view tests/test-url.py @ 14927:2aa3e07b2f07

posix, windows: introduce cachestat

This class contains a stat result, and possibly other file info to reliably
determine between two points in time whether a file has changed.

Uniquely identifying a file gives us that reliability because we either
atomic rename or append. So one of two will happen: the file 'id' will change,
or the size of the file will change.

posix implements it simply by calling os.stat() and checking if the result
has st_ino.

For now on Windows we always assume the path is uncacheable. This can be
improved on NTFS due to file IDs: http://msdn.microsoft.com/en-us/library/aa363788(v=vs.85).aspx

So we need to find out if a file path is on an NTFS drive, for that we have:

- GetVolumeInformation, which unfortunately only works with a root path (but is available on XP)
- GetVolumeInformationByHandleW, works on a full file path but requires Vista or higher

import sys

def check(a, b):
    if a != b:
        print (a, b)

def cert(cn):
    return dict(subject=((('commonName', cn),),))

from mercurial.sslutil import _verifycert

# Test non-wildcard certificates
check(_verifycert(cert('example.com'), 'example.com'),
      None)
check(_verifycert(cert('example.com'), 'www.example.com'),
      'certificate is for example.com')
check(_verifycert(cert('www.example.com'), 'example.com'),
      'certificate is for www.example.com')

# Test wildcard certificates
check(_verifycert(cert('*.example.com'), 'www.example.com'),
      None)
check(_verifycert(cert('*.example.com'), 'example.com'),
      'certificate is for *.example.com')
check(_verifycert(cert('*.example.com'), 'w.w.example.com'),
      'certificate is for *.example.com')

# Test subjectAltName
san_cert = {'subject': ((('commonName', 'example.com'),),),
            'subjectAltName': (('DNS', '*.example.net'),
                               ('DNS', 'example.net'))}
check(_verifycert(san_cert, 'example.net'),
      None)
check(_verifycert(san_cert, 'foo.example.net'),
      None)
# no fallback to subject commonName when subjectAltName has DNS
check(_verifycert(san_cert, 'example.com'),
      'certificate is for *.example.net, example.net')
# fallback to subject commonName when no DNS in subjectAltName
san_cert = {'subject': ((('commonName', 'example.com'),),),
            'subjectAltName': (('IP Address', '8.8.8.8'),)}
check(_verifycert(san_cert, 'example.com'), None)

# Avoid some pitfalls
check(_verifycert(cert('*.foo'), 'foo'),
      'certificate is for *.foo')
check(_verifycert(cert('*o'), 'foo'),
      'certificate is for *o')

check(_verifycert({'subject': ()},
                  'example.com'),
      'no commonName or subjectAltName found in certificate')
check(_verifycert(None, 'example.com'),
      'no certificate received')

# Unicode (IDN) certname isn't supported
check(_verifycert(cert(u'\u4f8b.jp'), 'example.jp'),
      'IDN in certificate not supported')

import doctest

def test_url():
    """
    >>> from mercurial.util import url

    This tests for edge cases in url.URL's parsing algorithm. Most of
    these aren't useful for documentation purposes, so they aren't
    part of the class's doc tests.

    Query strings and fragments:

    >>> url('http://host/a?b#c')
    <url scheme: 'http', host: 'host', path: 'a', query: 'b', fragment: 'c'>
    >>> url('http://host/a?')
    <url scheme: 'http', host: 'host', path: 'a'>
    >>> url('http://host/a#b#c')
    <url scheme: 'http', host: 'host', path: 'a', fragment: 'b#c'>
    >>> url('http://host/a#b?c')
    <url scheme: 'http', host: 'host', path: 'a', fragment: 'b?c'>
    >>> url('http://host/?a#b')
    <url scheme: 'http', host: 'host', path: '', query: 'a', fragment: 'b'>
    >>> url('http://host/?a#b', parsequery=False)
    <url scheme: 'http', host: 'host', path: '?a', fragment: 'b'>
    >>> url('http://host/?a#b', parsefragment=False)
    <url scheme: 'http', host: 'host', path: '', query: 'a#b'>
    >>> url('http://host/?a#b', parsequery=False, parsefragment=False)
    <url scheme: 'http', host: 'host', path: '?a#b'>

    IPv6 addresses:

    >>> url('ldap://[2001:db8::7]/c=GB?objectClass?one')
    <url scheme: 'ldap', host: '[2001:db8::7]', path: 'c=GB',
         query: 'objectClass?one'>
    >>> url('ldap://joe:xxx@[2001:db8::7]:80/c=GB?objectClass?one')
    <url scheme: 'ldap', user: 'joe', passwd: 'xxx', host: '[2001:db8::7]',
         port: '80', path: 'c=GB', query: 'objectClass?one'>

    Missing scheme, host, etc.:

    >>> url('://192.0.2.16:80/')
    <url path: '://192.0.2.16:80/'>
    >>> url('http://mercurial.selenic.com')
    <url scheme: 'http', host: 'mercurial.selenic.com'>
    >>> url('/foo')
    <url path: '/foo'>
    >>> url('bundle:/foo')
    <url scheme: 'bundle', path: '/foo'>
    >>> url('a?b#c')
    <url path: 'a?b', fragment: 'c'>
    >>> url('http://x.com?arg=/foo')
    <url scheme: 'http', host: 'x.com', query: 'arg=/foo'>
    >>> url('http://joe:xxx@/foo')
    <url scheme: 'http', user: 'joe', passwd: 'xxx', path: 'foo'>

    Just a scheme and a path:

    >>> url('mailto:[e-mail]')
    <url scheme: 'mailto', path: '[e-mail]'>
    >>> url('a:b:c:d')
    <url path: 'a:b:c:d'>
    >>> url('aa:bb:cc:dd')
    <url scheme: 'aa', path: 'bb:cc:dd'>

    SSH examples:

    >>> url('ssh://joe@host//home/joe')
    <url scheme: 'ssh', user: 'joe', host: 'host', path: '/home/joe'>
    >>> url('ssh://joe:xxx@host/src')
    <url scheme: 'ssh', user: 'joe', passwd: 'xxx', host: 'host', path: 'src'>
    >>> url('ssh://joe:xxx@host')
    <url scheme: 'ssh', user: 'joe', passwd: 'xxx', host: 'host'>
    >>> url('ssh://joe@host')
    <url scheme: 'ssh', user: 'joe', host: 'host'>
    >>> url('ssh://host')
    <url scheme: 'ssh', host: 'host'>
    >>> url('ssh://')
    <url scheme: 'ssh'>
    >>> url('ssh:')
    <url scheme: 'ssh'>

    Non-numeric port:

    >>> url('http://example.com:dd')
    <url scheme: 'http', host: 'example.com', port: 'dd'>
    >>> url('ssh://joe:xxx@host:ssh/foo')
    <url scheme: 'ssh', user: 'joe', passwd: 'xxx', host: 'host', port: 'ssh',
         path: 'foo'>

    Bad authentication credentials:

    >>> url('http://joe@joeville:123@4:@host/a?b#c')
    <url scheme: 'http', user: 'joe@joeville', passwd: '123@4:',
         host: 'host', path: 'a', query: 'b', fragment: 'c'>
    >>> url('http://!*#?/@!*#?/:@host/a?b#c')
    <url scheme: 'http', host: '!*', fragment: '?/@!*#?/:@host/a?b#c'>
    >>> url('http://!*#?@!*#?:@host/a?b#c')
    <url scheme: 'http', host: '!*', fragment: '?@!*#?:@host/a?b#c'>
    >>> url('http://!*@:!*@@host/a?b#c')
    <url scheme: 'http', user: '!*@', passwd: '!*@', host: 'host',
         path: 'a', query: 'b', fragment: 'c'>

    File paths:

    >>> url('a/b/c/d.g.f')
    <url path: 'a/b/c/d.g.f'>
    >>> url('/x///z/y/')
    <url path: '/x///z/y/'>
    >>> url('/foo:bar')
    <url path: '/foo:bar'>
    >>> url('\\\\foo:bar')
    <url path: '\\\\foo:bar'>
    >>> url('./foo:bar')
    <url path: './foo:bar'>

    Non-localhost file URL:

    >>> u = url('file://mercurial.selenic.com/foo')
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    Abort: file:// URLs can only refer to localhost

    Empty URL:

    >>> u = url('')
    >>> u
    <url path: ''>
    >>> str(u)
    ''

    Empty path with query string:

    >>> str(url('http://foo/?bar'))
    'http://foo/?bar'

    Invalid path:

    >>> u = url('http://foo/bar')
    >>> u.path = 'bar'
    >>> str(u)
    'http://foo/bar'

    >>> u = url('file:/foo/bar/baz')
    >>> u
    <url scheme: 'file', path: '/foo/bar/baz'>
    >>> str(u)
    'file:///foo/bar/baz'

    >>> u = url('file:///foo/bar/baz')
    >>> u
    <url scheme: 'file', path: '/foo/bar/baz'>
    >>> str(u)
    'file:///foo/bar/baz'

    >>> u = url('file:foo/bar/baz')
    >>> u
    <url scheme: 'file', path: 'foo/bar/baz'>
    >>> str(u)
    'file:foo/bar/baz'
    """

doctest.testmod(optionflags=doctest.NORMALIZE_WHITESPACE)