1
  /   ˆ{‹„„-i[D<†ªoZƒ›Ù«Â523_‘I|¥‰ÓMdœv ©ƒ†ü+=ÂÁ     ‹      Õ•QoÚ0Çßû)<?ì.¡ŒjÔñÔ¤¡Q@”iš¦irbC¬&6²”}ú9$V•­Ó6­{òùò÷Ýý.‰�<‹Ç½Ù‡I½™]ÑäÝëá ‡pàýy žÅÕƒv3ú#ŒpæÜ²°Ùlš›ó¦6˜M!sEÞëŒL]“;Žé)}ô‘L0îWDœt¹ ÅÅÒszå¬äÂŸÛZàb-SaKÛ‰
Qh³­—°JsÔÛ°«¤Ò¨‚ú4Pç!‰æÛ2}ø§Sùˆ;–ì2–¦¡ÄehÍr¹Pvz‰)‘ÅY“Fdª•…$gê¶¹�sŒXî"üqÐò:pYyš†2#æ~Õ‹F—ã(ÆtÄ
A€=¬¹öš+L‡Ì:Th.çRð“â›J|#¿œWšXØÔÈ¥“Z¤à�QS�Û%ó¬/<AfŽU•„ÿ¬,ý®“«i4‹Óª%¼q¨î7^¦f„r(–F¤Îûjª*Ås•Øåå~‹êš�\dS„èAå/€ÎuÎ…9F=¹ÇÂtoWz¯´VÐê4Â°¾DáE7è ô4Œ\³c¨­§Â°R·JoÔ1„ÿžñ™­|(£ñ�æàzV+DA§Û
NbùÛôí?Àº‡ô¿â¬ÄÚÿÖ˜Vëßxäíõ\ °@ªù÷i\Ï¯b       ˆÀ¿Ã590¾ÂÁ N    ‹      Õ–moÚ0Çß÷Sx~±wpº§6dêH¥¡QŠZ¦iš&dbC¬&vd_ÈØ§ŸCRA;Ð²jšè›ø|ùû|??$ç¿¯Ó¯“Kòqz5"“ÏFÃ¡-€/½@8«§íŽp9¦„ÆˆÙ@Qí¢×Öf	Óˆ1MNÁ¢‘¶9rœø¥/8!~,w-ñQb"‚¡ââÑ:G+¹pãÖ¸XÉHØÒF‘B*RmÖuã�‚Ò\t Ó…0>T�\h¨cûsÍ×å”Þ¿ï¢lÒeóÍ,¥ic²b‰\ª>E�ÑÀ—é’Xõ)ÈH+ó„©»öR.(a	öé·áàú»ÓÆåèÀg$6bÑ§ïýñùu?¤Á˜¥Â¶_så441‹$Õ\.¤àÅ·•øVþ<0¬4¡°‘‘J­¶Rpˆ»¨‘NlÆë+G›]U%áZ=XŒÉÅÍåxoª%áeˆmvOØ,L˜
I(�ˆÐ½ªiªÐ/ÕÜfç÷]Rçjä2FÒ"{•˜«;¥µËèÖo/³kÑ`Óì&ù(«n§ûºåy­®G:oÏ¼w„ºÛøé¦Ÿ£¶¹Í„â3.¶ž¥ÖÑìñ6†ëõŽÎéÑè„µÑÁ=�Áä
e*f,B¹³;ÉÝöìq6Gë)Ziÿ†VÚÍÞÑ¡	Uþgø«v<Ã»tO`‘a¾s«~óãw|@ÕwNðG'ð¡ÿïXnÙr‡kÓm¾_ožˆÓ°Þ€ºóaSæùPU–¿ Šï€s¼
    C   ˆl–äY>”êhŠe@=q�ÜªbÑ¿�þ]}ð,Ü¯²HÛm¶@?Å„d!?Ä_ŠI|¥�+Wé3ž¯ÄÃ ?÷     """Japanese search language: includes routine to split words."""

# Python Version of TinySegmenter
# (http://chasen.org/~taku/software/TinySegmenter/)
# TinySegmenter is super compact Japanese tokenizer.
#
# TinySegmenter was originally developed by Taku Kudo <taku(at)chasen.org>.
# Python Version was developed by xnights <programming.magic(at)gmail.com>.
# For details, see http://programming-magic.com/?id=170

import os
import re
import sys
from typing import Any, Dict, List

try:
    import MeCab
    native_module = True
except ImportError:
    native_module = False

try:
    import janome.tokenizer
    janome_module = True
except ImportError:
    janome_module = False

from sphinx.errors import ExtensionError, SphinxError
from sphinx.search import SearchLanguage
from sphinx.util import import_object


class BaseSplitter:
    def __init__(self, options: Dict) -> None:
        self.options = options

    def split(self, input: str) -> List[str]:
        """
        :param str input:
        :return:
        :rtype: list[str]
        """
        raise NotImplementedError


class MecabSplitter(BaseSplitter):
    def __init__(self, options: Dict) -> None:
        super().__init__(options)
        self.ctypes_libmecab: Any = None
        self.ctypes_mecab: Any = None
        if not native_module:
            self.init_ctypes(options)
        else:
            self.init_native(options)
        self.dict_encode = options.get('dic_enc', 'utf-8')

    def split(self, input: str) -> List[str]:
        if native_module:
            result = self.native.parse(input)
        else:
            result = self.ctypes_libmecab.mecab_sparse_tostr(
                self.ctypes_mecab, input.encode(self.dict_encode))
        return result.split(' ')

    def init_native(self, options: Dict) -> None:
        param = '-Owakati'
        dict = options.get('dict')
        if dict:
            param += ' -d %s' % dict
        self.native = MeCab.Tagger(param)

    def init_ctypes(self, options: Dict) -> None:
        import ctypes.util

        lib = options.get('lib')

        if lib is None:
            if sys.platform.startswith('win'):
                libname = 'libmecab.dll'
            else:
                libname = 'mecab'
            libpath = ctypes.util.find_library(libname)
        elif os.path.basename(lib) == lib:
            libpath = ctypes.util.find_library(lib)
        else:
            libpath = None
            if os.path.exists(lib):
                libpath = lib
        if libpath is None:
            raise RuntimeError('MeCab dynamic library is not available')

        param = 'mecab -Owakati'
        dict = options.get('dict')
        if dict:
            param += ' -d %s' % dict

        fs_enc = sys.getfilesystemencoding() or sys.getdefaultencoding(