def pdf_thread(self, url):
pdf_name = ''
exit_call = ''
csv_row = []
# save PDF to disk
try:
pdf_name = BytesIO(
url.split("/")[-1].encode('UTF-8')).read().__str__()[2:-1]
valid_chars = "-_.() %s%s" % (string.ascii_letters, string.digits)
regex = re.compile(valid_chars)
pdf_name = regex.sub('', pdf_name.__str__())
self.pdf_path = self.document_folder + regex.sub('', pdf_name)
r = requests.get(url, headers={'User-Agent': 'Mozilla/5.0'})
with open(self.pdf_path, 'wb') as code:
code.write(r.content)
code.close()
csv_row.insert(0, [self.csv_header[0], self.line_count.__str__()])
csv_row.insert(
1, [self.csv_header[1], url if url.__len__() > 0 else 'NULL'])
csv_row.insert(2, [
self.csv_header[2],
pdf_name if pdf_name.__len__() > 0 else 'NULL'
])
csv_row.insert(3, [
self.csv_header[3],
self.pdf_path if self.pdf_path.__len__() > 0 else 'NULL'
])
print(' >>>> PDF START:[' + url + '] ' +
self.line_count.__str__() + ' ' +
(datetime.datetime.now().__str__()[:-7]))
except Exception as e:
csv_row.insert(0, [self.csv_header[0], self.line_count.__str__()])
csv_row.insert(
1, [self.csv_header[1], url if url.__len__() > 0 else 'NULL'])
csv_row.insert(2, [self.csv_header[2], e.__str__()])
csv_row.insert(3, [
self.csv_header[3],
self.pdf_path if self.pdf_path.__len__() > 0 else 'NULL'
])
print(e)
pass
my_file = os.path.join(self.document_folder + pdf_name)
try:
fp = open(my_file, 'rb')
# self.pdf(fp, csv_row)
except Exception as e:
print(' PDF LOAD FAILED !!! ' + self.line_count.__str__() +
' : ' + self.pdf_path)
csv_row.pop(3)
csv_row.insert(3, [
self.csv_header[3], 'PDF FAILED TO OPEN:' +
self.pdf_path if self.pdf_path.__len__() > 0 else 'NULL'
])
# Write results
row = []
for i in range(csv_row.__len__()):
row.append(csv_row[i][1])
report_path = self.report_folder + self.report_name
row_append = [
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '',
'', '', '', '', '', ''
]
index = 4
for ii in row_append:
row.insert(index, ii)
index += 1
# OPEN FAILED
with open(report_path, 'a', encoding='utf8',
newline='') as csv_file:
writer = csv.writer(csv_file, quoting=csv.QUOTE_ALL)
writer.dialect.lineterminator.replace('\n', '')
writer.writerow(row)
return
try:
self.pdf(fp, csv_row)
except Exception as e:
print('PDF FAIL')
