# -*- coding: utf-8 -*-
from datetime import datetime
from os import listdir
from os.path import isfile, join
from concurrent import futures
from epoms.db import *
from epoms.entity_extract import EntityExtract
INDEX_NAME = 'epoms'
TIMEOUT = 300
MAX_WORKER = 1
config = EPOMSConfig()
en = EntityExtract()
news = (News().select())
for n in news:
print '>> Extracting Entity [%5d]' % ( n.id )
try:
names = en.extract_name( n.content )
keys = names.keys()
for i in keys:
for j in range(names[i]):
print i
except Exception as exc:
pass
start = new_start
return res
for n in news:
try:
content = n.content
sentences = sent_tokenize(content)
sentences = merge_sentence(sentences, mode)
print "Getting %d groups of text from document %d" % (len(sentences), n.id)
for s in sentences:
names = en.extract_name(s)
keys = names.keys()
for i in range(len(keys)):
name1, created = Name().get_or_create(name=keys[i])
orig = keys[i].replace(" ", "_")
for j in range(len(keys)):
if i != j:
# Save graph to db
name2, created = Name().get_or_create(name=keys[j])
Name_Graph.create(name1=name1, name2=name2, doc_id=n.id)
dest = keys[j].replace(" ", "_")
print "### ", orig, dest
except Exception as exc:
pass
# print '--> Error %s' % n.as_dict()
