You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
140 lines
3.8 KiB
Python
140 lines
3.8 KiB
Python
import logging
|
|
import re
|
|
|
|
from lxml.html import document_fromstring
|
|
from lxml.html import HTMLParser
|
|
from lxml.html import tostring
|
|
|
|
from cleaners import clean_attributes
|
|
from cleaners import normalize_spaces
|
|
from encoding import get_encoding
|
|
|
|
logging.getLogger().setLevel(logging.DEBUG)
|
|
utf8_parser = HTMLParser(encoding='utf-8')
|
|
|
|
|
|
LOG = logging.getLogger()
|
|
|
|
|
|
def build_doc(page):
|
|
"""Requires that the `page` not be None"""
|
|
if page is None:
|
|
LOG.error("Page content is None, can't build_doc")
|
|
return ''
|
|
if isinstance(page, unicode):
|
|
page_unicode = page
|
|
else:
|
|
enc = get_encoding(page)
|
|
page_unicode = page.decode(enc, 'replace')
|
|
doc = document_fromstring(
|
|
page_unicode.encode('utf-8', 'replace'),
|
|
parser=utf8_parser)
|
|
return doc
|
|
|
|
|
|
def js_re(src, pattern, flags, repl):
|
|
return re.compile(pattern, flags).sub(src, repl.replace('$', '\\'))
|
|
|
|
|
|
def normalize_entities(cur_title):
|
|
entities = {
|
|
u'\u2014': '-',
|
|
u'\u2013': '-',
|
|
u'—': '-',
|
|
u'–': '-',
|
|
u'\u00A0': ' ',
|
|
u'\u00AB': '"',
|
|
u'\u00BB': '"',
|
|
u'"': '"',
|
|
}
|
|
for c, r in entities.iteritems():
|
|
if c in cur_title:
|
|
cur_title = cur_title.replace(c, r)
|
|
|
|
return cur_title
|
|
|
|
|
|
def norm_title(title):
|
|
return normalize_entities(normalize_spaces(title))
|
|
|
|
|
|
def get_title(doc):
|
|
title_node = doc.find('.//title')
|
|
|
|
if not title_node:
|
|
return '[no-title]'
|
|
title = title_node.text
|
|
return norm_title(title)
|
|
|
|
|
|
def add_match(collection, text, orig):
|
|
text = norm_title(text)
|
|
if len(text.split()) >= 2 and len(text) >= 15:
|
|
if text.replace('"', '') in orig.replace('"', ''):
|
|
collection.add(text)
|
|
|
|
|
|
def shorten_title(doc):
|
|
title_node = doc.find('.//title')
|
|
if not title_node:
|
|
return ''
|
|
|
|
title = title_node.text
|
|
title = orig = norm_title(title)
|
|
|
|
candidates = set()
|
|
|
|
for item in ['.//h1', './/h2', './/h3']:
|
|
for e in list(doc.iterfind(item)):
|
|
if e.text:
|
|
add_match(candidates, e.text, orig)
|
|
if e.text_content():
|
|
add_match(candidates, e.text_content(), orig)
|
|
|
|
for item in ['#title', '#head', '#heading', '.pageTitle', '.news_title',
|
|
'.title', '.head', '.heading', '.contentheading', '.small_header_red']:
|
|
for e in doc.cssselect(item):
|
|
if e.text:
|
|
add_match(candidates, e.text, orig)
|
|
if e.text_content():
|
|
add_match(candidates, e.text_content(), orig)
|
|
|
|
if candidates:
|
|
title = sorted(candidates, key=len)[-1]
|
|
else:
|
|
for delimiter in [' | ', ' - ', ' :: ', ' / ']:
|
|
if delimiter in title:
|
|
parts = orig.split(delimiter)
|
|
if len(parts[0].split()) >= 4:
|
|
title = parts[0]
|
|
break
|
|
elif len(parts[-1].split()) >= 4:
|
|
title = parts[-1]
|
|
break
|
|
else:
|
|
if ': ' in title:
|
|
parts = orig.split(': ')
|
|
if len(parts[-1].split()) >= 4:
|
|
title = parts[-1]
|
|
else:
|
|
title = orig.split(': ', 1)[1]
|
|
|
|
if not 15 < len(title) < 150:
|
|
return orig
|
|
|
|
return title
|
|
|
|
|
|
def get_body(doc):
|
|
[elem.drop_tree() for elem in doc.xpath('.//script | .//link | .//style')]
|
|
raw_html = unicode(tostring(doc.body or doc))
|
|
cleaned = clean_attributes(raw_html)
|
|
try:
|
|
#BeautifulSoup(cleaned) #FIXME do we really need to try loading it?
|
|
return cleaned
|
|
except Exception: # FIXME find the equivalent lxml error
|
|
logging.error("cleansing broke html content: %s\n---------\n%s" % (
|
|
raw_html,
|
|
cleaned))
|
|
return raw_html
|