Skip to content

crawlersNoticias
crawlersNoticias
Commit 30d91be1 authored by Renán Sosa Guillen's avatar Renán Sosa Guillen
Browse files
Options

crawlers

parent 4c9562a7
Show whitespace changes
Inline Side-by-side
Showing with 657 additions and 433 deletions
descarga_por_dia/laJornada/laJornada/spiders/noticias.py
View file @ 30d91be1
from datetime import date, datetime, timedelta, tzinfo, time
import scrapy, re
from datetime import date, datetime, timedelta, tzinfo, time
from collections import OrderedDict
## scrapy crawl noticias -t json --nolog -o noticias.json -a year=2017 -a month=3 -a day=22
"""
scrapy crawl noticias -t json --nolog -o noticias.json -a year=2017 -a month=3 -a day=22
"""
TAG_RE = re.compile(r'<[^>]+>')
def remove_tags(text):
return TAG_RE.sub('', text)
# re: r'(Twitter:\s+)?(@[\w.%+-]+.)?'
TW_RE = re.compile(r"""
(Twitter: # inicio de bloque, contiene la cadena 'Twitter:' (case insensitive)
\s+ # cualquier espacio (\t\n\r\f), una o mas ocurrencias
)? # fin de bloque, ninguna o una ocurrencia del bloque
(@ # inicio de bloque, contiene caracter '@'
[\w.%+-]+ # cualquier caracter alfanumerico mas los signos (.%+-), una o mas ocurrencias
. # cualquier caracter, excepto '\n'
)? # fin de bloque, ninguna o una ocurrencia del bloque
""", re.X|re.I) # banderas: verbose|case insensitive
# re: r'(Facebook|Vk):\s+[\w.%+-]+.'
FB_RE = re.compile(r"""
(Facebook|Vk) # bloque, contiene la cadena 'Facebook' o 'Vk' (case insensitive)
: # contiene el caracter ':'
\s+ # cualquier espacio (\t\n\r\f), una o mas ocurrencias
[\w.%+-]+ # cualquier caracter alfanumerico mas los signos (.%+-), una o mas ocurrencias
. # cualquier caracter, excepto '\n'
""", re.X|re.I) # banderas: verbose|case insensitive
# re: r'\(?(Foro:\s+)?(https?:\/\/)?([w{3}.])?[\w%+-]+(\.[a-zA-Z]{2,6}){1,2}[/\w.#$%&+-]*\)?.'
URL_RE = re.compile(r"""
\(? # contiene o no caracter '(', ninguna o una vez
(Foro: # inicio de bloque, contiene la cadena 'Foro:' (case insensitive)
\s+ # cualquier espacio (\t\n\r\f), una o mas ocurrencias
)? # fin de bloque, ninguna o una ocurrencia del bloque
(http # inicio de bloque, contiene cadena 'http'
s? # contiene o no caracter 's'
:\/\/ # contiene cadena '://'
)? # fin de bloque, ninguna o una ocurrencia del bloque
([w{3}.])? # el caracter 'w' tres veces y/o punto (www.), ninguna o una vez
[\w%+-]+ # cualquier caracter alfanumerico mas los signos (%+-), una o mas ocurrencias
(\. # inicio de bloque, contiene caracter '.'
[a-zA-Z]{2,6} # 2 a 6 letras, minusculas o mayusculas
){1,2} # fin de bloque, bloque se repite de 1 a 2 veces
[/\w.#$%&+-]* # seguido de '/', cualquier caracter alfanumerico mas los signos (.#$%&+-), cero o mas ocurrencias
\)? # contiene o no caracter ')', ninguna o una vez
. # cualquier caracter, excepto '\n'
""", re.X|re.I) # banderas: verbose|case insensitive
# re: r'[\w.-]+@[\w-]+(\.[a-zA-Z]{2,6}){1,2}\s?'
EMAIL_RE = re.compile(r"""
[\w.-]+ # cualquier caracter alfanumerico mas los signos (.-), una o mas repeticiones
@ # seguido de '@'
[\w-]+ # cualquier caracter alfanumerico mas el signo '-', una o mas repeticiones
(\. # inicio de bloque, contiene '.'
[a-zA-Z]{2,6} # 2 a 6 letras, minusculas o mayusculas
){1,2} # fin de bloque, bloque se repite de 1 a 2 veces
\s? # cualquier espacio (\t\n\r\f), ninguna o una coincidencia
""", re.X|re.I) # banderas: verbose|case insensitive
DIVP_RE = re.compile(r'(<div class="(credito-(autor|titulo)|hemero)">.*?<\/div>|<p class="s-s">.{,35}<\/p>|<span class="loc">.*?<\/span>)', re.S)
TRANSLATION_RE = re.compile(r'Traducci.n.*', re.I|re.S)
def clean_raw(rawText):
text = rawText.replace("* * *", '')
text = DIVP_RE.sub('', text)
text = TRANSLATION_RE.sub('', text)
return text
def text_cleaning(text):
"""
Function for cleaning news text
"""
"""
Elimina los espacios dobles, triples o mayores, innecesarios dentro del texto. Primero divide el texto de
acuerdo a los saltos de linea, despues divide cada segmento en palabras sin tomar en cuenta espacios. Luego
las palabras son agrupadas en nuevos segmentos con un solo espacio entre ellas y se agregan los saltos de
linea necesarios.
"""
newText = ''
counter = 0
text = text.replace(u'\u0164', '')
text = text.replace("Afp", '')
for segment in text.split("\n"):
counter += 1
if counter == 1:
newText += " ".join(segment.split())
elif counter > 1:
newText += "\n" + " ".join(segment.split())
"""---------------------------------------------------------------------------------------------------"""
"""
Elimina del texto la info de facebook, twitter, foro y correo electronico.
"""
newText = TW_RE.sub('', newText)
newText = FB_RE.sub('', newText)
newText = EMAIL_RE.sub('', newText)
newText = URL_RE.sub('', newText)
newText = TRANSLATION_RE.sub('', newText)
"""---------------------------------------------------------------------------------------------------"""
return newText
class UTC(tzinfo):
"""clase para el 'time zone' (zona horaria)"""
......@@ -52,6 +147,10 @@ class QuotesSpider(scrapy.Spider):
self.comparison_date_7 = date(2009, 2, 15)
self.date = date(int(year), int(month), int(day))
self.parse_month = {'enero': 1, 'febrero': 2, 'marzo': 3, 'abril': 4,
'mayo': 5, 'junio': 6, 'julio': 7, 'agosto': 8,
'septiembre': 9, 'octubre': 10, 'noviembre': 11, 'diciembre': 12}
# self.section_list = ['opinion', 'politica', 'economia', 'mundo', 'estados', 'ciencias',
# 'capital', 'sociedad', 'cultura', 'espectaculos', 'deportes']
......@@ -63,12 +162,12 @@ class QuotesSpider(scrapy.Spider):
# else:
# yield scrapy.Request(url=self.baseURL+section, callback=self.parse_2)
if self.date <= self.comparison_date_2:
section_list = ['index.html', 'edito.html', 'opinion.html', 'correo.html', 'politica.html',
section_list = ['index.html', 'edito.html', 'opinion.html', 'politica.html',
'economia.html', 'cultura.html', 'espectaculos.html', 'estados.html',
'capital.html', 'mundo.html', 'soc-jus.html', 'deportes.html']
parse_s = {'index.html': 'Portada', 'edito.html': 'Editorial', 'opinion.html': 'Opinion',
'correo.html': 'Correo', 'politica.html': 'Politica', 'economia.html': 'Economia',
'politica.html': 'Politica', 'economia.html': 'Economia',
'cultura.html': 'Cultura', 'espectaculos.html': 'Espectaculos', 'estados.html': 'Estados',
'capital.html': 'Capital', 'mundo.html': 'Mundo', 'soc-jus.html': 'Sociedad',
'deportes.html': 'Deportes'}
......@@ -79,7 +178,7 @@ class QuotesSpider(scrapy.Spider):
item['date'] = datetime.combine(self.date, time()).replace(tzinfo=self.tz).isoformat('T')
item['topic'] = parse_s[s]
if s == 'edito.html' or s == 'correo.html':
if s == 'edito.html':
request = scrapy.Request(url=self.baseURL+s, callback=self.parse_item)
else:
request = scrapy.Request(url=self.baseURL+s, callback=self.parse)
......@@ -88,20 +187,20 @@ class QuotesSpider(scrapy.Spider):
yield request
elif self.date > self.comparison_date_2 and self.date <= self.comparison_date_3:
section_list = ['index.html', 'edito.html', 'opinion.html', 'correo.html', 'politica.html',
section_list = ['index.html', 'edito.html', 'opinion.html', 'politica.html',
'economia.html', 'cultura.html', 'espectaculos.html', 'estados.html',
'capital.html', 'mundo.html', 'soc-jus.html', 'deportes.html',
'index.php', 'edito.php', 'opinion.php', 'correo.php', 'politica.php',
'index.php', 'edito.php', 'opinion.php', 'politica.php',
'economia.php', 'cultura.php', 'espectaculos.php', 'estados.php',
'capital.php', 'mundo.php', 'soc-jus.php', 'deportes.php']
parse_s = {'index.html': 'Portada', 'edito.html': 'Editorial', 'opinion.html': 'Opinion',
'correo.html': 'Correo', 'politica.html': 'Politica', 'economia.html': 'Economia',
'politica.html': 'Politica', 'economia.html': 'Economia',
'cultura.html': 'Cultura', 'espectaculos.html': 'Espectaculos', 'estados.html': 'Estados',
'capital.html': 'Capital', 'mundo.html': 'Mundo', 'soc-jus.html': 'Sociedad',
'deportes.html': 'Deportes',
'index.php': 'Portada', 'edito.php': 'Editorial', 'opinion.php': 'Opinion',
'correo.php': 'Correo', 'politica.php': 'Politica', 'economia.php': 'Economia',
'politica.php': 'Politica', 'economia.php': 'Economia',
'cultura.php': 'Cultura', 'espectaculos.php': 'Espectaculos', 'estados.php': 'Estados',
'capital.php': 'Capital', 'mundo.php': 'Mundo', 'soc-jus.php': 'Sociedad',
'deportes.php': 'Deportes'}
......@@ -250,9 +349,9 @@ class QuotesSpider(scrapy.Spider):
def parse_5(self, response):
if ( response.url[:response.url.rfind('/')+1] == self.baseURL ): # verifica que se conserva la misma URL base
if response.url[:response.url.rfind('/')+1] == self.baseURL: # verifica que se conserva la misma URL base
section = response.url[response.url.rfind('/')+1:]
if ( section == 'opinion' ): # la seccion 'opinion' tiene una estructura diferente a las otras
if section == 'opinion': # la seccion 'opinion' tiene una estructura diferente a las otras
path_list = ['//*[@id="columnas"]/p/a/@href',
'//*[@id="opinion"]/p/a/@href']
else:
......@@ -266,66 +365,80 @@ class QuotesSpider(scrapy.Spider):
def parse_6(self, response):
if ( response.url[:response.url.rfind('/')+1] == self.baseURL ):
if response.url[:response.url.rfind('/')+1] == self.baseURL:
# linkSet = set()
# path_list = ['//*[@class="itemfirst"]/div/a/@href', '//*[@class="item start"]/div/a/@href',
# '//*[@class="item"]/div/a/@href']
#
# for path in path_list:
# for link in response.xpath(path).extract():
# if link not in linkSet:
# linkSet.add(link)
# yield scrapy.Request(url=self.baseURL+link, callback=self.parse_item_4)
linkSet = set()
path_list = ['//*[@class="itemfirst"]/div/a/@href', '//*[@class="item start"]/div/a/@href',
'//*[@class="item"]/div/a/@href']
for path in path_list:
for link in response.xpath(path).extract():
if link not in linkSet:
linkLst = []
linkLst.extend(response.xpath('//*[@class="itemfirst"]/div/a/@href').extract())
linkLst.extend(response.xpath('//*[@class="item start"]/div/a/@href').extract())
linkLst.extend(response.xpath('//*[@class="item"]/div/a/@href').extract())
for l in linkLst:
link = self.baseURL + l
if not link in linkSet:
linkSet.add(link)
yield scrapy.Request(url=self.baseURL+link, callback=self.parse_item_4)
yield scrapy.Request(url=link, callback=self.parse_item_4)
def parse_item(self, response):
"""
FECHAS <= 2004-12-12
"""
item = response.meta['item']
flag = True
text = ''
try:
title = response.xpath('//font[@size="5"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//font[@size="5"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="5"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="5"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="5"]').extract()[1]
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="5"]').extract()[1])
item['title'] = title
except:
try:
title = response.xpath('//font[@size="4"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//font[@size="4"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="4"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="4"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="4"][1]').extract()[1]
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="4"][1]').extract()[1])
item['title'] = title
except:
try:
title = response.xpath('//font[@size="3"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//font[@size="3"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="3"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="3"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//p/font[@size="3"][1]').extract()[1]
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//p/font[@size="3"][1]').extract()[1])
item['title'] = title
except:
try:
title = response.xpath('//font[@size="+1"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//font[@size="+1"]').extract_first())
item['title'] = title
except:
try:
title = response.xpath('//font[@size="+0"]').extract_first()
item['title'] = remove_tags(title)
title = remove_tags(response.xpath('//font[@size="+0"]').extract_first())
item['title'] = title
except:
if self.date <= date(1999, 10, 3): # en esta fecha hay un cambio respecto a las otras en cuanto al html de la pag
try:
......@@ -340,34 +453,86 @@ class QuotesSpider(scrapy.Spider):
if flag:
if self.date <= self.comparison_date_1:
"""
FECHAS > 1999-10-03 Y FECHAS <= 2001-12-07
"""
for p in response.css('p').extract():
text += remove_tags(p).replace('\r','') ## no toma en cuenta los primeros indices donde esta el titulo
text = text.replace('\t','')
# text += remove_tags(p).replace('\r','') ## no toma en cuenta los primeros indices donde esta el titulo
# text = text.replace('\t','')
p = clean_raw(p)
newsText = remove_tags(p)
text += text_cleaning(newsText)
m = re.search(title, text)
if title[-1] == "?": text = text[m.end()+1:]
else: text = text[m.end():]
text = text.lstrip("\n")
text = text.rstrip("\n")
elif self.date > self.comparison_date_1 and self.date <= self.comparison_date_3:
"""
FECHAS > 2001-12-07 Y FECHAS <= 2003-04-25
"""
for p in response.xpath('//table[@bordercolor="#CCCCCC"]').css('p').extract():
text += remove_tags(p).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p).replace('\r','')
# text = text.replace('\t','')
p = clean_raw(p)
newsText = remove_tags(p)
text += text_cleaning(newsText)
m = re.search(title, text)
if title[-1] == "?": text = text[m.end()+1:]
else: text = text[m.end():]
text = text.lstrip("\n")
text = text.rstrip("\n")
elif self.date > self.comparison_date_3 and self.date <= self.comparison_date_4:
"""
FECHAS > 2003-04-25 Y FECHAS <= 2004-11-16
"""
p = response.css('p').extract()
for i in range(0, len(p)):
text += remove_tags(p[i]).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p[i]).replace('\r','')
# text = text.replace('\t','')
aux = clean_raw(p[i])
newsText = remove_tags(aux).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
elif self.date > self.comparison_date_4 and self.date <= self.comparison_date_5:
"""
FECHAS > 2004-11-16 Y FECHAS <= 2004-12-12
"""
p = response.css('p').extract()
for i in range(3, len(p)):
text += remove_tags(p[i]).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p[i]).replace('\r','')
# text = text.replace('\t','')
aux = clean_raw(p[i])
newsText = remove_tags(aux).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
if text == '':
for i in range(0, len(p)):
text += remove_tags(p[i]).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p[i]).replace('\r','')
# text = text.replace('\t','')
aux = clean_raw(p[i])
newsText = remove_tags(aux).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
else:
text = remove_tags(response.body)
text = text[len(title):]
"""
FECHAS <= 1999-10-03
"""
# text = remove_tags(response.body)
# text = text[len(title):]
m = re.search(title, response.body)
body = response.body[m.end():]
body = clean_raw(body)
newsText = remove_tags(body).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
item['text'] = text
item['url'] = response.url
......@@ -376,27 +541,39 @@ class QuotesSpider(scrapy.Spider):
def parse_item_2(self, response):
"""
FECHAS > 2004-12-12 Y FECHAS <= 2005-01-31
"""
item = response.meta['item']
text = ''
title_list = []
title_list.extend(response.xpath('//*[@id="contenido"]/h1/text()').extract())
title_list.extend(response.xpath('//h1/text()').extract())
for t in title_list:
if t is not None or t != '':
# titleLst = []
# titleLst.extend(response.xpath('//*[@id="contenido"]/h1/text()').extract())
# titleLst.extend(response.xpath('//h1/text()').extract())
titleSet = set()
titleSet.add(response.xpath('//*[@id="contenido"]/h1').extract_first())
titleSet.add(response.xpath('//h1').extract_first())
for t in titleSet:
if t is not None and t != '':
title = remove_tags(t).replace('\r','')
title = title.replace('\t','')
item['title'] = title
p = response.css('p').extract()
for i in range(4, len(p)):
text += remove_tags(p[i]).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p[i]).replace('\r','')
# text = text.replace('\t','')
newsText = remove_tags(p[i]).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
if text == '':
for i in range(0, len(p)):
text += remove_tags(p[i]).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p[i]).replace('\r','')
# text = text.replace('\t','')
newsText = remove_tags(p[i]).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
item['text'] = text
item['url'] = response.url
......@@ -405,21 +582,36 @@ class QuotesSpider(scrapy.Spider):
def parse_item_3(self, response):
"""
FECHAS > 2005-01-31 Y FECHAS <= 2009-02-15
"""
item = NoticiasItem()
text = ''
titleSet = set()
# item['date'] = self.date
item['date'] = datetime.combine(self.date, time()).replace(tzinfo=self.tz).isoformat('T')
title = response.xpath('//*[@class="documentContent"]/h1[@class="title"]/text()').extract()
if ( len(title) > 0 ):
item['title'] = title[0]
else:
item['title'] = response.xpath('//*[@class="documentContent"]/h1/text()').extract_first()
# title = response.xpath('//*[@class="documentContent"]/h1[@class="title"]/text()').extract()
# if len(title) > 0:
# item['title'] = title[0]
# else:
# item['title'] = response.xpath('//*[@class="documentContent"]/h1/text()').extract_first()
titleSet.add(response.xpath('//*[@class="documentContent"]/h1[@class="title"]').extract_first())
titleSet.add(response.xpath('//*[@class="documentContent"]/h1').extract_first())
for t in titleSet:
if t is not None and t != '':
title = remove_tags(t).replace('\r','')
title = title.replace('\t','')
item['title'] = title
item['topic'] = response.xpath('//*[@id="portal-breadcrumbs"]/a[2]/text()').extract_first()
for p in response.xpath('//*[@class="documentContent"]/p').extract():
text += remove_tags(p).replace('\r','')
text = text.replace('\t','')
# text += remove_tags(p).replace('\r','')
# text = text.replace('\t','')
newsText = remove_tags(p).lstrip("\n")
newsText = newsText.rstrip("\n")
text += text_cleaning(newsText)
item['text'] = text
item['url'] = response.url
......@@ -428,23 +620,55 @@ class QuotesSpider(scrapy.Spider):
def parse_item_4(self, response):
"""
FECHAS > 2009-02-15
"""
d = response.xpath('//*[@class="main-fecha"]/text()').extract_first()
d = d.replace('de', '').replace(' ', ' ').split(' ')
newsDate = date(int(d[3]), self.parse_month[d[2].lower()], int(d[1]))
if newsDate == self.date:
item = NoticiasItem()
text = ''
# path_list = ['//*[@class="col"]/p', '//*[@class="col col1"]/p', '//*[@class="col col2"]/p']
path_list = ['//*[@class="col"]', '//*[@class="col col1"]', '//*[@class="col col2"]']
# path_list = ['//*[@class="col"]', '//*[@class="col col1"]', '//*[@class="col col2"]']
textLst = []
textLst.extend(response.xpath('//*[@class="col"]').extract())
textLst.extend(response.xpath('//*[@class="col col1"]').extract())
textLst.extend(response.xpath('//*[@class="col col2"]').extract())
# item['date'] = self.date
item['date'] = datetime.combine(self.date, time()).replace(tzinfo=self.tz).isoformat('T')
item['date'] = datetime.combine(newsDate, time()).replace(tzinfo=self.tz).isoformat('T')
item['title'] = remove_tags(response.xpath('//*[@class="cabeza"]').extract_first())
item['topic'] = response.xpath('//*[@class="breadcrumb gui"]/span[2]/a/text()').extract_first()
for path in path_list:
for p in response.xpath(path).extract():
text += remove_tags(p).replace('\r','')
text = text.replace('\t','')
author = response.xpath('//*[@class="credito-autor"]/text()').extract_first()
if author is None or author == '':
author = response.xpath('//*[@class="credito-articulo"]/text()').extract_first()
item['author'] = author
location = remove_tags(response.xpath('//p[@class="s-s"]').extract_first())
if location is not None and location != '' and len(location) <= 35:
item['location'] = location
else:
item['location'] = None
for p in textLst:
# text += remove_tags(p).replace('\r', '')
# text = text.replace('\t', '')
p = clean_raw(p)
# newsText = remove_tags(p).lstrip("\n")
# newsText = newsText.rstrip("\n")
# text += text_cleaning(newsText)
text += remove_tags(p)
text = text.lstrip("\n")
text = text.rstrip("\n")
text = text_cleaning(text)
item['text'] = text
item['url'] = response.url
# print item['title']
# print 'title: ' + item['title'] + '\nurl: ' + item['url'] + '\n'
yield item
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment