rhodecode/lib/utils2.py

# -*- coding: utf-8 -*-

# Copyright (C) 2011-2020 RhodeCode GmbH
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License, version 3
# (only), as published by the Free Software Foundation.
#
# 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 Affero General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
# This program is dual-licensed. If you wish to learn more about the
# RhodeCode Enterprise Edition, including its added features, Support services,
# and proprietary license terms, please see https://rhodecode.com/licenses/


"""
Some simple helper functions
"""

import collections
import datetime
import dateutil.relativedelta
import logging
import re
import sys
import time
import urllib.request, urllib.parse, urllib.error
import urlobject
import uuid
import getpass
import socket
import errno
import random
from functools import update_wrapper, partial, wraps
from contextlib import closing

import pygments.lexers
import sqlalchemy
import sqlalchemy.engine.url
import sqlalchemy.exc
import sqlalchemy.sql
import webob
import pyramid.threadlocal
from pyramid.settings import asbool

import rhodecode
from rhodecode.translation import _, _pluralize
from rhodecode.lib.str_utils import safe_str, safe_int, safe_bytes
from rhodecode.lib.hash_utils import md5, md5_safe, sha1, sha1_safe
from rhodecode.lib.type_utils import aslist, str2bool
from functools import reduce

#TODO: there's no longer safe_unicode, we mock it now, but should remove it
safe_unicode = safe_str


def __get_lem(extra_mapping=None):
    """
    Get language extension map based on what's inside pygments lexers
    """
    d = collections.defaultdict(lambda: [])

    def __clean(s):
        s = s.lstrip('*')
        s = s.lstrip('.')

        if s.find('[') != -1:
            exts = []
            start, stop = s.find('['), s.find(']')

            for suffix in s[start + 1:stop]:
                exts.append(s[:s.find('[')] + suffix)
            return [e.lower() for e in exts]
        else:
            return [s.lower()]

    for lx, t in sorted(pygments.lexers.LEXERS.items()):
        m = map(__clean, t[-2])
        if m:
            m = reduce(lambda x, y: x + y, m)
            for ext in m:
                desc = lx.replace('Lexer', '')
                d[ext].append(desc)

    data = dict(d)

    extra_mapping = extra_mapping or {}
    if extra_mapping:
        for k, v in extra_mapping.items():
            if k not in data:
                # register new mapping2lexer
                data[k] = [v]

    return data


def convert_line_endings(line, mode):
    """
    Converts a given line  "line end" accordingly to given mode

    Available modes are::
        0 - Unix
        1 - Mac
        2 - DOS

    :param line: given line to convert
    :param mode: mode to convert to
    :rtype: str
    :return: converted line according to mode
    """
    if mode == 0:
        line = line.replace('\r\n', '\n')
        line = line.replace('\r', '\n')
    elif mode == 1:
        line = line.replace('\r\n', '\r')
        line = line.replace('\n', '\r')
    elif mode == 2:
        line = re.sub('\r(?!\n)|(?<!\r)\n', '\r\n', line)
    return line


def detect_mode(line, default):
    """
    Detects line break for given line, if line break couldn't be found
    given default value is returned

    :param line: str line
    :param default: default
    :rtype: int
    :return: value of line end on of 0 - Unix, 1 - Mac, 2 - DOS
    """
    if line.endswith('\r\n'):
        return 2
    elif line.endswith('\n'):
        return 0
    elif line.endswith('\r'):
        return 1
    else:
        return default


def remove_suffix(s, suffix):
    if s.endswith(suffix):
        s = s[:-1 * len(suffix)]
    return s


def remove_prefix(s, prefix):
    if s.startswith(prefix):
        s = s[len(prefix):]
    return s


def find_calling_context(ignore_modules=None):
    """
    Look through the calling stack and return the frame which called
    this function and is part of core module ( ie. rhodecode.* )

    :param ignore_modules: list of modules to ignore eg. ['rhodecode.lib']

    usage::
        from rhodecode.lib.utils2 import find_calling_context

        calling_context = find_calling_context(ignore_modules=[
            'rhodecode.lib.caching_query',
            'rhodecode.model.settings',
        ])

        if calling_context:
            cc_str = 'call context %s:%s' % (
                calling_context.f_code.co_filename,
                calling_context.f_lineno,
            )
            print(cc_str)
    """

    ignore_modules = ignore_modules or []

    f = sys._getframe(2)
    while f.f_back is not None:
        name = f.f_globals.get('__name__')
        if name and name.startswith(__name__.split('.')[0]):
            if name not in ignore_modules:
                return f
        f = f.f_back
    return None


def ping_connection(connection, branch):
    if branch:
        # "branch" refers to a sub-connection of a connection,
        # we don't want to bother pinging on these.
        return

    # turn off "close with result".  This flag is only used with
    # "connectionless" execution, otherwise will be False in any case
    save_should_close_with_result = connection.should_close_with_result
    connection.should_close_with_result = False

    try:
        # run a SELECT 1.   use a core select() so that
        # the SELECT of a scalar value without a table is
        # appropriately formatted for the backend
        connection.scalar(sqlalchemy.sql.select([1]))
    except sqlalchemy.exc.DBAPIError as err:
        # catch SQLAlchemy's DBAPIError, which is a wrapper
        # for the DBAPI's exception.  It includes a .connection_invalidated
        # attribute which specifies if this connection is a "disconnect"
        # condition, which is based on inspection of the original exception
        # by the dialect in use.
        if err.connection_invalidated:
            # run the same SELECT again - the connection will re-validate
            # itself and establish a new connection.  The disconnect detection
            # here also causes the whole connection pool to be invalidated
            # so that all stale connections are discarded.
            connection.scalar(sqlalchemy.sql.select([1]))
        else:
            raise
    finally:
        # restore "close with result"
        connection.should_close_with_result = save_should_close_with_result


def engine_from_config(configuration, prefix='sqlalchemy.', **kwargs):
    """Custom engine_from_config functions."""
    log = logging.getLogger('sqlalchemy.engine')
    use_ping_connection = asbool(configuration.pop('sqlalchemy.db1.ping_connection', None))
    debug = asbool(configuration.pop('sqlalchemy.db1.debug_query', None))

    engine = sqlalchemy.engine_from_config(configuration, prefix, **kwargs)

    def color_sql(sql):
        color_seq = '\033[1;33m'  # This is yellow: code 33
        normal = '\x1b[0m'
        return ''.join([color_seq, sql, normal])

    if use_ping_connection:
        log.debug('Adding ping_connection on the engine config.')
        sqlalchemy.event.listen(engine, "engine_connect", ping_connection)

    if debug:
        # attach events only for debug configuration
        def before_cursor_execute(conn, cursor, statement,
                                  parameters, context, executemany):
            setattr(conn, 'query_start_time', time.time())
            log.info(color_sql(">>>>> STARTING QUERY >>>>>"))
            calling_context = find_calling_context(ignore_modules=[
                'rhodecode.lib.caching_query',
                'rhodecode.model.settings',
            ])
            if calling_context:
                log.info(color_sql('call context %s:%s' % (
                    calling_context.f_code.co_filename,
                    calling_context.f_lineno,
                )))

        def after_cursor_execute(conn, cursor, statement,
                                 parameters, context, executemany):
            delattr(conn, 'query_start_time')

        sqlalchemy.event.listen(engine, "before_cursor_execute", before_cursor_execute)
        sqlalchemy.event.listen(engine, "after_cursor_execute", after_cursor_execute)

    return engine


def get_encryption_key(config):
    secret = config.get('rhodecode.encrypted_values.secret')
    default = config['beaker.session.secret']
    return secret or default


def age(prevdate, now=None, show_short_version=False, show_suffix=True, short_format=False):
    """
    Turns a datetime into an age string.
    If show_short_version is True, this generates a shorter string with
    an approximate age; ex. '1 day ago', rather than '1 day and 23 hours ago'.

    * IMPORTANT*
    Code of this function is written in special way so it's easier to
    backport it to javascript. If you mean to update it, please also update
    `jquery.timeago-extension.js` file

    :param prevdate: datetime object
    :param now: get current time, if not define we use
        `datetime.datetime.now()`
    :param show_short_version: if it should approximate the date and
        return a shorter string
    :param show_suffix:
    :param short_format: show short format, eg 2D instead of 2 days
    :rtype: unicode
    :returns: unicode words describing age
    """

    def _get_relative_delta(now, prevdate):
        base = dateutil.relativedelta.relativedelta(now, prevdate)
        return {
            'year': base.years,
            'month': base.months,
            'day': base.days,
            'hour': base.hours,
            'minute': base.minutes,
            'second': base.seconds,
        }

    def _is_leap_year(year):
        return year % 4 == 0 and (year % 100 != 0 or year % 400 == 0)

    def get_month(prevdate):
        return prevdate.month

    def get_year(prevdate):
        return prevdate.year

    now = now or datetime.datetime.now()
    order = ['year', 'month', 'day', 'hour', 'minute', 'second']
    deltas = {}
    future = False

    if prevdate > now:
        now_old = now
        now = prevdate
        prevdate = now_old
        future = True
    if future:
        prevdate = prevdate.replace(microsecond=0)
    # Get date parts deltas
    for part in order:
        rel_delta = _get_relative_delta(now, prevdate)
        deltas[part] = rel_delta[part]

    # Fix negative offsets (there is 1 second between 10:59:59 and 11:00:00,
    # not 1 hour, -59 minutes and -59 seconds)
    offsets = [[5, 60], [4, 60], [3, 24]]
    for element in offsets:  # seconds, minutes, hours
        num = element[0]
        length = element[1]

        part = order[num]
        carry_part = order[num - 1]

        if deltas[part] < 0:
            deltas[part] += length
            deltas[carry_part] -= 1

    # Same thing for days except that the increment depends on the (variable)
    # number of days in the month
    month_lengths = [31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31]
    if deltas['day'] < 0:
        if get_month(prevdate) == 2 and _is_leap_year(get_year(prevdate)):
            deltas['day'] += 29
        else:
            deltas['day'] += month_lengths[get_month(prevdate) - 1]

        deltas['month'] -= 1

    if deltas['month'] < 0:
        deltas['month'] += 12
        deltas['year'] -= 1

    # Format the result
    if short_format:
        fmt_funcs = {
            'year': lambda d: u'%dy' % d,
            'month': lambda d: u'%dm' % d,
            'day': lambda d: u'%dd' % d,
            'hour': lambda d: u'%dh' % d,
            'minute': lambda d: u'%dmin' % d,
            'second': lambda d: u'%dsec' % d,
        }
    else:
        fmt_funcs = {
            'year': lambda d: _pluralize(u'${num} year', u'${num} years', d, mapping={'num': d}).interpolate(),
            'month': lambda d: _pluralize(u'${num} month', u'${num} months', d, mapping={'num': d}).interpolate(),
            'day': lambda d: _pluralize(u'${num} day', u'${num} days', d, mapping={'num': d}).interpolate(),
            'hour': lambda d: _pluralize(u'${num} hour', u'${num} hours', d, mapping={'num': d}).interpolate(),
            'minute': lambda d: _pluralize(u'${num} minute', u'${num} minutes', d, mapping={'num': d}).interpolate(),
            'second': lambda d: _pluralize(u'${num} second', u'${num} seconds', d, mapping={'num': d}).interpolate(),
        }

    i = 0
    for part in order:
        value = deltas[part]
        if value != 0:

            if i < 5:
                sub_part = order[i + 1]
                sub_value = deltas[sub_part]
            else:
                sub_value = 0

            if sub_value == 0 or show_short_version:
                _val = fmt_funcs[part](value)
                if future:
                    if show_suffix:
                        return _(u'in ${ago}', mapping={'ago': _val})
                    else:
                        return _(_val)

                else:
                    if show_suffix:
                        return _(u'${ago} ago', mapping={'ago': _val})
                    else:
                        return _(_val)

            val = fmt_funcs[part](value)
            val_detail = fmt_funcs[sub_part](sub_value)
            mapping = {'val': val, 'detail': val_detail}

            if short_format:
                datetime_tmpl = _(u'${val}, ${detail}', mapping=mapping)
                if show_suffix:
                    datetime_tmpl = _(u'${val}, ${detail} ago', mapping=mapping)
                    if future:
                        datetime_tmpl = _(u'in ${val}, ${detail}', mapping=mapping)
            else:
                datetime_tmpl = _(u'${val} and ${detail}', mapping=mapping)
                if show_suffix:
                    datetime_tmpl = _(u'${val} and ${detail} ago', mapping=mapping)
                    if future:
                        datetime_tmpl = _(u'in ${val} and ${detail}', mapping=mapping)

            return datetime_tmpl
        i += 1
    return _(u'just now')


def age_from_seconds(seconds):
    seconds = safe_int(seconds) or 0
    prevdate = time_to_datetime(time.time() + seconds)
    return age(prevdate, show_suffix=False, show_short_version=True)


def cleaned_uri(uri):
    """
    Quotes '[' and ']' from uri if there is only one of them.
        according to RFC3986 we cannot use such chars in uri
    :param uri:
    :return: uri without this chars
    """
    return urllib.parse.quote(uri, safe='@$:/')


def credentials_filter(uri):
    """
    Returns a url with removed credentials

    :param uri:
    """
    import urlobject
    if isinstance(uri, rhodecode.lib.encrypt.InvalidDecryptedValue):
        return 'InvalidDecryptionKey'

    url_obj = urlobject.URLObject(cleaned_uri(uri))
    url_obj = url_obj.without_password().without_username()

    return url_obj


def get_host_info(request):
    """
    Generate host info, to obtain full url e.g https://server.com
    use this
    `{scheme}://{netloc}`
    """
    if not request:
        return {}

    qualified_home_url = request.route_url('home')
    parsed_url = urlobject.URLObject(qualified_home_url)
    decoded_path = safe_unicode(urllib.parse.unquote(parsed_url.path.rstrip('/')))

    return {
        'scheme': parsed_url.scheme,
        'netloc': parsed_url.netloc+decoded_path,
        'hostname': parsed_url.hostname,
    }


def get_clone_url(request, uri_tmpl, repo_name, repo_id, repo_type, **override):
    qualified_home_url = request.route_url('home')
    parsed_url = urlobject.URLObject(qualified_home_url)
    decoded_path = safe_unicode(urllib.parse.unquote(parsed_url.path.rstrip('/')))

    args = {
        'scheme': parsed_url.scheme,
        'user': '',
        'sys_user': getpass.getuser(),
        # path if we use proxy-prefix
        'netloc': parsed_url.netloc+decoded_path,
        'hostname': parsed_url.hostname,
        'prefix': decoded_path,
        'repo': repo_name,
        'repoid': str(repo_id),
        'repo_type': repo_type
    }
    args.update(override)
    args['user'] = urllib.parse.quote(safe_str(args['user']))

    for k, v in args.items():
        uri_tmpl = uri_tmpl.replace('{%s}' % k, v)

    # special case for SVN clone url
    if repo_type == 'svn':
        uri_tmpl = uri_tmpl.replace('ssh://', 'svn+ssh://')

    # remove leading @ sign if it's present. Case of empty user
    url_obj = urlobject.URLObject(uri_tmpl)
    url = url_obj.with_netloc(url_obj.netloc.lstrip('@'))

    return safe_unicode(url)


def get_commit_safe(repo, commit_id=None, commit_idx=None, pre_load=None,
                    maybe_unreachable=False, reference_obj=None):
    """
    Safe version of get_commit if this commit doesn't exists for a
    repository it returns a Dummy one instead

    :param repo: repository instance
    :param commit_id: commit id as str
    :param commit_idx: numeric commit index
    :param pre_load: optional list of commit attributes to load
    :param maybe_unreachable: translate unreachable commits on git repos
    :param reference_obj: explicitly search via a reference obj in git. E.g "branch:123" would mean branch "123"
    """
    # TODO(skreft): remove these circular imports
    from rhodecode.lib.vcs.backends.base import BaseRepository, EmptyCommit
    from rhodecode.lib.vcs.exceptions import RepositoryError
    if not isinstance(repo, BaseRepository):
        raise Exception('You must pass an Repository '
                        'object as first argument got %s', type(repo))

    try:
        commit = repo.get_commit(
            commit_id=commit_id, commit_idx=commit_idx, pre_load=pre_load,
            maybe_unreachable=maybe_unreachable, reference_obj=reference_obj)
    except (RepositoryError, LookupError):
        commit = EmptyCommit()
    return commit


def datetime_to_time(dt):
    if dt:
        return time.mktime(dt.timetuple())


def time_to_datetime(tm):
    if tm:
        if isinstance(tm, str):
            try:
                tm = float(tm)
            except ValueError:
                return
        return datetime.datetime.fromtimestamp(tm)


def time_to_utcdatetime(tm):
    if tm:
        if isinstance(tm, str):
            try:
                tm = float(tm)
            except ValueError:
                return
        return datetime.datetime.utcfromtimestamp(tm)


MENTIONS_REGEX = re.compile(
    # ^@ or @ without any special chars in front
    r'(?:^@|[^a-zA-Z0-9\-\_\.]@)'
    # main body starts with letter, then can be . - _
    r'([a-zA-Z0-9]{1}[a-zA-Z0-9\-\_\.]+)',
    re.VERBOSE | re.MULTILINE)


def extract_mentioned_users(s):
    """
    Returns unique usernames from given string s that have @mention

    :param s: string to get mentions
    """
    usrs = set()
    for username in MENTIONS_REGEX.findall(s):
        usrs.add(username)

    return sorted(list(usrs), key=lambda k: k.lower())


class AttributeDictBase(dict):
    def __getstate__(self):
        odict = self.__dict__  # get attribute dictionary
        return odict

    def __setstate__(self, dict):
        self.__dict__ = dict

    __setattr__ = dict.__setitem__
    __delattr__ = dict.__delitem__


class StrictAttributeDict(AttributeDictBase):
    """
    Strict Version of Attribute dict which raises an Attribute error when
    requested attribute is not set
    """
    def __getattr__(self, attr):
        try:
            return self[attr]
        except KeyError:
            raise AttributeError('%s object has no attribute %s' % (
                self.__class__, attr))


class AttributeDict(AttributeDictBase):
    def __getattr__(self, attr):
        return self.get(attr, None)


def fix_PATH(os_=None):
    """
    Get current active python path, and append it to PATH variable to fix
    issues of subprocess calls and different python versions
    """
    if os_ is None:
        import os
    else:
        os = os_

    cur_path = os.path.split(sys.executable)[0]
    if not os.environ['PATH'].startswith(cur_path):
        os.environ['PATH'] = '%s:%s' % (cur_path, os.environ['PATH'])


def obfuscate_url_pw(engine):
    _url = engine or ''
    try:
        _url = sqlalchemy.engine.url.make_url(engine)
        if _url.password:
            _url.password = 'XXXXX'
    except Exception:
        pass
    return str(_url)


def get_server_url(environ):
    req = webob.Request(environ)
    return req.host_url + req.script_name


def unique_id(hexlen=32):
    alphabet = "23456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghjklmnpqrstuvwxyz"
    return suuid(truncate_to=hexlen, alphabet=alphabet)


def suuid(url=None, truncate_to=22, alphabet=None):
    """
    Generate and return a short URL safe UUID.

    If the url parameter is provided, set the namespace to the provided
    URL and generate a UUID.

    :param url to get the uuid for
    :truncate_to: truncate the basic 22 UUID to shorter version

    The IDs won't be universally unique any longer, but the probability of
    a collision will still be very low.
    """
    # Define our alphabet.
    _ALPHABET = alphabet or "23456789ABCDEFGHJKLMNPQRSTUVWXYZ"

    # If no URL is given, generate a random UUID.
    if url is None:
        unique_id = uuid.uuid4().int
    else:
        unique_id = uuid.uuid3(uuid.NAMESPACE_URL, url).int

    alphabet_length = len(_ALPHABET)
    output = []
    while unique_id > 0:
        digit = unique_id % alphabet_length
        output.append(_ALPHABET[digit])
        unique_id = int(unique_id / alphabet_length)
    return "".join(output)[:truncate_to]


def get_current_rhodecode_user(request=None):
    """
    Gets rhodecode user from request
    """
    pyramid_request = request or pyramid.threadlocal.get_current_request()

    # web case
    if pyramid_request and hasattr(pyramid_request, 'user'):
        return pyramid_request.user

    # api case
    if pyramid_request and hasattr(pyramid_request, 'rpc_user'):
        return pyramid_request.rpc_user

    return None


def action_logger_generic(action, namespace=''):
    """
    A generic logger for actions useful to the system overview, tries to find
    an acting user for the context of the call otherwise reports unknown user

    :param action: logging message eg 'comment 5 deleted'
    :param type: string

    :param namespace: namespace of the logging message eg. 'repo.comments'
    :param type: string

    """

    logger_name = 'rhodecode.actions'

    if namespace:
        logger_name += '.' + namespace

    log = logging.getLogger(logger_name)

    # get a user if we can
    user = get_current_rhodecode_user()

    logfunc = log.info

    if not user:
        user = '<unknown user>'
        logfunc = log.warning

    logfunc('Logging action by {}: {}'.format(user, action))


def escape_split(text, sep=',', maxsplit=-1):
    r"""
    Allows for escaping of the separator: e.g. arg='foo\, bar'

    It should be noted that the way bash et. al. do command line parsing, those
    single quotes are required.
    """
    escaped_sep = r'\%s' % sep

    if escaped_sep not in text:
        return text.split(sep, maxsplit)

    before, _mid, after = text.partition(escaped_sep)
    startlist = before.split(sep, maxsplit)  # a regular split is fine here
    unfinished = startlist[-1]
    startlist = startlist[:-1]

    # recurse because there may be more escaped separators
    endlist = escape_split(after, sep, maxsplit)

    # finish building the escaped value. we use endlist[0] becaue the first
    # part of the string sent in recursion is the rest of the escaped value.
    unfinished += sep + endlist[0]

    return startlist + [unfinished] + endlist[1:]  # put together all the parts


class OptionalAttr(object):
    """
    Special Optional Option that defines other attribute. Example::

        def test(apiuser, userid=Optional(OAttr('apiuser')):
            user = Optional.extract(userid)
            # calls

    """

    def __init__(self, attr_name):
        self.attr_name = attr_name

    def __repr__(self):
        return '<OptionalAttr:%s>' % self.attr_name

    def __call__(self):
        return self


# alias
OAttr = OptionalAttr


class Optional(object):
    """
    Defines an optional parameter::

        param = param.getval() if isinstance(param, Optional) else param
        param = param() if isinstance(param, Optional) else param

    is equivalent of::

        param = Optional.extract(param)

    """

    def __init__(self, type_):
        self.type_ = type_

    def __repr__(self):
        return '<Optional:%s>' % self.type_.__repr__()

    def __call__(self):
        return self.getval()

    def getval(self):
        """
        returns value from this Optional instance
        """
        if isinstance(self.type_, OAttr):
            # use params name
            return self.type_.attr_name
        return self.type_

    @classmethod
    def extract(cls, val):
        """
        Extracts value from Optional() instance

        :param val:
        :return: original value if it's not Optional instance else
            value of instance
        """
        if isinstance(val, cls):
            return val.getval()
        return val


def glob2re(pat):
    """
    Translate a shell PATTERN to a regular expression.

    There is no way to quote meta-characters.
    """

    i, n = 0, len(pat)
    res = ''
    while i < n:
        c = pat[i]
        i = i+1
        if c == '*':
            #res = res + '.*'
            res = res + '[^/]*'
        elif c == '?':
            #res = res + '.'
            res = res + '[^/]'
        elif c == '[':
            j = i
            if j < n and pat[j] == '!':
                j = j+1
            if j < n and pat[j] == ']':
                j = j+1
            while j < n and pat[j] != ']':
                j = j+1
            if j >= n:
                res = res + '\\['
            else:
                stuff = pat[i:j].replace('\\','\\\\')
                i = j+1
                if stuff[0] == '!':
                    stuff = '^' + stuff[1:]
                elif stuff[0] == '^':
                    stuff = '\\' + stuff
                res = '%s[%s]' % (res, stuff)
        else:
            res = res + re.escape(c)
    return res + '\Z(?ms)'


def parse_byte_string(size_str):
    match = re.match(r'(\d+)(MB|KB)', size_str, re.IGNORECASE)
    if not match:
        raise ValueError('Given size:%s is invalid, please make sure '
                         'to use format of <num>(MB|KB)' % size_str)

    _parts = match.groups()
    num, type_ = _parts
    return int(num) * {'mb': 1024*1024, 'kb': 1024}[type_.lower()]


class CachedProperty(object):
    """
    Lazy Attributes. With option to invalidate the cache by running a method

    >>> class Foo(object):
    ...
    ...    @CachedProperty
    ...    def heavy_func(self):
    ...        return 'super-calculation'
    ...
    ... foo = Foo()
    ... foo.heavy_func() # first computation
    ... foo.heavy_func() # fetch from cache
    ... foo._invalidate_prop_cache('heavy_func')

    # at this point calling foo.heavy_func() will be re-computed
    """

    def __init__(self, func, func_name=None):

        if func_name is None:
            func_name = func.__name__
        self.data = (func, func_name)
        update_wrapper(self, func)

    def __get__(self, inst, class_):
        if inst is None:
            return self

        func, func_name = self.data
        value = func(inst)
        inst.__dict__[func_name] = value
        if '_invalidate_prop_cache' not in inst.__dict__:
            inst.__dict__['_invalidate_prop_cache'] = partial(
                self._invalidate_prop_cache, inst)
        return value

    def _invalidate_prop_cache(self, inst, name):
        inst.__dict__.pop(name, None)


def retry(func=None, exception=Exception, n_tries=5, delay=5, backoff=1, logger=True):
    """
    Retry decorator with exponential backoff.

    Parameters
    ----------
    func : typing.Callable, optional
        Callable on which the decorator is applied, by default None
    exception : Exception or tuple of Exceptions, optional
        Exception(s) that invoke retry, by default Exception
    n_tries : int, optional
        Number of tries before giving up, by default 5
    delay : int, optional
        Initial delay between retries in seconds, by default 5
    backoff : int, optional
        Backoff multiplier e.g. value of 2 will double the delay, by default 1
    logger : bool, optional
        Option to log or print, by default False

    Returns
    -------
    typing.Callable
        Decorated callable that calls itself when exception(s) occur.

    Examples
    --------
    >>> import random
    >>> @retry(exception=Exception, n_tries=3)
    ... def test_random(text):
    ...    x = random.random()
    ...    if x < 0.5:
    ...        raise Exception("Fail")
    ...    else:
    ...        print("Success: ", text)
    >>> test_random("It works!")
    """

    if func is None:
        return partial(
            retry,
            exception=exception,
            n_tries=n_tries,
            delay=delay,
            backoff=backoff,
            logger=logger,
        )

    @wraps(func)
    def wrapper(*args, **kwargs):
        _n_tries, n_delay = n_tries, delay
        log = logging.getLogger('rhodecode.retry')

        while _n_tries > 1:
            try:
                return func(*args, **kwargs)
            except exception as e:
                e_details = repr(e)
                msg = "Exception on calling func {func}: {e}, " \
                      "Retrying in {n_delay} seconds..."\
                    .format(func=func, e=e_details, n_delay=n_delay)
                if logger:
                    log.warning(msg)
                else:
                    print(msg)
                time.sleep(n_delay)
                _n_tries -= 1
                n_delay *= backoff

        return func(*args, **kwargs)

    return wrapper


def user_agent_normalizer(user_agent_raw, safe=True):
    log = logging.getLogger('rhodecode.user_agent_normalizer')
    ua = (user_agent_raw or '').strip().lower()
    ua = ua.replace('"', '')

    try:
        if 'mercurial/proto-1.0' in ua:
            ua = ua.replace('mercurial/proto-1.0', '')
            ua = ua.replace('(', '').replace(')', '').strip()
            ua = ua.replace('mercurial ', 'mercurial/')
        elif ua.startswith('git'):
            parts = ua.split(' ')
            if parts:
                ua = parts[0]
            ua = re.sub('\.windows\.\d', '', ua).strip()

        return ua
    except Exception:
        log.exception('Failed to parse scm user-agent')
        if not safe:
            raise

    return ua


def get_available_port(min_port=40000, max_port=55555, use_range=False):
    hostname = ''
    for _ in range(min_port, max_port):
        pick_port = 0
        if use_range:
            pick_port = random.randint(min_port, max_port)

        with closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
            try:
                s.bind((hostname, pick_port))
                s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
                return s.getsockname()[1]
            except OSError:
                continue
            except socket.error as e:
                if e.args[0] in [errno.EADDRINUSE, errno.ECONNREFUSED]:
                    continue
                raise