Changeset - 7a4e2c6ec02f
[Not reviewed]
stable
0 2 0
Mads Kiilerich - 4 years ago 2021-09-28 14:40:55
mads@kiilerich.com
setup: support Bleach 4.2 for Python 3.10 support

Changelog shows no significant API changes.

Bleach 3.2 and later are however even more unreasonably paranoid than 3.1, and the test
thus has to be updated and we stop supporting 3.1 .
2 files changed with 2 insertions and 2 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/markup_renderer.py
Show inline comments
 
@@ -24,193 +24,193 @@ Original author and date, and relevant c
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 

	
 
import hashlib
 
import logging
 
import re
 
import traceback
 

	
 
import bleach
 
import markdown as markdown_mod
 
from docutils.core import publish_parts
 
from docutils.parsers.rst import directives
 

	
 
from kallithea.lib import webutils
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class MarkupRenderer(object):
 
    RESTRUCTUREDTEXT_DISALLOWED_DIRECTIVES = ['include', 'meta', 'raw']
 

	
 
    MARKDOWN_PAT = re.compile(r'md|mkdn?|mdown|markdown', re.IGNORECASE)
 
    RST_PAT = re.compile(r're?st', re.IGNORECASE)
 
    PLAIN_PAT = re.compile(r'readme', re.IGNORECASE)
 

	
 
    @classmethod
 
    def _detect_renderer(cls, source, filename):
 
        """
 
        runs detection of what renderer should be used for generating html
 
        from a markup language
 

	
 
        filename can be also explicitly a renderer name
 
        """
 

	
 
        if cls.MARKDOWN_PAT.findall(filename):
 
            return cls.markdown
 
        elif cls.RST_PAT.findall(filename):
 
            return cls.rst
 
        elif cls.PLAIN_PAT.findall(filename):
 
            return cls.rst
 
        return cls.plain
 

	
 
    @classmethod
 
    def _flavored_markdown(cls, text):
 
        """
 
        Github style flavored markdown
 

	
 
        :param text:
 
        """
 

	
 
        # Extract pre blocks.
 
        extractions = {}
 

	
 
        def pre_extraction_callback(matchobj):
 
            digest = hashlib.sha1(matchobj.group(0)).hexdigest()
 
            extractions[digest] = matchobj.group(0)
 
            return "{gfm-extraction-%s}" % digest
 
        pattern = re.compile(r'<pre>.*?</pre>', re.MULTILINE | re.DOTALL)
 
        text = re.sub(pattern, pre_extraction_callback, text)
 

	
 
        # Prevent foo_bar_baz from ending up with an italic word in the middle.
 
        def italic_callback(matchobj):
 
            s = matchobj.group(0)
 
            if list(s).count('_') >= 2:
 
                return s.replace('_', r'\_')
 
            return s
 
        text = re.sub(r'^(?! {4}|\t)\w+_\w+_\w[\w_]*', italic_callback, text)
 

	
 
        # In very clear cases, let newlines become <br /> tags.
 
        def newline_callback(matchobj):
 
            if len(matchobj.group(1)) == 1:
 
                return matchobj.group(0).rstrip() + '  \n'
 
            else:
 
                return matchobj.group(0)
 
        pattern = re.compile(r'^[\w\<][^\n]*(\n+)', re.MULTILINE)
 
        text = re.sub(pattern, newline_callback, text)
 

	
 
        # Insert pre block extractions.
 
        def pre_insert_callback(matchobj):
 
            return '\n\n' + extractions[matchobj.group(1)]
 
        text = re.sub(r'{gfm-extraction-([0-9a-f]{32})\}',
 
                      pre_insert_callback, text)
 

	
 
        return text
 

	
 
    @classmethod
 
    def render(cls, source, filename=None):
 
        """
 
        Renders a given filename using detected renderer
 
        it detects renderers based on file extension or mimetype.
 
        At last it will just do a simple html replacing new lines with <br/>
 

	
 
        >>> MarkupRenderer.render('''<img id="a" style="margin-top:-1000px;color:red" src="http://example.com/test.jpg">''', '.md')
 
        '<p><img id="a" src="http://example.com/test.jpg" style="color: red;"></p>'
 
        '<p><img id="a" src="http://example.com/test.jpg" style=""></p>'
 
        >>> MarkupRenderer.render('''<img class="c d" src="file://localhost/test.jpg">''', 'b.mkd')
 
        '<p><img class="c d"></p>'
 
        >>> MarkupRenderer.render('''<a href="foo">foo</a>''', 'c.mkdn')
 
        '<p><a href="foo">foo</a></p>'
 
        >>> MarkupRenderer.render('''<script>alert(1)</script>''', 'd.mdown')
 
        '&lt;script&gt;alert(1)&lt;/script&gt;'
 
        >>> MarkupRenderer.render('''<div onclick="alert(2)">yo</div>''', 'markdown')
 
        '<div>yo</div>'
 
        >>> MarkupRenderer.render('''<a href="javascript:alert(3)">yo</a>''', 'md')
 
        '<p><a>yo</a></p>'
 
        """
 

	
 
        renderer = cls._detect_renderer(source, filename)
 
        readme_data = renderer(source)
 
        # Allow most HTML, while preventing XSS issues:
 
        # no <script> tags, no onclick attributes, no javascript
 
        # "protocol", and also limit styling to prevent defacing.
 
        return bleach.clean(readme_data,
 
            tags=['a', 'abbr', 'b', 'blockquote', 'br', 'code', 'dd',
 
                  'div', 'dl', 'dt', 'em', 'h1', 'h2', 'h3', 'h4', 'h5',
 
                  'h6', 'hr', 'i', 'img', 'li', 'ol', 'p', 'pre', 'span',
 
                  'strong', 'sub', 'sup', 'table', 'tbody', 'td', 'th',
 
                  'thead', 'tr', 'ul'],
 
            attributes=['class', 'id', 'style', 'label', 'title', 'alt', 'href', 'src'],
 
            styles=['color'],
 
            protocols=['http', 'https', 'mailto'],
 
            )
 

	
 
    @classmethod
 
    def plain(cls, source, universal_newline=True):
 
        """
 
        >>> MarkupRenderer.plain('https://example.com/')
 
        '<br /><a href="https://example.com/">https://example.com/</a>'
 
        """
 
        if universal_newline:
 
            newline = '\n'
 
            source = newline.join(source.splitlines())
 

	
 
        def url_func(match_obj):
 
            url_full = match_obj.group(0)
 
            return '<a href="%(url)s">%(url)s</a>' % ({'url': url_full})
 
        source = webutils.url_re.sub(url_func, source)
 
        return '<br />' + source.replace("\n", '<br />')
 

	
 
    @classmethod
 
    def markdown(cls, source, safe=True, flavored=False):
 
        """
 
        Convert Markdown (possibly GitHub Flavored) to INSECURE HTML, possibly
 
        with "safe" fall-back to plaintext. Output from this method should be sanitized before use.
 

	
 
        >>> MarkupRenderer.markdown('''<img id="a" style="margin-top:-1000px;color:red" src="http://example.com/test.jpg">''')
 
        '<p><img id="a" style="margin-top:-1000px;color:red" src="http://example.com/test.jpg"></p>'
 
        >>> MarkupRenderer.markdown('''<img class="c d" src="file://localhost/test.jpg">''')
 
        '<p><img class="c d" src="file://localhost/test.jpg"></p>'
 
        >>> MarkupRenderer.markdown('''<a href="foo">foo</a>''')
 
        '<p><a href="foo">foo</a></p>'
 
        >>> MarkupRenderer.markdown('''<script>alert(1)</script>''')
 
        '<script>alert(1)</script>'
 
        >>> MarkupRenderer.markdown('''<div onclick="alert(2)">yo</div>''')
 
        '<div onclick="alert(2)">yo</div>'
 
        >>> MarkupRenderer.markdown('''<a href="javascript:alert(3)">yo</a>''')
 
        '<p><a href="javascript:alert(3)">yo</a></p>'
 
        >>> MarkupRenderer.markdown('''## Foo''')
 
        '<h2>Foo</h2>'
 
        >>> print(MarkupRenderer.markdown('''
 
        ...     #!/bin/bash
 
        ...     echo "hello"
 
        ... '''))
 
        <table class="code-highlighttable"><tr><td class="linenos"><div class="linenodiv"><pre>1
 
        2</pre></div></td><td class="code"><div class="code-highlight"><pre><span></span><span class="ch">#!/bin/bash</span>
 
        <span class="nb">echo</span> <span class="s2">&quot;hello&quot;</span>
 
        </pre></div>
 
        </td></tr></table>
 
        """
 
        try:
 
            if flavored:
 
                source = cls._flavored_markdown(source)
 
            return markdown_mod.markdown(
 
                source,
 
                extensions=['markdown.extensions.codehilite', 'markdown.extensions.extra'],
 
                extension_configs={'markdown.extensions.codehilite': {'css_class': 'code-highlight'}})
 
        except Exception:
 
            log.error(traceback.format_exc())
 
            if safe:
 
                log.debug('Falling back to render in plain mode')
 
                return cls.plain(source)
 
            else:
 
                raise
 

	
 
    @classmethod
 
    def rst(cls, source, safe=True):
 
        try:
 
            docutils_settings = dict([(alias, None) for alias in
 
                                cls.RESTRUCTUREDTEXT_DISALLOWED_DIRECTIVES])
 

	
 
            docutils_settings.update({'input_encoding': 'unicode',
setup.py
Show inline comments
 
#!/usr/bin/env python3
 
# -*- coding: utf-8 -*-
 
import os
 
import platform
 
import re
 
import sys
 

	
 
import setuptools
 
# monkey patch setuptools to use distutils owner/group functionality
 
from setuptools.command import sdist
 

	
 

	
 
if sys.version_info < (3, 6):
 
    raise Exception('Kallithea requires Python 3.6 or later')
 

	
 

	
 
here = os.path.abspath(os.path.dirname(__file__))
 

	
 

	
 
def _get_meta_var(name, data, callback_handler=None):
 
    matches = re.compile(r'(?:%s)\s*=\s*(.*)' % name).search(data)
 
    if matches:
 
        s = eval(matches.groups()[0])
 
        if callable(callback_handler):
 
            return callback_handler(s)
 
        return s
 

	
 
_meta = open(os.path.join(here, 'kallithea', '__init__.py'), 'r')
 
_metadata = _meta.read()
 
_meta.close()
 

	
 
def callback(V):
 
    return '.'.join(map(str, V[:3])) + '.'.join(V[3:])
 
__version__ = _get_meta_var('VERSION', _metadata, callback)
 
__license__ = _get_meta_var('__license__', _metadata)
 
__author__ = _get_meta_var('__author__', _metadata)
 
__url__ = _get_meta_var('__url__', _metadata)
 
# defines current platform
 
__platform__ = platform.system()
 

	
 
is_windows = __platform__ in ['Windows']
 

	
 
requirements = [
 
    "alembic >= 1.0.10, < 1.5",
 
    "gearbox >= 0.1.0, < 1",
 
    "waitress >= 0.8.8, < 1.5",
 
    "WebOb >= 1.8, < 1.9",
 
    "backlash >= 0.1.2, < 1",
 
    "TurboGears2 >= 2.4, < 2.5",
 
    "tgext.routes >= 0.2.0, < 1",
 
    "Beaker >= 1.10.1, < 2",
 
    "WebHelpers2 >= 2.0, < 2.1",
 
    "FormEncode >= 1.3.1, < 2.1",
 
    "SQLAlchemy >= 1.2.9, < 1.4",
 
    "Mako >= 0.9.1, < 1.2",
 
    "Pygments >= 2.2.0, < 2.7",
 
    "Whoosh >= 2.7.1, < 2.8",
 
    "celery >= 5, < 5.1",
 
    "Babel >= 1.3, < 2.9",
 
    "python-dateutil >= 2.1.0, < 2.9",
 
    "Markdown >= 2.2.1, < 3.2",
 
    "docutils >= 0.11, < 0.17",
 
    "URLObject >= 2.3.4, < 2.5",
 
    "Routes >= 2.0, < 2.5",
 
    "dulwich >= 0.19.0, < 0.20",
 
    "mercurial >= 5.2, < 5.9",
 
    "decorator >= 4.2.1, < 4.5",
 
    "Paste >= 2.0.3, < 3.5",
 
    "bleach >= 3.0, < 3.1.4",
 
    "bleach >= 3.2, < 4.2",
 
    "Click >= 7.0, < 8",
 
    "ipaddr >= 2.2.0, < 2.3",
 
    "paginate >= 0.5, < 0.6",
 
    "paginate_sqlalchemy >= 0.3.0, < 0.4",
 
    "bcrypt >= 3.1.0, < 3.2",
 
    "pip >= 20.0, < 999",
 
    "chardet >= 3",
 
]
 

	
 
dependency_links = [
 
]
 

	
 
classifiers = [
 
    'Development Status :: 4 - Beta',
 
    'Environment :: Web Environment',
 
    'Framework :: Pylons',
 
    'Intended Audience :: Developers',
 
    'License :: OSI Approved :: GNU General Public License (GPL)',
 
    'Operating System :: OS Independent',
 
    'Programming Language :: Python :: 3.6',
 
    'Programming Language :: Python :: 3.7',
 
    'Programming Language :: Python :: 3.8',
 
    'Topic :: Software Development :: Version Control',
 
]
 

	
 

	
 
# additional files from project that goes somewhere in the filesystem
 
# relative to sys.prefix
 
data_files = []
 

	
 
description = ('Kallithea is a fast and powerful management tool '
 
               'for Mercurial and Git with a built in push/pull server, '
 
               'full text search and code-review.')
 

	
 
keywords = ' '.join([
 
    'kallithea', 'mercurial', 'git', 'code review',
 
    'repo groups', 'ldap', 'repository management', 'hgweb replacement',
 
    'hgwebdir', 'gitweb replacement', 'serving hgweb',
 
])
 

	
 
# long description
 
README_FILE = 'README.rst'
 
try:
 
    long_description = open(README_FILE).read()
 
except IOError as err:
 
    sys.stderr.write(
 
        "[WARNING] Cannot find file specified as long_description (%s): %s\n"
 
        % (README_FILE, err)
 
    )
 
    long_description = description
 

	
 

	
 
sdist_org = sdist.sdist
 
class sdist_new(sdist_org):
 
    def initialize_options(self):
 
        sdist_org.initialize_options(self)
 
        self.owner = self.group = 'root'
 
sdist.sdist = sdist_new
 

	
 
packages = setuptools.find_packages(exclude=['ez_setup'])
 

	
 
setuptools.setup(
 
    name='Kallithea',
 
    version=__version__,
 
    description=description,
 
    long_description=long_description,
 
    keywords=keywords,
 
    license=__license__,
 
    author=__author__,
 
    author_email='kallithea@sfconservancy.org',
 
    dependency_links=dependency_links,
 
    url=__url__,
 
    install_requires=requirements,
 
    classifiers=classifiers,
 
    data_files=data_files,
 
    packages=packages,
 
    include_package_data=True,
 
    message_extractors={'kallithea': [
 
            ('**.py', 'python', None),
 
            ('templates/**.mako', 'mako', {'input_encoding': 'utf-8'}),
 
            ('templates/**.html', 'mako', {'input_encoding': 'utf-8'}),
 
            ('public/**', 'ignore', None)]},
 
    zip_safe=False,
 
    entry_points="""
 
    [console_scripts]
 
    kallithea-api =    kallithea.bin.kallithea_api:main
 
    kallithea-gist =   kallithea.bin.kallithea_gist:main
 
    kallithea-cli =    kallithea.bin.kallithea_cli:cli
 

	
 
    [paste.app_factory]
 
    main = kallithea.config.application:make_app
 
    """,
 
)
0 comments (0 inline, 0 general)