def crawpages(urlp, selector, start=1, limit=-1):
if urlp.find('%d') == -1: return
soup = ''
links = []
index = start
while True:
page = urlp % index
print '-->', page
new_soup = select(BeautifulSoup(urlopen(urlp % index).read()),
selector)
if new_soup is None or len(new_soup) == 0 or new_soup == soup:
print 'repeat or none'
print 'End at %s' % index
break
if limit > 0 and limit + 1 <= index:
print 'limit ', limit
print 'End at %s' % index
break
print getlisttext(new_soup)
if (new_soup[0].name == 'a'):
for link in new_soup:
if ('href' in dict(link.attrs)):
url = urljoin(page, link['href'])
#if url.find("'")!=-1: continue why?
url = url.split('#')[0] # remove location portion
if url[0:4] == 'http':
links.append(url)
else:
links.append(page)
soup = new_soup
index += 1
return links
def get_page_links(url):
soup = Soup(load_page(url))
print 'open %s %s' % (soup.title.string.encode('utf-8'), url)
links = select(soup, 'a.post-article')
#links = soup.findAll('a', href=re.compile(r'^http://duoduovision.diandian.com/post/\d{4}-\d{2}-\d{2}/'))
#return list(set([nomalize_url(a['href']) for a in links]))
return [nomalize_url(a['href']) for a in links]
def get_page_links(url):
soup = Soup(load_page(url))
print 'open %s %s' % (soup.title.string.encode('utf-8'), url)
links = select(soup, 'a.post-article')
#links = soup.findAll('a', href=re.compile(r'^http://duoduovision.diandian.com/post/\d{4}-\d{2}-\d{2}/'))
#return list(set([nomalize_url(a['href']) for a in links]))
return [nomalize_url(a['href']) for a in links]
def crawpages(urlp, selector, start=1, limit=-1):
if urlp.find('%d') == -1: return
soup = ''
links = []
index = start
while True:
page = urlp % index
print '-->', page
new_soup = select(BeautifulSoup(urlopen(urlp % index).read()), selector)
if new_soup is None or len(new_soup)== 0 or new_soup == soup:
print 'repeat or none'
print 'End at %s' % index
break
if limit > 0 and limit + 1 <= index:
print 'limit ', limit
print 'End at %s' % index
break
print getlisttext(new_soup)
if(new_soup[0].name == 'a'):
for link in new_soup:
if ('href' in dict(link.attrs)):
url=urljoin(page, link['href'])
#if url.find("'")!=-1: continue why?
url=url.split('#')[0] # remove location portion
if url[0:4]=='http':
links.append(url)
else: links.append(page)
soup = new_soup
index += 1
return links
def save_pagenates(leaf, limit= -1):
"""抓取有分页的url，他们的URL大多类似"""
url_param = leaf.url
index = 1
url = url_param % (index)
filename = os.path.join(leaf.dir, "%s.html" % (index))
while file_exist("%s\t%s" % (filename, url)):
print '%s is already downloaded.' % (url)
index += 1
url = url_param % (index)
filename = os.path.join(leaf.dir, "%s.html" % (index))
page = urlopen(url).read().decode(encode, 'ignore')
soup = select(BeautifulSoup(page), config[leaf.level]['selector'])
if soup == None or len(soup) == 0: print 'break 1'; return
else: print 'page: ', index, getlisttext(soup)
try:
file = open(filename, 'w').write(page.encode(encode, 'ignore'))
except IOError as err:
print 'error: ', str(err), filename
open(logfile, 'a').write(filename + '\t' + url + '\r\n')
index += 1
url = url_param % (index)
filename = os.path.join(leaf.dir, "%s.html" % (index))
while True:
while file_exist("%s\t%s" % (filename, url)):
print '%s is already downloaded.' % (url)
index += 1
url = url_param % (index)
filename = os.path.join(leaf.dir, "%s.html" % (index))
tmp_page = urlopen(url).read().decode(encode, 'ignore')
tmp_soup = select(BeautifulSoup(tmp_page), config[leaf.level]['selector'])
print tmp_soup
print tmp_page == page, tmp_soup == soup
if tmp_page == page or tmp_soup == soup or tmp_soup == None or len(tmp_soup) == 0: print 'break 1'; break
else: print 'page: ', index, getlisttext(tmp_soup[0])
try:
open(filename, 'w').write(tmp_page.encode(encode, 'ignore'))
except:
print filename
open(logfile, 'a').write(filename + '\t' + url + '\r\n')
page = tmp_page
soup = tmp_soup
index += 1
if limit > 0 and limit < index: break
url = url_param % (index)
filename = os.path.join(leaf.dir, "%s.html" % (index))
def create_children(map):
if map.level >= len(config) - 1:
return
current_config = config[map.level]
# here we create soup by url or soup from the parent 'map'
if map.url != None:
soup = BeautifulSoup(urlopen(map.url).read().decode(encode, 'ignore'))
elif map.soup != None:
soup = map.soup
# we use the 'selector' to create html fragments we called soups
soups = select(soup, current_config['selector'])
# use keywords as a filter
if 'keywords' in current_config:
soups = keyword_filter(soups, current_config['keywords'])
children = []
for i in range(len(soups)):
# if the element is 'a' then we record url and make soup None,
# if the element is sth else, we make url None and soup the
# elements.
if soups[i].name == 'a':
# print soups[i]['href'], soups[i].string
url = soups[i]['href']
# use the content as the title
title = soups[i].string
# if process is in the config, we will use it for url
# transformation.
if 'process' in current_config:
url = current_config['process'](url)
soup = None
else:
# if the name is not 'a', we may use the 'title' in
# config, if 'title' startswith 'css:', we will use the
# element's content as the title, or we use the 'title'
# config directly.
if current_config['title'].startswith('css:'):
title_css = current_config['title'].split(':', 1)[1]
title = select(soups[i], title_css)[0].string
else: title = current_config['title']
url = None
soup = soups[i]
# if 'collapse' is True then, it will not create a new dir
# with title as its name
if 'collapse' not in current_config or current_config['collapse'] == False:
if map.dir is None or map.dir == '': dir = title
else: dir = os.path.join(map.dir, title)
else:
dir = map.dir
children.append(LinkMap(
parent=map,
url=url,
soup=soup,
seletor=config[map.level + 1]['selector'],
title=title,
level=map.level + 1,
dir=utils.formatname(dir)))
# when there is no sub links in the title, we should give another
# link, use the 'default_selector'
if len(children) == 0 and 'default_selector' in current_config.keys():
soups = select(soup, current_config['default_selector'])
url = soups[0]['href']
if 'process' in current_config: url = current_config['process'](url)
if 'collapse' not in current_config or current_config['collapse'] == False:
dir = os.path.join(map.dir, title)
else:
dir = map.dir
children.append(LinkMap(
parent=map,
url=url,
seletor=config[map.level + 1]['selector'],
title=soups[0].string,
level=map.level + 1,
dir=utils.formatname(dir)))
map.children = children
for child in map.children:
create_children(child)
def get_img_links(url):
soup = Soup(load_page(url))
imgs = select(soup, 'a.post-meidaurl img')
return [img['src'] for img in imgs]
def get_img_links(url):
soup = Soup(load_page(url))
imgs = select(soup, 'a.post-meidaurl img')
return [img['src'] for img in imgs]
