# Copyright (C) 2007-2017 Damon Lynch <damonlynch@gmail.com>

# This file is part of Rapid Photo Downloader.
#
# Rapid Photo Downloader 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 3 of the License, or
# (at your option) any later version.
#
# Rapid Photo Downloader 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 Rapid Photo Downloader.  If not,
# see <http://www.gnu.org/licenses/>.

__author__ = 'Damon Lynch'
__copyright__ = "Copyright 2007-2017, Damon Lynch"

import contextlib
import locale
import logging
import os
import random
import re
import string
import sys
import tempfile
import time
import tarfile
from collections import namedtuple, defaultdict
from datetime import datetime
from gettext import gettext as _
from itertools import groupby, zip_longest
from typing import Optional, List, Union, Any
import struct
import ctypes
import signal
import pkg_resources

import arrow
import psutil

import raphodo.__about__ as __about__


# Linux specific code to ensure child processes exit when parent dies
# See http://stackoverflow.com/questions/19447603/
# how-to-kill-a-python-child-process-created-with-subprocess-check-output-when-t/
libc = ctypes.CDLL("libc.so.6")
def set_pdeathsig(sig = signal.SIGTERM):
    def callable():
        return libc.prctl(1, sig)
    return callable


def available_cpu_count(physical_only=False) -> int:
    """
    Determine the number of CPUs available.

    A CPU is "available" if cpuset has not restricted the number of
    cpus. Portions of this code from
    http://stackoverflow.com/questions/1006289/how-to-find-out-the-number-of-
    cpus-using-python

    :return available CPU count, or 1 if cannot be determined.
     Value guaranteed to be >= 1.
    """

    # cpuset may restrict the number of *available* processors
    available = None
    if sys.platform.startswith('linux'):
        try:
            m = re.search(r'(?m)^Cpus_allowed:\s*(.*)$',
                          open('/proc/self/status').read())
            if m:
                available = bin(int(m.group(1).replace(',', ''), 16)).count('1')
                if available > 0 and not physical_only:
                    return available
        except IOError:
            pass

    if physical_only:
        physical = psutil.cpu_count(logical=False)
        if physical is not None:
            if available is not None:
                return min(available, physical)
            return physical

    c = os.cpu_count()
    if c is not None:
        return max(c, 1)
    c = psutil.cpu_count()
    if c is not None:
        return max(c, 1)
    else:
        return 1

def confirm(prompt: Optional[str]=None, resp: Optional[bool]=False) -> bool:
    r"""
    Prompts for yes or no response from the user.

    :param prompt: prompt displayed to user
    :param resp: the default value assumed by the caller when user
     simply types ENTER.
    :return: True for yes and False for no.
    """

    # >>> confirm(prompt='Create Directory?', resp=True)
    # Create Directory? [y]|n:
    # True
    # >>> confirm(prompt='Create Directory?', resp=False)
    # Create Directory? [n]|y:
    # False
    # >>> confirm(prompt='Create Directory?', resp=False)
    # Create Directory? [n]|y: y
    # True

    if prompt is None:
        prompt = 'Confirm'

    if resp:
        prompt = '%s [%s]|%s: ' % (prompt, 'y', 'n')
    else:
        prompt = '%s [%s]|%s: ' % (prompt, 'n', 'y')

    while True:
        ans = input(prompt)
        if not ans:
            return resp
        if ans not in ['y', 'Y', 'n', 'N']:
            print('please enter y or n.')
            continue
        return ans in ['y', 'Y']


@contextlib.contextmanager
def stdchannel_redirected(stdchannel, dest_filename):
    """
    A context manager to temporarily redirect stdout or stderr

    Usage:
    with stdchannel_redirected(sys.stderr, os.devnull):
       do_work()

    Source:
    http://marc-abramowitz.com/archives/2013/07/19/python-context-manager-for-redirected-stdout-and-stderr/
    """
    oldstdchannel = dest_file = None
    try:
        oldstdchannel = os.dup(stdchannel.fileno())
        dest_file = open(dest_filename, 'w')
        os.dup2(dest_file.fileno(), stdchannel.fileno())
        yield
    finally:
        if oldstdchannel is not None:
            os.dup2(oldstdchannel, stdchannel.fileno())
        if dest_file is not None:
            dest_file.close()

@contextlib.contextmanager
def show_errors():
    yield

# Translators: these values are file size suffixes like B representing bytes, KB representing
# kilobytes, etc.
suffixes = [_('B'), _('KB'), _('MB'), _('GB'), _('TB'), _('PB'), _('EB'), _('ZB'), _('YB')]

def format_size_for_user(size_in_bytes: int, 
                         zero_string: str='', 
                         no_decimals: int=2) -> str:
    r"""
    Humanize display of bytes.

    Uses Microsoft style i.e. 1000 Bytes = 1 KB

    :param size: size in bytes
    :param zero_string: string to use if size == 0

    >>> locale.setlocale(locale.LC_ALL, ('en_US', 'utf-8'))
    'en_US.UTF-8'
    >>> format_size_for_user(0)
    ''
    >>> format_size_for_user(1)
    '1 B'
    >>> format_size_for_user(123)
    '123 B'
    >>> format_size_for_user(1000)
    '1 KB'
    >>> format_size_for_user(1024)
    '1.02 KB'
    >>> format_size_for_user(1024, no_decimals=0)
    '1 KB'
    >>> format_size_for_user(1100, no_decimals=2)
    '1.1 KB'
    >>> format_size_for_user(1000000, no_decimals=2)
    '1 MB'
    >>> format_size_for_user(1000001, no_decimals=2)
    '1 MB'
    >>> format_size_for_user(1020001, no_decimals=2)
    '1.02 MB'
    """

    if size_in_bytes == 0: return zero_string
    i = 0
    while size_in_bytes >= 1000 and i < len(suffixes)-1:
        size_in_bytes /= 1000
        i += 1

    if no_decimals:
        s = '{:.{prec}f}'.format(size_in_bytes, prec=no_decimals).rstrip('0').rstrip('.')
    else:
        s = '{:.0f}'.format(size_in_bytes)
    return s + ' ' + suffixes[i]

def divide_list(source: list, no_pieces: int) -> list:
    r"""
    Returns a list containing no_pieces lists, with the items
    of the original list evenly distributed
    :param source: the list to divide
    :param no_pieces: the nubmer of pieces the lists
    :return: the new list

    >>> divide_list(list(range(12)), 4)
    [[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]]
    >>> divide_list(list(range(11)), 4)
    [[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10]]
    """
    source_size = len(source)
    slice_size = source_size // no_pieces
    remainder = source_size % no_pieces
    result = []

    extra = 0
    for i in range(no_pieces):
        start = i * slice_size + extra
        source_slice = source[start:start + slice_size]
        if remainder:
            source_slice += [source[start + slice_size]]
            remainder -= 1
            extra += 1
        result.append(source_slice)
    return result

def divide_list_on_length(source: List, length: int) -> List:

    r"""
    Break a list into lists no longer than length.

    >>> l=list(range(11))
    >>> divide_list_on_length(l, 3)
    [[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10]]
    >>> l=list(range(12))
    >>> divide_list_on_length(l, 3)
    [[0, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11]]
    """

    return [source[i:i+length] for i in range(0, len(source), length)]

def addPushButtonLabelSpacer(s: str) -> str:
    return ' ' + s


class GenerateRandomFileName:
    def __init__(self):
        # the characters used to generate temporary filenames
        self.filename_characters = list(string.ascii_letters + string.digits)

    def name(self, extension: str=None) -> str:
        """

        :return: filename 5 characters long without any extension
        """
        if extension is not None:
            return '{}.{}'.format(''.join(
                random.sample(self.filename_characters, 5)),
                extension)
        else:
            return ''.join(random.sample(self.filename_characters, 5))


TempDirs = namedtuple('TempDirs', 'photo_temp_dir, video_temp_dir')
CacheDirs = namedtuple('CacheDirs', 'photo_cache_dir, video_cache_dir')


def create_temp_dir(folder: Optional[str]=None,
                    prefix: Optional[str]=None,
                    force_no_prefix: bool=False) -> str:
    """
    Creates a temporary director and logs errors
    :param folder: the folder in which the temporary directory should
     be created. If not specified, uses the tempfile.mkstemp default.
    :param prefix: any name the directory should start with. If None,
     default rpd-tmp will be used as prefix, unless force_no_prefix
     is True
    :param force_no_prefix: if True, a directory prefix will never
     be used
    :return: full path of the temporary directory
    """
    if prefix is None and not force_no_prefix:
        prefix = "rpd-tmp-"
    try:
        temp_dir = tempfile.mkdtemp(prefix=prefix, dir=folder)
    except OSError as inst:
        msg = "Failed to create temporary directory in %s: %s %s" % (
                      folder,
                      inst.errno,
                      inst.strerror)
        logging.critical(msg)
        temp_dir = None
    return temp_dir


def create_temp_dirs(photo_download_folder: str,
                     video_download_folder: str) -> TempDirs:
    """
    Create pair of temporary directories for photo and video download
    :param photo_download_folder: where photos will be downloaded to
    :param video_download_folder: where videos will be downloaded to
    :return: the directories
    """
    photo_temp_dir = video_temp_dir = None
    if photo_download_folder is not None:
        photo_temp_dir = create_temp_dir(photo_download_folder)
        logging.debug("Photo temporary directory: %s", photo_temp_dir)
    if video_download_folder is not None:
        video_temp_dir = create_temp_dir(video_download_folder)
        logging.debug("Video temporary directory: %s", video_temp_dir)
    return TempDirs(photo_temp_dir, video_temp_dir)


def same_device(file1: str, file2: str) -> bool:
    """
    Returns True if the files / directories are on the same device (partition).

    No error checking.

    :param file1: first file / directory to check
    :param file2: second file / directory to check
    :return: True if the same file system, else false
    """

    dev1 = os.stat(file1).st_dev
    dev2 = os.stat(file2).st_dev
    return dev1 == dev2


def find_mount_point(path: str) -> str:
    """
    Find the mount point of a path
    See:
    http://stackoverflow.com/questions/4453602/how-to-find-the-mountpoint-a-file-resides-on
    
    >>> print(find_mount_point('/crazy/path'))
    /
    
    :param path: 
    :return: 
    """
    path = os.path.realpath(path)
    while not os.path.ismount(path):
        path = os.path.dirname(path)
    return path


def make_internationalized_list(items: List[str]) -> str:
    r"""
    Makes a string of items conforming to i18n

    >>> print(make_internationalized_list([]))
    <BLANKLINE>
    >>> print(make_internationalized_list(['one']))
    one
    >>> print(make_internationalized_list(['one', 'two']))
    one and two
    >>> print(make_internationalized_list(['one', 'two', 'three']))
    one, two and three
    >>> print(make_internationalized_list(['one', 'two', 'three', 'four']))
    one, two, three and four

    Loosely follows the guideline here:
    http://cldr.unicode.org/translation/lists

    :param items: the list of items to make a string out of
    :return: internationalized string
    """
    if len(items) == 1:
        return items[0]
    if len(items) == 2:
        # two things in a list e.g. "device1 and device2"
        return _('%(first_item)s and %(last_item)s') % dict(
            first_item=items[0], last_item=items[1])
    if len(items) > 2:
        s = items[0]
        for item in items[1:-1]:
            # the middle of a list of things
            s =  '%(first_items)s, %(last_items)s'% dict(first_items=s,
                                                         last_items=item)
        # the end of a list of things
        s = '%(start_items)s and %(last_item)s' % dict(start_items=s,
                                                       last_item=items[-1])
        return s
    return ''


def thousands(i: int) -> str:
    """
    Add a thousands seperator (or its locale equivalent) to an
    integer. Assumes the module level locale setting has already been
    set.
    :param i: the integer e.g. 1000
    :return: string with seperators e.g. '1,000'
    """
    try:
        return locale.format("%d", i, grouping=True)
    except TypeError:
        return i


# Source of class AdjacentKey, first_and_last and runs:
# http://stupidpythonideas.blogspot.com/2014/01/grouping-into-runs-of-adjacent-values.html
class AdjacentKey:
    r"""
    >>> [list(g) for k, g in groupby([0, 1, 2, 3, 5, 6, 7, 10, 11, 13, 16], AdjacentKey)]
    [[0, 1, 2, 3], [5, 6, 7], [10, 11], [13], [16]]
    """
    __slots__ = ['obj']

    def __init__(self, obj) -> None:
        self.obj = obj

    def __eq__(self, other) -> bool:
        ret = self.obj - 1 <= other.obj <= self.obj + 1
        if ret:
            self.obj = other.obj
        return ret


def first_and_last(iterable):
    start = end = next(iterable)
    for end in iterable: pass
    return start, end


def runs(iterable):
    r"""
    identify adjacent elements in pre-sorted data

    :param iterable: sorted data

    >>> list(runs([0, 1, 2, 3, 5, 6, 7, 10, 11, 13, 16]))
    [(0, 3), (5, 7), (10, 11), (13, 13), (16, 16)]
    >>> list(runs([0]))
    [(0, 0)]
    >>> list(runs([0, 1, 10, 100, 101]))
    [(0, 1), (10, 10), (100, 101)]
    """

    for k, g in groupby(iterable, AdjacentKey):
        yield first_and_last(g)

numbers = namedtuple('numbers', 'number, plural')

long_numbers = {
    1: _('one'),
    2: _('two'),
    3: _('three'),
    4: _('four'),
    5: _('five'),
    6: _('six'),
    7: _('seven'),
    8: _('eight'),
    9: _('nine'),
    10: _('ten'),
    11: _('eleven'),
    12: _('twelve'),
    13: _('thirteen'),
    14: _('fourteen'),
    15: _('fifteen'),
    16: _('sixteen'),
    17: _('seventeen'),
    18: _('eighteen'),
    19: _('ninenteen'),
    20: _('twenty')
}


def number(value: int) -> numbers:
    r"""
    Convert integer to written form, e.g. one, two, etc.

    Will propagate TypeError or KeyError on
    failure.

    >>> number(1)
    numbers(number='one', plural=False)
    >>> number(2)
    numbers(number='two', plural=True)
    >>> number(10)
    numbers(number='ten', plural=True)
    >>> number(20)
    numbers(number='twenty', plural=True)
    >>>

    :param value: int between 1 and 20
    :return: tuple of str and whether it is plural
    """

    plural = value > 1
    text = long_numbers[value]
    return numbers(text, plural)


def datetime_roughly_equal(dt1: Union[datetime, float], dt2: Union[datetime, float],
                           seconds: int=120) -> bool:
    r"""
    Check to see if date times are equal, give or take n seconds
    :param dt1: python datetime, or timestamp, to check
    :param dt2:python datetime, or timestamp to check
    :param seconds: number of seconds leeway
    :return: True if "equal", False otherwise

    >>> dt1 = datetime.now()
    >>> time.sleep(.1)
    >>> dt2 = datetime.now()
    >>> datetime_roughly_equal(dt1, dt2, 1)
    True
    >>> dt1 = 1458561776.0
    >>> dt2 = 1458561776.0
    >>> datetime_roughly_equal(dt1, dt2, 120)
    True
    >>> dt2 += 450
    >>> datetime_roughly_equal(dt1, dt2, 120)
    False
    >>> datetime_roughly_equal(dt1, dt2, 500)
    True
    """

    at1 = arrow.get(dt1)
    at2 = arrow.get(dt2)
    return at1.replace(seconds=-seconds) < at2 < at1.replace(seconds=+seconds)


def process_running(process_name: str, partial_name: bool=True) -> bool:
    """
    Search the list of the system's running processes to see if a process with this
    name is running

    :param process_name: the name of the process to search for
    :param partial_name: if True, the process_name argument can be a
     partial match
    :return: True if found, else False
    """

    for proc in psutil.process_iter():
        try:
            name = proc.name()
        except psutil.NoSuchProcess:
            pass
        else:
            if partial_name:
                if name.find(process_name) >= 0:
                    return True
            else:
                if name == process_name:
                    return True
    return False

def make_html_path_non_breaking(path: str) -> str:
    """
    When /some/path is displayed in rich text, it will be word-wrapped on the
    slashes. Inhibit that using a special unicode character.

    :param path: the path
    :return: the path containing the special characters
    """

    return path.replace(os.sep, '{}&#8288;'.format(os.sep))


def prefs_list_from_gconftool2_string(value: str) -> List[str]:
    r"""
    Take a raw string preference value as returned by gconftool-2
    and convert it to a list of strings.

    Handles escaped characters

    :param value: the raw value as returned by gconftool-2
    :return: the list of strings

    >>> prefs_list_from_gconftool2_string( # doctest: +ELLIPSIS
    ... '[Text,IMG_,,Sequences,Stored number,Four digits,Filename,Extension,UPPERCASE]')
    ... # doctest: +NORMALIZE_WHITESPACE
    ['Text', 'IMG_', '', 'Sequences', 'Stored number', 'Four digits', 'Filename', 'Extension',
    'UPPERCASE']
    >>> prefs_list_from_gconftool2_string('[Text,IMG_\,\\;+=|!@\,#^&*()$%/",,]')
    ['Text', 'IMG_,\\;+=|!@,#^&*()$%/"', '', '']
    >>> prefs_list_from_gconftool2_string('[Manila,Dubai,London]')
    ['Manila', 'Dubai', 'London']
    """
    # Trim the left and right square brackets
    value = value[1:-1]

    # Split on the comma, but not commas that were escaped.
    # Use a regex with a negative lookbehind assertion
    splits = re.split(r'(?<!\\),', value)
    # Replace the escaped commas with just plain commas
    return [s.replace('\\,', ',') for s in splits]


def pref_bool_from_gconftool2_string(value: str) -> bool:
    if value == 'true':
        return True
    elif value == 'false':
        return False
    raise ValueError


def remove_last_char_from_list_str(items: List[str]) -> List[str]:
    r"""
    Remove the last character from a list of strings, modifying the list in place,
    such that the last item is never empty

    :param items: the list to modify
    :return: in place copy

    >>> remove_last_char_from_list_str([' abc', 'def', 'ghi'])
    [' abc', 'def', 'gh']
    >>> remove_last_char_from_list_str([' abc', 'def', 'gh'] )
    [' abc', 'def', 'g']
    >>> remove_last_char_from_list_str([' abc', 'def', 'g'] )
    [' abc', 'def']
    >>> remove_last_char_from_list_str([' a'])
    [' ']
    >>> remove_last_char_from_list_str([' '])
    []
    >>> remove_last_char_from_list_str([])
    []
    """
    if items:
        if not items[-1]:
            items = items[:-1]
        else:
            items[-1] = items[-1][:-1]
            if items and not items[-1]:
                items = items[:-1]
    return items


def platform_c_maxint() -> int:
    """
    See http://stackoverflow.com/questions/13795758/what-is-sys-maxint-in-python-3

    :return: the maximum size of an int in C when compiled the same way Python was
    """
    return 2 ** (struct.Struct('i').size * 8 - 1) - 1


def commonprefix(*paths) -> str:
    """
    Python 3.4 compatible.

    Remove when Python 3.5 becomes the minimum.
    """

    return os.path.dirname(os.path.commonprefix(paths))


def _recursive_identify_depth(*paths, depth) -> int:
    basenames = [os.path.basename(path) for path in paths]
    if len(basenames) != len(set(basenames)):
        duplicates = _collect_duplicates(basenames, paths)

        for basename in duplicates:
            chop = len(basename) + 1
            chopped = (path[:-chop] for path in duplicates[basename])
            depth = max(depth, _recursive_identify_depth(*chopped, depth=depth + 1))
    return depth


def _collect_duplicates(basenames, paths):
    duplicates = defaultdict(list)
    for basename, path in zip(basenames, paths):
        duplicates[basename].append(path)
    return {basename: paths for basename, paths in duplicates.items() if len(paths) > 1}


def make_path_end_snippets_unique(*paths) -> List[str]:
    r"""
    Make list of path ends unique given possible common path endings.  
    
    A snippet starts from the end of the path, in extreme cases possibly up the path start. 

    :param paths: sequence of paths to generate unique end snippets for
    :return: list of unique snippets
    
    >>> p0 = '/home/damon/photos'
    >>> p1 = '/media/damon/backup1/photos'
    >>> p2 = '/media/damon/backup2/photos'
    >>> p3 = '/home/damon/videos'
    >>> p4 = '/media/damon/backup1/videos'
    >>> p5 = '/media/damon/backup2/videos'
    >>> p6 = '/media/damon/drive1/home/damon/photos'
    >>> s0 = make_path_end_snippets_unique(p0, p3)
    >>> print(s0)
    ['photos', 'videos']
    >>> s1 = make_path_end_snippets_unique(p0, p1, p2)
    >>> print(s1)
    ['damon/photos', 'backup1/photos', 'backup2/photos']
    >>> s2 = make_path_end_snippets_unique(p0, p1, p2, p3)
    >>> print(s2)
    ['damon/photos', 'backup1/photos', 'backup2/photos', 'videos']
    >>> s3 = make_path_end_snippets_unique(p3, p4, p5)
    >>> print(s3)
    ['damon/videos', 'backup1/videos', 'backup2/videos']
    >>> s4 = make_path_end_snippets_unique(p0, p1, p2, p3, p6)
    >>> print(s4) #doctest: +NORMALIZE_WHITESPACE
    ['/home/damon/photos', '/media/damon/backup1/photos', '/media/damon/backup2/photos', 'videos',
     'drive1/home/damon/photos']
    >>> s5 = make_path_end_snippets_unique(p1, p2, p3, p6)
    >>> print(s5)
    ['backup1/photos', 'backup2/photos', 'videos', 'damon/photos']
    """

    basenames = [os.path.basename(path) for path in paths]

    if len(basenames) != len(set(basenames)):
        names = []
        depths = defaultdict(int)
        duplicates = _collect_duplicates(basenames, paths)

        for basename, path in zip(basenames, paths):
            if basename in duplicates:
                depths[basename] = _recursive_identify_depth(*duplicates[basename], depth=0)

        for basename, path in zip(basenames, paths):
            depth = depths[basename]
            if depth:
                dirs = path.split(os.sep)
                index = len(dirs) - depth - 1
                name = (os.sep.join(dirs[max(index, 0): ]))
                if index > 1:
                    pass
                    # name = '...' + name
                elif index == 1:
                    name = os.sep + name
            else:
                name = basename
            names.append(name)
        return names
    else:
        return basenames

have_logged_os_release = False


def log_os_release() -> None:
    """
    Log the entired contents of /etc/os-release, but only if
    we didn't do so already.
    """

    global have_logged_os_release

    if not have_logged_os_release:
        try:
            with open('/etc/os-release', 'r') as f:
                for line in f:
                    logging.debug(line.rstrip('\n'))
        except:
            pass
        have_logged_os_release = True


def extract_file_from_tar(full_tar_path, member_filename) -> bool:
    """
    Extracts a file from a tar.gz and places it beside the tar file
    :param full_tar_path: path and filename of the tar.gz file
    :param member_filename: file wanted
    :return: True if successful, False otherwise
    """

    tar_dir, tar_name = os.path.split(full_tar_path)
    tar_name = tar_name[:len('.tar.gz') * -1]
    member = os.path.join(tar_name, member_filename)
    try:
        with tarfile.open(full_tar_path) as tar:
            tar.extractall(members=(tar.getmember(member),), path=tar_dir)
    except Exception:
        logging.error('Unable to extract %s from tarfile', member_filename)
        return False
    else:
        try:
            src = os.path.join(tar_dir, tar_name, member_filename)
            dst = os.path.join(tar_dir, member_filename)
            os.rename(src, dst)
            os.rmdir(os.path.join(tar_dir, tar_name))
            return True
        except OSError:
            logging.error('Unable to move %s to new location', member_filename)
            return False


def current_version_is_dev_version(current_version=None) -> bool:
    if current_version is None:
        current_version = pkg_resources.parse_version(__about__.__version__)
    return current_version.is_prerelease


def remove_topmost_directory_from_path(path: str) -> str:
    if os.sep not in path:
        return path
    return path[path[1:].find(os.sep) + 1:]


def arrow_locale() -> str:
    """
    Test if locale is suitable for use with Arrow.
    :return: Return user locale if it works with Arrow, else Arrow default ('en_us')
    """

    default = 'en_us'
    try:
        lang = locale.getdefaultlocale()[0]
    except Exception:
        return default

    try:
        arrow.locales.get_locale(lang)
        return lang
    except (ValueError, AttributeError):
        return default


def letters(x: int) -> str:
    """
    Return a letter representation of a positive number.

    Adapted from algorithm at
    http://en.wikipedia.org/wiki/Hexavigesimal

    >>> letters(0)
    'a'
    >>> letters(1)
    'b'
    >>> letters(2)
    'c'
    >>> letters(25)
    'z'
    >>> letters(26)
    'aa'
    >>> letters(27)
    'ab'
    >>> letters(28)
    'ac'
    """

    v = ''
    while x > 25:
        r = x % 26
        x = x // 26 - 1
        v = string.ascii_lowercase[r] + v
    v = string.ascii_lowercase[x] + v

    return v