# -*- coding: utf-8 -*- import sys import re import os from meneame.obtenerNoticias import ObtenerNoticias from meneame.utils import limpia, limpia1 from meneame.properties import MENEAME_BASE, MENEAME_PENDIENTES, DIROUT from meneame.utils import get_path from hashlib import md5 if len(sys.argv) != 4: print 'python noticiasSql.py <pag_inicial> <pag_final> <published|pending>' sys.exit(1) o = ObtenerNoticias() p = get_path() d = os.listdir(p) if DIROUT not in d: os.mkdir(DIROUT) path = p+'/'+DIROUT+'/' init = int(sys.argv[1]) max = int(sys.argv[2]) status = sys.argv[3] url_news = '' all_news = [] if not status in ('published', 'pending'):
import os import datetime import time # FIXED PARAMETERS MINUTES = 10 # CADA CUANTO SE VA A EJECUTAR NUMBER_OF_DAYS = 1 # CUANTOS DIAS SE VA A EJECUTAR SLEEP = 60*MINUTES end = datetime.datetime.now() + datetime.timedelta(days=NUMBER_OF_DAYS) while datetime.datetime.now() < end: # check if exists output dir path = get_path() d = os.listdir(path) if DIROUT not in d: os.mkdir(DIROUT) stats_base = obtenerEstadisticas(path, MENEAME_BASE_RSS) stats_comentarios = obtenerEstadisticas(path, MENEAME_COMENTARIOS_RSS) stats_pendientes = obtenerEstadisticas(path, MENEAME_PENDIENTES_RSS) stats_base.getStats() stats_comentarios.getStats() stats_pendientes.getStats() print "Durmiendo..." time.sleep(SLEEP) # Sleep 5 minutos