/brz/remove-bazaar

# Copyright (C) 2006-2010 Canonical Ltd

#

# 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

"""A collection of function for handling URL operations."""

from __future__ import absolute_import

import os

import re

import sys

try:

    import urlparse

except ImportError:

    from urllib import parse as urlparse

from .lazy_import import lazy_import

lazy_import(globals(), """

from posixpath import split as _posix_split

from breezy import (

    errors,

    osutils,

    )

""")

from .sixish import (

    text_type,

    )

def basename(url, exclude_trailing_slash=True):

    """Return the last component of a URL.

    :param url: The URL in question

    :param exclude_trailing_slash: If the url looks like "path/to/foo/"

        ignore the final slash and return 'foo' rather than ''

    :return: Just the final component of the URL. This can return ''

        if you don't exclude_trailing_slash, or if you are at the

        root of the URL.

    """

    return split(url, exclude_trailing_slash=exclude_trailing_slash)[1]

def dirname(url, exclude_trailing_slash=True):

    """Return the parent directory of the given path.

    :param url: Relative or absolute URL

    :param exclude_trailing_slash: Remove a final slash

        (treat http://host/foo/ as http://host/foo, but

        http://host/ stays http://host/)

    :return: Everything in the URL except the last path chunk

    """

    # TODO: jam 20060502 This was named dirname to be consistent

    #       with the os functions, but maybe "parent" would be better

    return split(url, exclude_trailing_slash=exclude_trailing_slash)[0]

# Private copies of quote and unquote, copied from Python's

# urllib module because urllib unconditionally imports socket, which imports

# ssl.

always_safe = (b'ABCDEFGHIJKLMNOPQRSTUVWXYZ'

               b'abcdefghijklmnopqrstuvwxyz'

               b'0123456789' b'_.-')

_safe_map = {}

for i, c in zip(range(256), bytes(bytearray(range(256)))):

    _safe_map[c] = c if (i < 128 and c in always_safe) else '%{0:02X}'.format(i).encode('ascii')

_safe_quoters = {}

def quote(s, safe=b'/'):

    """quote('abc def') -> 'abc%20def'

    Each part of a URL, e.g. the path info, the query, etc., has a

    different set of reserved characters that must be quoted.

    RFC 2396 Uniform Resource Identifiers (URI): Generic Syntax lists

    the following reserved characters.

    reserved    = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |

                  "$" | ","

    Each of these characters is reserved in some component of a URL,

    but not necessarily in all of them.

    By default, the quote function is intended for quoting the path

    section of a URL.  Thus, it will not encode '/'.  This character

    is reserved, but in typical usage the quote function is being

    called on a path where the existing slash characters are used as

    reserved characters.

    """

    # fastpath

    if not s:

        if s is None:

            raise TypeError('None object cannot be quoted')

        return s

    cachekey = (safe, always_safe)

    try:

        (quoter, safe) = _safe_quoters[cachekey]

    except KeyError:

        safe_map = _safe_map.copy()

        safe_map.update([(c, c) for c in safe])

        quoter = safe_map.__getitem__

        safe = always_safe + safe

        _safe_quoters[cachekey] = (quoter, safe)

    if not s.rstrip(safe):

        return s

    return b''.join(map(quoter, s))

_hexdig = '0123456789ABCDEFabcdef'

_hextochr = dict((a + b, chr(int(a + b, 16)))

                 for a in _hexdig for b in _hexdig)

def unquote(s):

    """unquote('abc%20def') -> 'abc def'."""

    res = s.split(b'%')

    # fastpath

    if len(res) == 1:

        return s

    s = res[0]

    for item in res[1:]:

        try:

            s += _hextochr[item[:2]] + item[2:]

        except KeyError:

            s += b'%' + item

        except UnicodeDecodeError:

            s += unichr(int(item[:2], 16)) + item[2:]

    return s

def escape(relpath):

    """Escape relpath to be a valid url."""

    if isinstance(relpath, text_type):

        relpath = relpath.encode('utf-8')

    return quote(relpath, safe=b'/~')

def file_relpath(base, path):

    """Compute just the relative sub-portion of a url

    This assumes that both paths are already fully specified file:// URLs.

    """

    if len(base) < MIN_ABS_FILEURL_LENGTH:

        raise ValueError('Length of base (%r) must equal or'

            ' exceed the platform minimum url length (which is %d)' %

            (base, MIN_ABS_FILEURL_LENGTH))

    base = osutils.normpath(local_path_from_url(base))

    path = osutils.normpath(local_path_from_url(path))

    return escape(osutils.relpath(base, path))

def _find_scheme_and_separator(url):

    """Find the scheme separator (://) and the first path separator

    This is just a helper functions for other path utilities.

    It could probably be replaced by urlparse

    """

    m = _url_scheme_re.match(url)

    if not m:

        return None, None

    scheme = m.group('scheme')

    path = m.group('path')

    # Find the path separating slash

    # (first slash after the ://)

    first_path_slash = path.find(b'/')

    if first_path_slash == -1:

        return len(scheme), None

    return len(scheme), first_path_slash+m.start('path')

def is_url(url):

    """Tests whether a URL is in actual fact a URL."""

    return _url_scheme_re.match(url) is not None

def join(base, *args):

    """Create a URL by joining sections.

    This will normalize '..', assuming that paths are absolute

    (it assumes no symlinks in either path)

    If any of *args is an absolute URL, it will be treated correctly.

    Example:

        join('http://foo', 'http://bar') => 'http://bar'

        join('http://foo', 'bar') => 'http://foo/bar'

        join('http://foo', 'bar', '../baz') => 'http://foo/baz'

    """

    if not args:

        return base

    scheme_end, path_start = _find_scheme_and_separator(base)

    if scheme_end is None and path_start is None:

        path_start = 0

    elif path_start is None:

        path_start = len(base)

    path = base[path_start:]

    for arg in args:

        arg_scheme_end, arg_path_start = _find_scheme_and_separator(arg)

        if arg_scheme_end is None and arg_path_start is None:

            arg_path_start = 0

        elif arg_path_start is None:

            arg_path_start = len(arg)

        if arg_scheme_end is not None:

            base = arg

            path = arg[arg_path_start:]

            scheme_end = arg_scheme_end

            path_start = arg_path_start

        else:

            path = joinpath(path, arg)

    return base[:path_start] + path

def joinpath(base, *args):

    """Join URL path segments to a URL path segment.

    This is somewhat like osutils.joinpath, but intended for URLs.

    XXX: this duplicates some normalisation logic, and also duplicates a lot of

    path handling logic that already exists in some Transport implementations.

    We really should try to have exactly one place in the code base responsible

    for combining paths of URLs.

    """

    path = base.split(b'/')

    if len(path) > 1 and path[-1] == b'':

        #If the path ends in a trailing /, remove it.

        path.pop()

    for arg in args:

        if arg.startswith(b'/'):

            path = []

        for chunk in arg.split(b'/'):

            if chunk == b'.':

                continue

            elif chunk == b'..':

                if path == [b'']:

                    raise errors.InvalidURLJoin('Cannot go above root',

                            base, args)

                path.pop()

            else:

                path.append(chunk)

    if path == [b'']:

        return b'/'

    else:

        return b'/'.join(path)

# jam 20060502 Sorted to 'l' because the final target is 'local_path_from_url'

def _posix_local_path_from_url(url):

    """Convert a url like file:///path/to/foo into /path/to/foo"""

    url = split_segment_parameters_raw(url)[0]

    file_localhost_prefix = b'file://localhost/'

    if url.startswith(file_localhost_prefix):

        path = url[len(file_localhost_prefix) - 1:]

    elif not url.startswith(b'file:///'):

        raise errors.InvalidURL(

            url, 'local urls must start with file:/// or file://localhost/')

    else:

        path = url[len(b'file://'):]

    # We only strip off 2 slashes

    return unescape(path)

def _posix_local_path_to_url(path):

    """Convert a local path like ./foo into a URL like file:///path/to/foo

    This also handles transforming escaping unicode characters, etc.

    """

    # importing directly from posixpath allows us to test this

    # on non-posix platforms

    return b'file://' + escape(osutils._posix_abspath(path))

def _win32_local_path_from_url(url):

    """Convert a url like file:///C:/path/to/foo into C:/path/to/foo"""

    if not url.startswith('file://'):

        raise errors.InvalidURL(url, 'local urls must start with file:///, '

                                     'UNC path urls must start with file://')

    url = split_segment_parameters_raw(url)[0]

    # We strip off all 3 slashes

    win32_url = url[len('file:'):]

    # check for UNC path: //HOST/path

    if not win32_url.startswith('///'):

        if (win32_url[2] == '/'

            or win32_url[3] in '|:'):

            raise errors.InvalidURL(url, 'Win32 UNC path urls'

                ' have form file://HOST/path')

        return unescape(win32_url)

    # allow empty paths so we can serve all roots

    if win32_url == '///':

        return '/'

    # usual local path with drive letter

    if (len(win32_url) < 6

        or win32_url[3] not in ('abcdefghijklmnopqrstuvwxyz'

                                'ABCDEFGHIJKLMNOPQRSTUVWXYZ')

        or win32_url[4] not in  '|:'

        or win32_url[5] != '/'):

        raise errors.InvalidURL(url, 'Win32 file urls start with'

                ' file:///x:/, where x is a valid drive letter')

    return win32_url[3].upper() + u':' + unescape(win32_url[5:])

def _win32_local_path_to_url(path):

    """Convert a local path like ./foo into a URL like file:///C:/path/to/foo

    This also handles transforming escaping unicode characters, etc.

    """

    # importing directly from ntpath allows us to test this

    # on non-win32 platform

    # FIXME: It turns out that on nt, ntpath.abspath uses nt._getfullpathname

    #       which actually strips trailing space characters.

    #       The worst part is that on linux ntpath.abspath has different

    #       semantics, since 'nt' is not an available module.

    if path == '/':

        return 'file:///'

    win32_path = osutils._win32_abspath(path)

    # check for UNC path \\HOST\path

    if win32_path.startswith('//'):

        return 'file:' + escape(win32_path)

    return ('file:///' + str(win32_path[0].upper()) + ':' +

        escape(win32_path[2:]))

local_path_to_url = _posix_local_path_to_url

local_path_from_url = _posix_local_path_from_url

MIN_ABS_FILEURL_LENGTH = len('file:///')

WIN32_MIN_ABS_FILEURL_LENGTH = len('file:///C:/')

if sys.platform == 'win32':

    local_path_to_url = _win32_local_path_to_url

    local_path_from_url = _win32_local_path_from_url

    MIN_ABS_FILEURL_LENGTH = WIN32_MIN_ABS_FILEURL_LENGTH

_url_scheme_re = re.compile(b'^(?P<scheme>[^:/]{2,}):(//)?(?P<path>.*)$')

_url_hex_escapes_re = re.compile(b'(%[0-9a-fA-F]{2})')

def _unescape_safe_chars(matchobj):

    """re.sub callback to convert hex-escapes to plain characters (if safe).

    e.g. '%7E' will be converted to '~'.

    """

    hex_digits = matchobj.group(0)[1:]

    char = chr(int(hex_digits, 16))

    if char in _url_dont_escape_characters:

        return char

    else:

        return matchobj.group(0).upper()

def normalize_url(url):

    """Make sure that a path string is in fully normalized URL form.

    This handles URLs which have unicode characters, spaces,

    special characters, etc.

    It has two basic modes of operation, depending on whether the

    supplied string starts with a url specifier (scheme://) or not.

    If it does not have a specifier it is considered a local path,

    and will be converted into a file:/// url. Non-ascii characters

    will be encoded using utf-8.

    If it does have a url specifier, it will be treated as a "hybrid"

    URL. Basically, a URL that should have URL special characters already

    escaped (like +?&# etc), but may have unicode characters, etc

    which would not be valid in a real URL.

    :param url: Either a hybrid URL or a local path

    :return: A normalized URL which only includes 7-bit ASCII characters.

    """

    scheme_end, path_start = _find_scheme_and_separator(url)

    if scheme_end is None:

        return local_path_to_url(url)

    prefix = url[:path_start]

    path = url[path_start:]

    if not isinstance(url, unicode):

        for c in url:

            if c not in _url_safe_characters:

                raise errors.InvalidURL(url, 'URLs can only contain specific'

                                            ' safe characters (not %r)' % c)

        path = _url_hex_escapes_re.sub(_unescape_safe_chars, path)

        return str(prefix + ''.join(path))

    # We have a unicode (hybrid) url

    path_chars = list(path)

    for i in range(len(path_chars)):

        if path_chars[i] not in _url_safe_characters:

            chars = path_chars[i].encode('utf-8')

            path_chars[i] = ''.join(

                ['%%%02X' % ord(c) for c in path_chars[i].encode('utf-8')])

    path = ''.join(path_chars)

    path = _url_hex_escapes_re.sub(_unescape_safe_chars, path)

    return str(prefix + path)

def relative_url(base, other):

    """Return a path to other from base.

    If other is unrelated to base, return other. Else return a relative path.

    This assumes no symlinks as part of the url.

    """

    dummy, base_first_slash = _find_scheme_and_separator(base)

    if base_first_slash is None:

        return other

    dummy, other_first_slash = _find_scheme_and_separator(other)

    if other_first_slash is None:

        return other

    # this takes care of differing schemes or hosts

    base_scheme = base[:base_first_slash]

    other_scheme = other[:other_first_slash]

    if base_scheme != other_scheme:

        return other

    elif sys.platform == 'win32' and base_scheme == 'file://':

        base_drive = base[base_first_slash+1:base_first_slash+3]

        other_drive = other[other_first_slash+1:other_first_slash+3]

        if base_drive != other_drive:

            return other

    base_path = base[base_first_slash+1:]

    other_path = other[other_first_slash+1:]

    if base_path.endswith('/'):

        base_path = base_path[:-1]

    base_sections = base_path.split('/')

    other_sections = other_path.split('/')

    if base_sections == ['']:

        base_sections = []

    if other_sections == ['']:

        other_sections = []

    output_sections = []

    for b, o in zip(base_sections, other_sections):

        if b != o:

            break

        output_sections.append(b)

    match_len = len(output_sections)

    output_sections = ['..' for x in base_sections[match_len:]]

    output_sections.extend(other_sections[match_len:])

    return "/".join(output_sections) or "."

def _win32_extract_drive_letter(url_base, path):

    """On win32 the drive letter needs to be added to the url base."""

    # Strip off the drive letter

    # path is currently /C:/foo

    if len(path) < 4 or path[2] not in ':|' or path[3] != '/':

        raise errors.InvalidURL(url_base + path,

            'win32 file:/// paths need a drive letter')

    url_base += path[0:3] # file:// + /C:

    path = path[3:] # /foo

    return url_base, path

def split(url, exclude_trailing_slash=True):

    """Split a URL into its parent directory and a child directory.

    :param url: A relative or absolute URL

    :param exclude_trailing_slash: Strip off a final '/' if it is part

        of the path (but not if it is part of the protocol specification)

    :return: (parent_url, child_dir).  child_dir may be the empty string if we're at

        the root.

    """

    scheme_loc, first_path_slash = _find_scheme_and_separator(url)

    if first_path_slash is None:

        # We have either a relative path, or no separating slash

        if scheme_loc is None:

            # Relative path

            if exclude_trailing_slash and url.endswith('/'):

                url = url[:-1]

            return _posix_split(url)

        else:

            # Scheme with no path

            return url, ''

    # We have a fully defined path

    url_base = url[:first_path_slash] # http://host, file://

    path = url[first_path_slash:] # /file/foo

    if sys.platform == 'win32' and url.startswith('file:///'):

        # Strip off the drive letter

        # url_base is currently file://

        # path is currently /C:/foo

        url_base, path = _win32_extract_drive_letter(url_base, path)

        # now it should be file:///C: and /foo

    if exclude_trailing_slash and len(path) > 1 and path.endswith('/'):

        path = path[:-1]

    head, tail = _posix_split(path)

    return url_base + head, tail

def split_segment_parameters_raw(url):

    """Split the subsegment of the last segment of a URL.

    :param url: A relative or absolute URL

    :return: (url, subsegments)

    """

    # GZ 2011-11-18: Dodgy removing the terminal slash like this, function

    #                operates on urls not url+segments, and Transport classes

    #                should not be blindly adding slashes in the first place. 

    lurl = strip_trailing_slash(url)

    # Segments begin at first comma after last forward slash, if one exists

    segment_start = lurl.find(b",", lurl.rfind(b"/")+1)

    if segment_start == -1:

        return (url, [])

    return (lurl[:segment_start], lurl[segment_start+1:].split(b","))

def split_segment_parameters(url):

    """Split the segment parameters of the last segment of a URL.

    :param url: A relative or absolute URL

    :return: (url, segment_parameters)

    """

    (base_url, subsegments) = split_segment_parameters_raw(url)

    parameters = {}

    for subsegment in subsegments:

        (key, value) = subsegment.split("=", 1)

        parameters[key] = value

    return (base_url, parameters)

def join_segment_parameters_raw(base, *subsegments):

    """Create a new URL by adding subsegments to an existing one. 

    This adds the specified subsegments to the last path in the specified

    base URL. The subsegments should be bytestrings.

    :note: You probably want to use join_segment_parameters instead.

    """

    if not subsegments:

        return base

    for subsegment in subsegments:

        if not isinstance(subsegment, str):

            raise TypeError("Subsegment %r is not a bytestring" % subsegment)

        if "," in subsegment:

            raise errors.InvalidURLJoin(", exists in subsegments",

                                        base, subsegments)

    return ",".join((base,) + subsegments)

def join_segment_parameters(url, parameters):

    """Create a new URL by adding segment parameters to an existing one.

    The parameters of the last segment in the URL will be updated; if a

    parameter with the same key already exists it will be overwritten.

    :param url: A URL, as string

    :param parameters: Dictionary of parameters, keys and values as bytestrings

    """

    (base, existing_parameters) = split_segment_parameters(url)

    new_parameters = {}

    new_parameters.update(existing_parameters)

    for key, value in parameters.items():

        if not isinstance(key, str):

            raise TypeError("parameter key %r is not a bytestring" % key)

        if not isinstance(value, str):

            raise TypeError("parameter value %r for %s is not a bytestring" %

                (key, value))

        if "=" in key:

            raise errors.InvalidURLJoin("= exists in parameter key", url,

                parameters)

        new_parameters[key] = value

    return join_segment_parameters_raw(base, 

        *["%s=%s" % item for item in sorted(new_parameters.items())])

def _win32_strip_local_trailing_slash(url):

    """Strip slashes after the drive letter"""

    if len(url) > WIN32_MIN_ABS_FILEURL_LENGTH:

        return url[:-1]

    else:

        return url

def strip_trailing_slash(url):

    """Strip trailing slash, except for root paths.

    The definition of 'root path' is platform-dependent.

    This assumes that all URLs are valid netloc urls, such that they

    form:

    scheme://host/path

    It searches for ://, and then refuses to remove the next '/'.

    It can also handle relative paths

    Examples:

        path/to/foo       => path/to/foo

        path/to/foo/      => path/to/foo

        http://host/path/ => http://host/path

        http://host/path  => http://host/path

        http://host/      => http://host/

        file:///          => file:///

        file:///foo/      => file:///foo

        # This is unique on win32 platforms, and is the only URL

        # format which does it differently.

        file:///c|/       => file:///c:/

    """

    if not url.endswith(b'/'):

        # Nothing to do

        return url

    if sys.platform == 'win32' and url.startswith(b'file://'):

        return _win32_strip_local_trailing_slash(url)

    scheme_loc, first_path_slash = _find_scheme_and_separator(url)

    if scheme_loc is None:

        # This is a relative path, as it has no scheme

        # so just chop off the last character

        return url[:-1]

    if first_path_slash is None or first_path_slash == len(url)-1:

        # Don't chop off anything if the only slash is the path

        # separating slash

        return url

    return url[:-1]

def unescape(url):

    """Unescape relpath from url format.

    This returns a Unicode path from a URL

    """

    # jam 20060427 URLs are supposed to be ASCII only strings

    #       If they are passed in as unicode, unquote

    #       will return a UNICODE string, which actually contains

    #       utf-8 bytes. So we have to ensure that they are

    #       plain ASCII strings, or the final .decode will

    #       try to encode the UNICODE => ASCII, and then decode

    #       it into utf-8.

    if isinstance(url, text_type):

        try:

            url = url.encode("ascii")

        except UnicodeError as e:

            raise errors.InvalidURL(url, 'URL was not a plain ASCII url: %s' % (e,))

    unquoted = unquote(url)

    try:

        unicode_path = unquoted.decode('utf-8')

    except UnicodeError as e:

        raise errors.InvalidURL(url, 'Unable to encode the URL as utf-8: %s' % (e,))

    return unicode_path

# These are characters that if escaped, should stay that way

_no_decode_chars = ';/?:@&=+$,#'

_no_decode_ords = [ord(c) for c in _no_decode_chars]

_no_decode_hex = (['%02x' % o for o in _no_decode_ords]

                + ['%02X' % o for o in _no_decode_ords])

_hex_display_map = dict(([('%02x' % o, chr(o)) for o in range(256)]

                    + [('%02X' % o, chr(o)) for o in range(256)]))

#These entries get mapped to themselves

_hex_display_map.update((hex,'%'+hex) for hex in _no_decode_hex)