def convert_images_to_string(images, output_file=False):
#whole_oa_string = ''
ocr_dict = {}
count = 1
print('converting images to string')
for img in images:
#whole_oa_string += pytesseract.image_to_string(img)
text = pytesseract.image_to_string(img)
ocr_dict[count] = text
print("page {} converted".format(count))
count += 1
print('done converting to string')
if output_file:
output_string = ""
for page_no, text in ocr_dict.items():
output_string += str(page_no) + "\n" + text + "\n\n"
fw.print_string_to_txt(output_string, "output.txt")
#return whole_oa_string
return ocr_dict
def convert_pdf_to_txt(pdf_file_path,use_pickle=True):
output_path = rename_pdf_to_txt(pdf_file_path)
#output_folder = "output/"
#output_path = output_folder + output_text_file
if use_pickle:
filehandler = open('images.p', 'r')
images = pickle.load(filehandler)
else:
images = convert_pdf_to_images(pdf_file_path)
raw_string = pytess.convertImagesToString(images)
file_writer.print_string_to_txt(raw_string,output_path)
return output_path
def test_main():
#only detailed description
raw_text = file_reader.get_string_from_txt('test.test')
words = nltk.word_tokenize(raw_text)
file_writer.print_string_to_txt("|\n".join(words), 'words.test')
regex_all_num = re.compile(r'\d+', re.M)
ref_numerals = []
for i, w in enumerate(words):
matchObj = re.fullmatch(regex_all_num, w)
if matchObj:
#print(matchObj.group(0))
ref_numerals.append((w, i))
print(ref_numeral_features(w))
for j in range(int(start_para), int(end_para)+1):
if j not in input_paragraphs:
input_paragraphs.append(j)
print(input_paragraphs)
#test
#input_paragraphs = [103,104,105,106,107,108,109,29]
input_path = "input/test_reference.txt"
input_paragraphs.sort()
reference_string = file_reader.get_string_from_txt(input_path)
#print(reference_string)
reference_split = reference_string.split("\n")
#print(reference_split)
output_string = []
for para in input_paragraphs:
print(para)
regex = r'\[\d+' + str(para) + r'\].+'
match_obj = re.search(regex, reference_string)
if (match_obj):
print(match_obj[0])
output_string.append(match_obj[0])
file_writer.print_string_to_txt("\n\n".join(output_string),"output/analyzed_reference.txt")
@author: alanyliu
"""
import file_reader
import file_writer
import re
input_path = "training/2018-09-21 15694060 nonfinal rejection.txt"
text = file_reader.get_string_from_txt(input_path)
print(text)
text = re.sub(r'(]\.)', '\g<1>\n', text)
text = re.sub(r'\n\s+\n', '\n', text)
text = re.sub(r'\n{2,}', '\n', text)
#print(re.findall(r'\w\n\w',text))
#print(re.findall(r'(\w)(\n)(\w)',text))
text = re.sub(r'(\w|,)(\n)(\w)', '\g<1> \g<3>', text)
text = re.sub(r'(\.)\s', '\g<1>\n', text)
file_writer.print_string_to_txt(text, "training/train.txt")
#punct-features ideas
#next word is punctuation
#last letter of previous word is capitalized
#probably doesn't end with acronym
#false positives, false negatives
#single double quotes
#claim interpretation