def run(self):
if (not self.document.pages):
return
FileHelper.writeToFile("{}-response.json".format(self.fileName),
json.dumps(self.response))
print("Total Pages in Document: {}".format(len(self.document.pages)))
p = 1
for page in self.document.pages:
FileHelper.writeToFile(
"{}-page-{}-response.json".format(self.fileName, p),
json.dumps(page.blocks))
self._outputWords(page, p)
self._outputText(page, p)
if (self.forms):
self._outputForm(page, p)
self._outputFormTranslate(page, p)
if (self.tables):
self._outputTable(page, p)
self._outputTablePretty(page, p)
self._outputTablePrettyTranslate(page, p)
p = p + 1
def _outputText(self, page, p):
text = page.text
FileHelper.writeToFile("{}-page-{}-text.txt".format(self.fileName, p),
text)
textInReadingOrder = page.getTextInReadingOrder()
FileHelper.writeToFile(
"{}-page-{}-text-inreadingorder.txt".format(self.fileName, p),
textInReadingOrder)
def _outputTablePretty(self, page, p, table_format='github'):
for table_number, table in enumerate(page.tables):
rows_list = list()
for row in table.rows:
one_row = list()
for cell in row.cells:
one_row = one_row + [cell.text]
rows_list.append(one_row)
pretty_table = tabulate(rows_list, tablefmt=table_format)
FileHelper.writeToFile(
"{}-page-{}-table-{}-tables-pretty.txt".format(
self.fileName, p, table_number), pretty_table)
def _generateInsightsPerDocument(self, page, p, insights, medicalInsights, translate, ta, tma, tt):
maxLen = 2000
text = page.text
start = 0
sl = len(text)
sentiment = []
syntax = []
entities = []
keyPhrases = []
medicalEntities = []
phi = []
translation = ""
while(start < sl):
end = start + maxLen
if(end > sl):
end = sl
subText = text[start:end]
if(insights):
self._insights(start, text, sentiment, syntax, entities, keyPhrases, ta)
if(medicalInsights):
self._medicalInsights(start, text, medicalEntities, phi, tma)
if(translate):
translation = translation + tt.getTranslation(subText) + "\n"
start = end
if(insights):
FileHelper.writeCSV("{}-page-{}-insights-sentiment.csv".format(self.fileName, p),
["Sentiment"], sentiment)
FileHelper.writeCSV("{}-page-{}-insights-entities.csv".format(self.fileName, p),
["Type", "Text", "Score", "BeginOffset", "EndOffset"], entities)
FileHelper.writeCSV("{}-page-{}-insights-syntax.csv".format(self.fileName, p),
["PartOfSpeech-Tag", "PartOfSpeech-Score", "Text", "BeginOffset", "EndOffset"], syntax)
FileHelper.writeCSV("{}-page-{}-insights-keyPhrases.csv".format(self.fileName, p),
["Text", "Score", "BeginOffset", "EndOffset"], keyPhrases)
if(medicalInsights):
FileHelper.writeCSV("{}-page-{}-medical-insights-entities.csv".format(self.fileName, p),
["Text", "Type", "Category", "Score", "BeginOffset", "EndOffset"], medicalEntities)
FileHelper.writeToFile("{}-page-{}-medical-insights-phi.json".format(self.fileName, p), json.dumps(phi))
if(translate):
FileHelper.writeToFile("{}-page-{}-text-translation.txt".format(self.fileName, p), translation)
def _outputTablePrettyTranslate(self, page, p, table_format='github'):
tt = None
tt = TextTranslater('auto', 'en', 'us-east-1')
for table_number, table in enumerate(page.tables):
rows_list = list()
for row in table.rows:
one_row = list()
for cell in row.cells:
if cell.text != "":
one_row = one_row + [tt.getTranslation(cell.text)]
else:
one_row = one_row + [cell.text]
rows_list.append(one_row)
pretty_table = tabulate(rows_list, tablefmt=table_format)
FileHelper.writeToFile(
"{}-page-{}-table-{}-tables-pretty-translated.txt".format(
self.fileName, p, table_number), pretty_table)
def _outputText(self, page, p):
text = page.text
FileHelper.writeToFile("{}-text.txt".format(self.fileName), text)
'''textInReadingOrder = page.getTextInReadingOrder()