def get_text_of_a_document(source_file,
keep_txt=False,
reuse_txt=False,
output_folder=None):
global EXTERNAl_CONVERTORS
ec = EXTERNAl_CONVERTORS
_, file_extension = os.path.splitext(source_file)
file_extension = file_extension.lower()
if output_folder is None:
txt_file = source_file + ".txt"
else:
txt_file = os.path.join(output_folder,
os.path.basename(source_file) + ".txt")
if reuse_txt and os.path.exists(txt_file):
pass
elif file_extension == ".xlsx":
ec.run_xlsx2csv(source_file, txt_file)
elif file_extension == ".xls":
res = ec.run_xls2csv(source_file, txt_file)
if res != 0:
temp_fname = source_file + ".xlsx"
shutil.copy(source_file, temp_fname)
ec.run_xlsx2csv(temp_fname, txt_file)
os.unlink(temp_fname)
elif file_extension == ".docx":
ec.run_office2txt(source_file, txt_file)
elif file_extension == ".pdf":
temp_file = source_file + ".docx"
sha256 = build_dislosures_sha256(source_file)
if TDocConversionClient(
TDocConversionClient.parse_args([])).retrieve_document(
sha256, temp_file) and os.path.exists(temp_file):
ec.run_office2txt(temp_file, txt_file)
else:
# the worse case, let's use calibre
ec.run_calibre(source_file, txt_file)
if os.path.exists(temp_file):
os.unlink(temp_file)
elif file_extension in {".html", ".rtf", ".htm"}:
ec.run_calibre(source_file, txt_file)
elif file_extension == ".doc":
res = ec.run_catdoc(source_file, txt_file)
if res != 0:
temp_fname = source_file + ".docx"
shutil.copy(source_file, temp_fname)
ec.run_office2txt(temp_fname, txt_file)
os.unlink(temp_fname)
else:
return None
if os.path.exists(txt_file):
doc_text = read_input_text(txt_file)
if not keep_txt:
os.unlink(txt_file)
return doc_text
else:
return None
def __init__(self, args):
self.logger = setup_logging(log_file_name="export_human_files.log")
self.args = args
if self.args.tmp_folder is None:
self.args.tmp_folder = tempfile.mkdtemp("export_human")
self.logger.debug("create folder {}".format(self.args.tmp_folder))
else:
self.logger.debug("rm folder {}".format(self.args.tmp_folder))
shutil.rmtree(self.args.tmp_folder, ignore_errors=True)
os.mkdir(self.args.tmp_folder)
self.source_doc_client = TSourceDocClient(TSourceDocClient.parse_args([]), self.logger)
self.pdf_conversion_client = TDocConversionClient(TDocConversionClient.parse_args([]), self.logger)
self.smart_parser_server_client = TSmartParserCacheClient(TSmartParserCacheClient.parse_args([]), self.logger)
self.new_pdfs = set()
def __init__(self, args):
self.register_task_result_error_count = 0
self.logger = setup_logging(log_file_name=args.log_file_name,
append_mode=True)
self.conversion_client = TDocConversionClient(
TDocConversionClient.parse_args([]), self.logger)
self.args = args
rounds = TDeclarationRounds(args.round_file)
self.dlrobot_remote_calls = TRemoteDlrobotCallList(
logger=self.logger,
file_name=args.remote_calls_file,
min_start_time_stamp=rounds.start_time_stamp)
self.worker_2_running_tasks = defaultdict(list)
self.worker_2_continuous_failures_count = defaultdict(int)
offices = TOfficeTableInMemory()
offices.read_from_local_file(self.args.offices_file)
self.web_sites_db = TDeclarationWebSiteList(self.logger,
offices=offices)
if not os.path.exists(self.args.result_folder):
os.makedirs(self.args.result_folder)
self.web_sites_to_process = self.find_projects_to_process()
self.cloud_id_to_worker_ip = dict()
self.config = TRobotConfig.read_by_config_type(
self.args.dlrobot_config_type)
self.last_remote_call = None # for testing
host, port = self.args.server_address.split(":")
self.logger.debug("start server on {}:{}".format(host, port))
super().__init__((host, int(port)), TDlrobotRequestHandler)
self.last_service_action_time_stamp = time.time()
self.service_action_count = 0
self.decl_sender = TDeclarationSender(
self.logger, self.args.enable_smart_parser,
self.args.enable_source_doc_server)
self.stop_process = False
if self.args.enable_ip_checking:
self.permitted_hosts = set(
str(x)
for x in ipaddress.ip_network('192.168.100.0/24').hosts())
self.permitted_hosts.add('127.0.0.1')
self.permitted_hosts.add('95.165.96.61') # disclosures.ru
self.logger.debug("init complete")
self.send_to_telegram("start dlrobot central with {} tasks".format(
len(self.web_sites_to_process)))
def process_all_tasks(self):
if len(self.ocr_tasks) == 0:
self.ocr_queue_is_empty_last_time_stamp = time.time()
self.try_convert_with_winword()
new_files_from_ocr = self.process_docx_from_ocr()
if new_files_from_ocr:
self.got_ocred_file_last_time_stamp = time.time()
# file garbage tasks
current_time = time.time()
if current_time - self.file_garbage_collection_timestamp >= 60: # just not too often
self.file_garbage_collection_timestamp = current_time
if not self.http_server_is_working:
return
self.process_ocr_logs()
if not self.http_server_is_working:
return
self.process_stalled_files()
if current_time - self.self_server_ping_timestamp >= 3600: # just not too often
args = TDocConversionClient.parse_args(
["--server-address", self.args.server_address])
client = TDocConversionClient(args, self.logger)
if not client.assert_declarator_conv_alive(raise_exception=False):
self.logger.error("cannot ping itself, exit")
self.stop_http_server(run_shutdown=False)
sys.exit(1)
self.self_server_ping_timestamp = current_time
current_time = time.time()
if current_time - self.got_ocred_file_last_time_stamp > self.args.ocr_restart_time and \
current_time - self.ocr_queue_is_empty_last_time_stamp > self.args.ocr_restart_time :
self.logger.debug(
"last ocr file was received long ago and all this time the ocr queue was not empty"
)
if not self.http_server_is_working:
return
self.restart_ocr()
self.got_ocred_file_last_time_stamp = time.time(
) #otherwize restart will be too often
def process_with_client(self,
input_files,
timeout=None,
rebuild=False,
skip_receiving=False,
log_name="client",
input_task_timeout=5):
output_files = list(os.path.basename(i) + ".docx" for i in input_files)
for o in output_files:
if os.path.exists(o):
os.unlink(o)
client_args = [
"--server-address",
self.server_address,
"--conversion-timeout",
"180",
"--output-folder",
".",
] + input_files
if timeout is not None:
client_args.extend(['--conversion-timeout', str(timeout)])
if rebuild:
client_args.append('--rebuild')
if skip_receiving:
client_args.append('--skip-receiving')
if self.client_count >= 0 and log_name == "client":
log_name = log_name + str(self.client_count)
logger = setup_logging(logger_name=log_name)
try:
self.client_count += 1
self.client = TDocConversionClient(
TDocConversionClient.parse_args(client_args), logger=logger)
self.client.input_task_timeout = input_task_timeout
self.client.start_conversion_thread()
self.client.process_files()
return output_files
finally:
close_logger(logger)
import json
from ConvStorage.conversion_client import TDocConversionClient
import argparse
import os
import time
def parse_args():
parser = argparse.ArgumentParser()
parser.add_argument("--history-file", dest='history_file', default=None)
return parser.parse_args()
if __name__ == '__main__':
args = parse_args()
conv_client = TDocConversionClient(TDocConversionClient.parse_args([]), logging)
stats = conv_client.get_stats()
if args.history_file is None:
print(json.dumps(stats))
else:
lines = list()
if os.path.exists(args.history_file):
with open(args.history_file, "r", encoding="utf-8") as inp:
for l in inp:
lines.append(l)
lines.append("{}\t{}\n".format(int(time.time()), json.dumps(stats)))
lines = lines[-400:]
with open(args.history_file, "w", encoding="utf-8") as out:
for l in lines:
out.write(l)
from ConvStorage.conversion_client import TDocConversionClient
from common.logging_wrapper import setup_logging
import sys
if __name__ == '__main__':
logger = setup_logging(log_file_name="convert_pdf.log")
client = TDocConversionClient(
TDocConversionClient.parse_args(sys.argv[1:]), logger)
client.start_conversion_thread()
exit_code = client.process_files()
sys.exit(exit_code)
