mirror of
https://github.com/Hopiu/linkchecker.git
synced 2026-03-22 17:00:25 +00:00
67 lines
2.2 KiB
Python
67 lines
2.2 KiB
Python
# Copyright (C) 2000-2018 Petr Dlouhy
|
|
#
|
|
# This program is free software; you can redistribute it and/or modify
|
|
# it under the terms of the GNU General Public License as published by
|
|
# the Free Software Foundation; either version 2 of the License, or
|
|
# (at your option) any later version.
|
|
#
|
|
# This program is distributed in the hope that it will be useful,
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
# GNU General Public License for more details.
|
|
#
|
|
# You should have received a copy of the GNU General Public License along
|
|
# with this program; if not, write to the Free Software Foundation, Inc.,
|
|
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
|
"""
|
|
HTML parser implemented using Beautiful Soup and html.parser.
|
|
|
|
USAGE
|
|
|
|
Two functions are provided, one to make a BeautifulSoup object from markup and
|
|
another to call a handler's callbacks for each element in a BeautifulSoup
|
|
object it can process.
|
|
|
|
The used callback of a handler is:
|
|
|
|
- Start tag: <tag {attr1:value1, attr2:value2, ..}>
|
|
def start_element (tag, attrs, text, line, column)
|
|
@param tag: tag name
|
|
@type tag: string
|
|
@param attrs: tag attributes
|
|
@type attrs: dict
|
|
@param text: element text
|
|
@type tag: string
|
|
@param line: tag line number
|
|
@type tag: integer
|
|
@param column: tag column number
|
|
@type tag: integer
|
|
|
|
EXAMPLE
|
|
|
|
# Create a new BeautifulSoup object.
|
|
soup = htmlutil.htmlsoup.make_soup("<html><body>Blubb</body></html>")
|
|
# Process the soup with the chosen handler as a parameter.
|
|
htmlutil.htmlsoup.proces_soup(handler, soup)
|
|
|
|
"""
|
|
|
|
from warnings import filterwarnings
|
|
|
|
filterwarnings("ignore",
|
|
message="The soupsieve package is not installed. CSS selectors cannot be used.",
|
|
category=UserWarning, module="bs4")
|
|
|
|
from bs4 import BeautifulSoup
|
|
|
|
|
|
def make_soup(markup, from_encoding=None):
|
|
return BeautifulSoup(markup, "html.parser", from_encoding=from_encoding,
|
|
multi_valued_attributes=None)
|
|
|
|
def process_soup(handler, soup):
|
|
for element in soup.find_all(True):
|
|
handler.start_element(
|
|
element.name, element.attrs, element.text.strip(),
|
|
element.sourceline,
|
|
None if element.sourcepos is None else element.sourcepos + 1)
|