def _get_main_image_dict(self, html_string):
""" Try to find the main image in the given html string """
# Transform the html string into an lxml tree
doc = build_doc(html_string)
# Build list of img tags
tags = []
for t in self.tags(doc, 'img'):
tags.append(t)
# Convert image tags into informational dicts
images = [self._img_tag_to_dict(t) for t in tags]
# If we have no images we return an empty dict
if not images:
return {}
# If there is only one image then we will use it as the main image
if len(images) == 1:
return images[0]
# If we make it here then we have more than 1 image. We will return the
# largest image.
largest_pix_area = 0
largest_image_dict = {}
for i in images:
if i['pix-area'] > largest_pix_area:
largest_image_dict = i
largest_pix_area = i['pix-area']
return largest_image_dict
def _get_main_image_dict(self, html_string):
""" Try to find the main image in the given html string """
# Transform the html string into an lxml tree
doc = build_doc(html_string)
# Build list of img tags
tags = []
for t in self.tags(doc, 'img'):
tags.append(t)
# Convert image tags into informational dicts
images = [self._img_tag_to_dict(t) for t in tags]
# If we have no images we return an empty dict
if not images:
return {}
# If there is only one image then we will use it as the main image
if len(images) == 1:
return images[0]
# If we make it here then we have more than 1 image. We will return the
# largest image.
largest_pix_area = 0
largest_image_dict = {}
for i in images:
if i['pix-area'] > largest_pix_area:
largest_image_dict = i
return largest_image_dict
def _parse(self, input):
doc, self.encoding = build_doc(input)
doc = html_cleaner.clean_html(doc)
base_href = self.options.get("url", None)
if base_href:
doc.make_links_absolute(base_href, resolve_base_href=True)
else:
doc.resolve_base_href()
return doc
def _parse(self, input):
doc, self.encoding = build_doc(input)
doc = html_cleaner.clean_html(doc)
base_href = self.options.get('url', None)
if base_href:
doc.make_links_absolute(base_href, resolve_base_href=True)
else:
doc.resolve_base_href()
return doc
def _parse(self, input): doc = build_doc(input) doc = html_cleaner.clean_html(doc) base_href = self.options['url'] if base_href: doc.make_links_absolute(base_href, resolve_base_href=True) else: doc.resolve_base_href() return doc
def process_url(f,url,partial_html=False,keep_going=True):
text_input = f.read()
clean_doc = Document(text_input,
debug=options.verbose,
url=url)
clean_doc.summary(partial_html=partial_html,encoding=enc)
dirty_doc = build_doc(text_input)
del text_input
if keep_going:
try:
next_page = urlparse.urljoin(url,find_next_page(dirty_doc)[0])
if next_page not in visited_urls:
if options.verbose: sys.stderr.write(next_page+"\n")
if clean_doc.html.tag == 'html':
clean_doc.html[0][-1].append(process_url(urllib.urlopen(next_page),next_page,partial_html=True).html[0])
else:
clean_doc.html.append(process_url(urllib.urlopen(next_page),next_page,partial_html=True).html[0])
except IndexError:
pass
else:
visited_urls.insert([next_page])
return clean_doc
def __init__(self, url, text=None, page=1, min_article_length=250, min_article_percentage=0.075):
"""
:param url: the url of the document
:param text: optionally the string value of the page may be passed in
:param page: if this is one in a series of documents in an article this should be set
:param min_article_length: if an article is less than this number of characters it's not an article
:param min_article_percentage: an article must be this % of the text on the page
"""
self.url = url
self.page = page
self._article = None
self.min_article_length = min_article_length
self.min_article_percentage = min_article_percentage
if text:
self.text = text
else:
self.text = requests.get(url).text
# parses the HTML and cleans it up removing elements this doesn't want to deal with (e.g., head, script, form)
doc, self.encoding = build_doc(self.text)
doc = html_cleaner.clean_html(doc)
doc.make_links_absolute(self.url, resolve_base_href=True)
self.html = doc
