def store_record(self, index_name, doc_name, content):
if not self.is_connected():
logger.error('Error. Not connected to Elasticsearch')
return
if type(index_name) is not str:
logger.error('Error. Index name must be a str')
return
if type(doc_name) is not str:
logger.error('Error. Missing document name to store in Elasticsearch')
return
if not isinstance(content, dict):
logger.error('Error. Missing content to store in Elasticsearch')
return
t1 = time()
logger.debug("Gevent (before es_obj.index): '{}'".format(gevent.getcurrent().name))
try:
res = self.es.index(index=index_name, doc_type=doc_name, body=content)
except Exception as ex:
logger.error('Error. Something went wrong storing the data')
return
else:
logger.debug("Gevent (after es_obj.index: '{}' - {}".
format(gevent.getcurrent().name, time() - t1))
return res
def connect(self):
self.es = Elasticsearch([{'host': self.host, 'port': self.port}])
if self.is_connected():
msg = 'Connected to ElasticSearch on'
logger.info('{msg} {host}:{port}'.format(msg=msg, host=self.host,
port=self.port))
else:
msg = 'Error. Failed to connect to Elasticsearch on'
logger.error('{msg} {host}:{port}'.format(msg=msg, host=self.host,
port=self.port))
def create_index(self, index_name, mapping=None):
if not self.is_connected():
logger.error('Error. Not connected to Elasticsearch')
return
if type(index_name) is not str:
logger.error('Error. Index name must be a str')
return
if mapping and not isinstance(mapping, dict):
logger.error('Error. Mapping must be a dictionary')
return
try:
if not self.es.indices.exists(index_name):
# Ignore 400 means to ignore "Index Already Exist" error.
res = self.es.indices.create(index=index_name,
body=mapping,
ignore=[400, 404])
except Exception as ex:
logger.error("Error creating the index '{}'.Error: {}".
format(index_name, str(ex)))
return
else:
logger.info(("Index '{}' was created successfully").
format(index_name))
return True
def search(self, index_name, content):
if not self.is_connected():
logger.error('Error. Not connected to Elasticsearch')
return
if type(index_name) is not str:
logger.error('Error. Index name must be a str')
return
if type(content) is not str:
loggin.error('Error. Content must be a dictionary')
return
return self.es.search(index=index_name, body=content)
def es_init(es_addr, es_port):
es = ES(es_addr, es_port)
es.connect()
for idx in list(mappings.keys()):
if not es.secure_delete_index(idx):
logger.error("Error deleting index '{}'".format(idx))
return
if not es.create_index(idx, mappings.get(idx, '')):
logger.error("Error creating index '{}'".format(idx))
return
return
def secure_delete_index(self, index_name):
if not self.is_connected():
logger.error('Error. Not connected to Elasticsearch')
return
if type(index_name) is not str:
logger.error('Error. Index name must be a str')
return
msg = "Do you want to delete the index '{}'?".format(index_name)
if self.es.indices.exists(index_name):
if utils.query_yes_no(msg, False):
res = self.es.indices.delete(index=index_name)
logger.info("The index {} was deleted successfully".
format(index_name))
return True
def parse_pdf2img(filename, folder_img):
try:
with tempfile.TemporaryDirectory() as tmppath:
images = convert_from_path(filename,
dpi=80,
fmt='jpeg',
strict=False,
last_page=10,
output_folder=tmppath)
utils.create_directory(folder_img)
files = os.listdir(tmppath)
for file in files:
src = os.path.join(tmppath, file)
shutil.move(src, folder_img)
return True
except:
logger.error(("pdf2image could not convert " +
" the document '{}'").format(filename))
return False
def parse_pdf(root, file_name, file_extension, folder='', encoding='utf-8'):
t0 = time()
content = {}
file_path = os.path.join(root, folder, file_name + file_extension)
logger.debug('Gevent (init parse_pdf): {}. File: {}'.format(
gevent.getcurrent().name, file_path))
status = 'error'
clean_text = ''
content = {}
exclude_sent_with_words = []
file_exc_words = './exclude_words.txt'
if os.path.isfile(file_exc_words):
exclude_sent_with_words = utils.read_txt_file(file_exc_words)
if file_extension != '.pdf':
logger.error("File extension of '{}' is not '.pdf'".format(file_path))
else:
eof = subprocess.check_output(['tail', '-n', '1', file_path])
# %%EOF, %%EOF\n, %%EOF\r, %%EOF\r\n
eof = eof.replace(b'\r', b'')
eof = eof.replace(b'\n', b'')
if (b'%%EOF' in eof[-4:]):
logger.error("Error reading EOF bytes '{}' from '{}'".format(
eof.decode('utf-8'), file_path))
else:
t1 = time()
pdfinfo = get_pdfinfo(file_path)
numpages = pdfinfo.get('pages', -1)
logger.debug('Gevent (before textract.process): {}'.format(
gevent.getcurrent().name))
try:
text = textract.process(file_path, encoding=encoding)
except:
logger.error(
("Unexpected error while parsing PDF file_path '{}' " +
"using textract").format(file_path))
return {'status': status, 'args': file_path, 'data': content}
logger.debug('Gevent (after textract.process: {} - {}'.format(
gevent.getcurrent().name,
time() - t1))
text = text.decode("utf-8")
text = utils.remove_non_printable_chars(text)
text = text.split('\n')
for line in text:
if not line and clean_text[-2:] != '\n\n':
clean_text += '\n'
else:
if "disclosure" in line.lower(): break
for exc_words in exclude_sent_with_words:
if '@' in line.lower(): break
if re.search(
r'\b' + exc_words.lower().replace('+', '\+') +
r'\b', line.lower()):
break
else:
if text.count(line) <= max(numpages - 10, 4):
#remove extra spaces
clean_line = re.sub(r'\s+', ' ', line)
clean_line = utils.remove_nonsense_lines(
str(clean_line), 6)
if clean_line:
clean_text += clean_line + '\n'
if not clean_text:
logger.error(
("textract was unable to parse " +
"the contents of the document '{}'").format(file_path))
return {'status': status, 'args': file_path, 'data': content}
summary, freq_words, sentiment = text_summary(clean_text, 20)
tags = list(freq_words)[:5] if len(freq_words) > 5 else list(
freq_words)
clean_text_bytes = bytes(clean_text, encoding=encoding)
clean_text_b64str = base64.b64encode(clean_text_bytes).decode(
'utf-8')
hash_object = hashlib.sha512(clean_text_bytes)
hex_dig = hash_object.hexdigest()
content = {
'meta': {
'dir_root': root,
'folder_file': folder,
'filename': file_name,
'extension': file_extension,
'content_sha512_hex': hex_dig,
**pdfinfo
},
'content': clean_text,
'content_base64': clean_text_b64str,
'summary': summary,
'created': datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
'tags': tags,
'sentiment': sentiment
}
logger.debug('Gevent (end parse_pdf): {} - {}'.format(
gevent.getcurrent().name,
time() - t0))
if not content:
status = 'error'
logger.error("Empty content for '{}'".format(file_path))
else:
status = 'ok'
return {'status': status, 'args': file_path, 'data': content}
def on_exception(greenlet):
logger.error("Greenlet '{}' died unexpectedly. Args: '{}'".format(
greenlet, greenlet.args))
if not es.create_index(idx, mappings.get(idx, '')):
logger.error("Error creating index '{}'".format(idx))
return
return
if __name__ == '__main__':
scheduler = GeventScheduler()
config_app = config.get('app')
config_es = config.get('elasticsearch')
interval = config.get('freq_min', 5)
if not config_app:
logger.error('Missing: config > app')
sys.exit(1)
if not config_es:
logger.error('Missing: config > elasticsearch')
sys.exit(1)
es_addr = config_es.get('host', '127.0.01')
es_port = config_es.get('port', 9200)
dir_root = config_app.get('dir_root')
dir_processed = config_app.get('dir_processed')
dir_error = config_app.get('dir_errors')
if not dir_root:
logger.error('Missing: config > app > dir_root')
sys.exit(1)
