAll(self.getTextOnly(self.sentences),
self.getPatterns())
self.writeMatched(matches, '-tree')
matchesr = self.removeTroublesomeSentencesNeural(
matches, 'Article Sentence')
return matchesr
def findNeuralMatches(self, resultType, lengthType):
if self.textLoaded:
if not self.Neural:
self.Neural = NeuralClass()
self.Neural.loadModel()
#text = testText
text = self.getTextOnly(self.sentences)
#A list of sentences
grammarTextList = self.st.listToToken(text)
grammarText = []
#Turns [('word0','POS0'),('word1','POS1')] to 'POS0 POS1'
for senList in grammarTextList:
sentence = ""
for word in senList:
sentence += word[1] + " "
grammarText.append(sentence)
#Load patterns
#[0] Reduced
#[1] Full
#[2] Dependency
#[3] Head
patterns = self.getPatterns()
fullSentence = []
fullPOS = []
shortSentence = []
shortPOS = []
for x in patterns:
fullSentence.append(x['FullSentence'])
fullPOS.append(x['FullPOS'])
shortSentence.append(x['ShortSentence'])
shortPOS.append(x['ShortPOS'])
matchSentence = fullSentence
matchPOS = fullPOS
if (lengthType == "short"):
matchSentence = shortSentence
matchPOS = shortPOS
getNeuText = False
getNeuPOS = True
getNeuDep = False
getNeuHead = False
results = {"Text": "F", "POS": "F", "Dependency": "F", "Head": "F"}
if (getNeuText):
results["Text"] = self.Neural.runAndPlotPatterns(
matchSentence, text)
if (getNeuPOS):
results["POS"] = self.Neural.runAndPlotPatterns(
matchPOS, grammarText)
if (getNeuDep):
results["Dep"] = self.Neural.runAndPlotPatterns(
matchPOS, grammarText)
if (getNeuHead):
results["Head"] = self.Neural.runAndPlotPatterns(
matchPOS, grammarText)
matchList = []
for idx, x in enumerate(results["POS"]):
for idy, y in enumerate(x):
#Text = results[]
ps = results["POS"][idx][idy]
resultDict = {
self.nameListNeural[0]: y,
self.nameListNeural[1]: matchPOS[idy],
self.nameListNeural[2]: matchSentence[idy],
self.nameListNeural[3]: grammarText[idx],
self.nameListNeural[4]: text[idx],
self.nameListNeural[5]: resultType,
}
# matchList.append([ y,grammarPattern[idy],sentencesFull[idy],grammarText[idx],text[idx]])
matchList.append(resultDict)
rt = self.convertListString(self.viewMatches(matchList))
rem = self.removeTroublesomeSentencesNeural(
rt, self.nameListNeural[4])
rems = rem[:100]
noDupes = []
for x in rems:
ok = True
for y in noDupes:
if x[self.nameListNeural[4]] == y[self.nameListNeural[4]]:
ok = False
if ok:
noDupes.append(x)
return noDupes
def convertListString(self, data):
for x in data:
for key, value in x.items():
x[key] = str(value)
#x[self.nameListNeural[0]] = str(x[self.nameListNeural[0]])
return data
def writeMatched(self, data, typed):
titleEnd = self.title
if titleEnd.endswith(' '):
titleEnd = titleEnd[:-1]
f = open('{0}\\{1}{2}.txt'.format(self.resultFolder, titleEnd, typed),
"a",
encoding="utf-8")
# f.write(json.dumps(str(data)))
# json_data = f.read()
#data = json.loads(json_data)
for x in data:
if x:
for key in x:
f.write('{0}:{1}'.format(key, x[key]))
f.write("\n")
#f.write('{0}:{1}'.format(self.nameListNeural[4],x[self.nameListNeural[4]]))
#f.write("\n")
#f.write('{0}:{1}'.format(self.nameListNeural[2],x[self.nameListNeural[2]]))
#f.write("\n")
#f.write('{0}:{1}'.format(self.nameListNeural[0],x[self.nameListNeural[0]]))
#f.write("\n")
f.write('END:')
f.write("\n")
def viewMatches(self, neuralMatchList):
'Sorts so higher values are first'
#columnIndex = 0
#sortedArr = neuralMatchList
#sortedArr.sort(key=lambda x: x[columnIndex],reverse=True)
sortedArr = sorted(neuralMatchList,
key=lambda i: i[self.nameListNeural[0]],
reverse=True)
self.writeMatched(sortedArr, '-neural')
return sortedArr
def saveSentences(self, title, sentences):
self.textLoaded = True
self.title = title
self.sentences = sentences
import time
from PIL import ImageGrab, Image
import sys
from baidu import BaiDuapi
from getText import GetText
def screenShot():
#监听按压事件
if keyboard.wait(hotkey='shift+a') == None:
time.sleep(0.01)
if keyboard.wait(hotkey='esc') == None:
im = ImageGrab.grabclipboard()
if isinstance(im, Image.Image):
im.save('图片.jpg')
#print('图片保存完成')
else:
print('请重新截图')
if __name__ == "__main__":
a = BaiDuapi('api.ini')
for i in range(sys.maxsize):
screenShot()
cc = a.pictureText('图片.jpg')
print(cc)
GetText.setText(cc)
GetText.getText()