pinyinparser 1.0.0__py3-none-any.whl

pinyinparser-1.0.0.dist-info/METADATA

@@ -0,0 +1,7 @@
+Metadata-Version: 2.4
+Name: pinyinparser
+Version: 1.0.0
+Summary: 拼音解析器
+Author-email: WFLing-seaer <WFLing_seaer@163.com>
+License: WTFPL
+Requires-Python: >=3.12

pinyinparser-1.0.0.dist-info/RECORD

@@ -0,0 +1,5 @@
+pinyinparser.py,sha256=tGq4VRRcICA1U5gFDMMY_fC4SQ7_wweXzChzlLXucV0,34208
+pinyinparser-1.0.0.dist-info/METADATA,sha256=lIhxK-NOkWLPYZ1HFj19mfDRDFedIu6HmVKQyHqLKNE,177
+pinyinparser-1.0.0.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+pinyinparser-1.0.0.dist-info/top_level.txt,sha256=6EGG--K0y1GoDpu1iJyOdAleVlfhNtImpxo7QWjj5I4,13
+pinyinparser-1.0.0.dist-info/RECORD,,

pinyinparser-1.0.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

pinyinparser-1.0.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+pinyinparser

pinyinparser.py

@@ -0,0 +1,1035 @@
+import base64  # 解压那坨位图用的（嗯确实是解压因为a85编码比直接写hex还短这何尝不是一种压缩
+import re  # 过滤用的
+from collections.abc import Iterable
+from enum import IntEnum, StrEnum
+from functools import cache
+from itertools import chain, pairwise
+from typing import cast, overload
+from unicodedata import normalize
+from warnings import warn
+
+
+class Initial(IntEnum):
+    missing = 0
+    unspec = 0x0001
+    nul = 0x0002
+    H = 0x0003  # Hhm Hhng
+    R = 0x0004  # Rri
+    b = 0x0005
+    c = 0x0006
+    ch = 0x0007
+    d = 0x0008
+    f = 0x0009
+    g = 0x000A
+    h = 0x000B
+    j = 0x000C
+    k = 0x000D
+    l = 0x000E
+    m = 0x000F
+    n = 0x0010
+    p = 0x0011
+    q = 0x0012
+    r = 0x0013
+    s = 0x0014
+    sh = 0x0015
+    t = 0x0016
+    w = 0x0017
+    x = 0x0018
+    y = 0x0019
+    z = 0x001A
+    zh = 0x001B
+    M = 0x001C  # Mm Mn Mng
+
+
+class Final(IntEnum):
+    missing = 0
+    unspec = 0x0100
+    nul = 0x0200
+    a = 0x0300
+    ai = 0x0400
+    an = 0x0500
+    ang = 0x0600
+    ao = 0x0700
+    e = 0x0800
+    ei = 0x0900
+    en = 0x0A00
+    eng = 0x0B00
+    er = 0x0C00
+    hm = 0x0D00
+    hng = 0x0E00
+    i = 0x0F00  # [i] ji qi xi
+    ia = 0x1000
+    ian = 0x1100
+    iang = 0x1200
+    iao = 0x1300
+    ieh = 0x1400
+    ien = 0x1500
+    ii = 0x1600  # [z]/[ɿ] zi ci si
+    ieng = 0x1700
+    iong = 0x1800
+    iou = 0x1900
+    ng = 0x1A00  # [ŋ̊]
+    o = 0x1B00
+    ong = 0x1C00
+    ou = 0x1D00
+    ri = 0x1E00  # [ʅ] zhi chi shi [ʐ]/[ʅ] ri
+    u = 0x1F00
+    ua = 0x2000
+    uai = 0x2100
+    uan = 0x2200
+    uang = 0x2300
+    uei = 0x2400
+    uen = 0x2500
+    ueng = 0x2600
+    uo = 0x2700
+    v = 0x2800
+    van = 0x2900
+    veh = 0x2A00
+    ven = 0x2B00
+    m = 0x2C00
+    n = 0x2D00
+    eh = 0x2E00  # ê
+
+
+class Tone(IntEnum):
+    missing = 0
+    unspec = 0x0020
+    nul = 0x0040
+    t1 = 0x0060
+    t2 = 0x0080
+    t3 = 0x00A0
+    t4 = 0x00C0
+    t5 = 0x00E0
+
+
+FINAL2STR = {
+    Final.ii: "i",
+    Final.ri: "i",
+    Final.iou: "iu",
+    Final.uei: "ui",
+    Final.ien: "in",
+    Final.uen: "un",
+    Final.v: "ü",
+    Final.veh: "üe",
+    Final.van: "üan",
+    Final.ven: "ün",
+    Final.eh: "ê",
+    Final.ieh: "ie",
+    Final.ieng: "ing",
+}
+TONE2STR = {
+    Tone.t1: "1",
+    Tone.t2: "2",
+    Tone.t3: "3",
+    Tone.t4: "4",
+    Tone.t5: "5",
+}
+FINAL2TONED = {
+    Tone.t1: {"a": "ā", "e": "ē", "i": "ī", "o": "ō", "u": "ū", "ü": "ǖ", "ê": "ê̄", "m": "m̄", "n": "n̄"},
+    Tone.t2: {"a": "á", "e": "é", "i": "í", "o": "ó", "u": "ú", "ü": "ǘ", "ê": "ế", "m": "ḿ", "n": "ń"},
+    Tone.t3: {"a": "ǎ", "e": "ě", "i": "ǐ", "o": "ǒ", "u": "ǔ", "ü": "ǚ", "ê": "ê̌", "m": "m̌", "n": "ň"},
+    Tone.t4: {"a": "à", "e": "è", "i": "ì", "o": "ò", "u": "ù", "ü": "ǜ", "ê": "ề", "m": "m̀", "n": "ǹ"},
+}
+SYLLMAP_Y = {
+    Final.ia: "ya",
+    Final.ieh: "ye",
+    Final.iao: "yao",
+    Final.iou: "you",
+    Final.ian: "yan",
+    Final.ien: "yin",
+    Final.iang: "yang",
+    Final.ieng: "ying",
+    Final.iong: "yong",
+    Final.v: "yu",
+    Final.veh: "yue",
+    Final.van: "yuan",
+    Final.ven: "yun",
+}
+SYLLMAP_W = {
+    Final.ua: "wa",
+    Final.uo: "wo",
+    Final.uai: "wai",
+    Final.uei: "wei",
+    Final.uan: "wan",
+    Final.uen: "wen",
+    Final.uang: "wang",
+    Final.ueng: "weng",
+}
+SYLL1SEP: set[Final] = {
+    Final.m,
+    Final.n,
+    Final.ng,
+    Final.an,
+    Final.ang,
+    Final.en,
+    Final.eng,
+    Final.er,
+    Final.hm,
+    Final.hng,
+    Final.i,
+    Final.ri,
+    Final.ii,
+    Final.ian,
+    Final.iang,
+    Final.ien,
+    Final.ieng,
+    Final.iong,
+    Final.ong,
+    Final.uan,
+    Final.uang,
+    Final.uen,
+    Final.ueng,
+    Final.van,
+    Final.ven,
+}
+SYLL2SEP: set[Final] = {Final.m, Final.n, Final.ng, Final.eh}
+SYLLSEP: dict[Final, set[Final]] = {
+    Final.a: {Final.o, Final.ong, Final.ou},
+    Final.ia: {Final.o, Final.ong, Final.ou},
+    Final.u: {Final.a, Final.ai, Final.an, Final.ang, Final.ao, Final.eng, Final.o, Final.ong, Final.ou},
+    Final.v: {Final.an, Final.ang, Final.e, Final.ei, Final.en, Final.eng, Final.er, Final.o},
+}
+
+
+class ToneStyle(StrEnum):
+    ABOVE = "above"
+    RIGHT = "right"
+    AFTER = "after"
+
+
+class IncompatibleWarning(UserWarning):
+    pass
+
+
+class _SyllMeta(type):
+    def __getattr__(self, name):
+        try:
+            return parse_single(name)
+        except ValueError:
+            raise AttributeError(name)
+
+
+class Syllable(metaclass=_SyllMeta):
+    @overload
+    def __init__(self, i: Initial = ..., f: Final = ..., t: Tone = ...): ...
+
+    @overload
+    def __init__(self, i: int): ...
+
+    @overload
+    def __init__(self, i: str): ...
+
+    def __init__(self, i: int | Initial | str = Initial.missing, f: Final = Final.missing, t: Tone = Tone.missing):
+        if isinstance(i, str):
+            if (f is not Final.missing) or (t is not Tone.missing):
+                raise ValueError("不能同时使用字符串初始化和声韵调初始化。")
+            sp = parse_single(i)
+            self.initial = sp.initial
+            self.final = sp.final
+            self.tone = sp.tone
+        elif isinstance(i, Initial):
+            self.initial = i
+            self.final = f
+            self.tone = t
+        else:
+            if (f is not Final.missing) or (t is not Tone.missing):
+                raise ValueError("不能同时使用uint16初始化和声韵调初始化。")
+            if i & ~0x3FFF:
+                raise ValueError("int中存在无效的位。使用整数初始化时仅接受高2位为0的uint16。")
+            try:
+                self.initial = Initial(i & 0x001F)
+                self.final = Final(i & 0x3F00)
+                self.tone = Tone(i & 0x00E0)
+            except ValueError:
+                raise ValueError("无效的uint16。")
+
+    def __int__(self):
+        return self.initial | self.final | self.tone
+
+    def __repr__(self):
+        return f"<{self.initial.name}·{self.final.name}·{self.tone.name.removeprefix("t")}{"" if self.is_valid() else "(N/E)"}>"
+
+    @cache
+    def to_str(self, tone_style: ToneStyle = ToneStyle.ABOVE, NO_INCOMPAT_WARNING: bool = False) -> str:
+        match self.initial:
+            case Initial.missing | Initial.unspec | Initial.nul:
+                initial_str = ""
+            case Initial.R:
+                initial_str = "r"
+            case Initial.H:
+                initial_str = "h"
+            case _:
+                initial_str = self.initial.name
+
+        if self.final in (Final.missing, Final.unspec, Final.nul):
+            final_str = ""
+        else:
+            final_str = FINAL2STR.get(self.final, self.final.name)
+
+        match self.initial:
+            case Initial.y:
+                base_str = SYLLMAP_Y.get(self.final, f"y{final_str}")
+            case Initial.w:
+                base_str = SYLLMAP_W.get(self.final, f"w{final_str}")
+            case Initial.j | Initial.q | Initial.x:
+                base_str = initial_str + final_str.replace("ü", "u")
+            case Initial.R if self.final == Final.ri:
+                base_str = "ri"
+            case Initial.H if self.final in (Final.hm, Final.hng):
+                base_str = final_str
+            case Initial.M if self.final in (Final.m, Final.n, Final.ng):
+                base_str = final_str
+            case _:
+                base_str = initial_str + final_str
+
+        if not base_str:
+            return ""
+
+        if tone_style == ToneStyle.AFTER:
+            return base_str + TONE2STR.get(self.tone, "")
+
+        if self.tone in (Tone.missing, Tone.unspec, Tone.nul):
+            return base_str
+
+        if tone_style == ToneStyle.ABOVE and self.tone == Tone.t5:
+            return base_str
+
+        toned = FINAL2TONED.get(self.tone, {})
+
+        for v in ["a", "ê", "e", "o"]:
+            if (pos := base_str.find(v)) != -1:
+                break
+        else:
+            for pos in range(len(base_str) - 1, -1, -1):
+                if base_str[pos] in {"i", "u", "ü"}:
+                    break
+            else:
+                for v in ["n", "m"]:
+                    if (pos := base_str.find(v)) != -1:
+                        break
+                else:
+                    return base_str
+
+        if tone_style == ToneStyle.ABOVE:
+            if base_str[pos] in toned:
+                return f"{base_str[:pos]}{toned[base_str[pos]]}{base_str[pos+1:]}"
+            return base_str
+        elif tone_style == ToneStyle.RIGHT:
+            if not NO_INCOMPAT_WARNING:
+                warn(
+                    "注意：使用RIGHT（数字附标）模式产出的拼音字符串并不通用，且不再能被本解析器解析。如果你明确知道你在做什么，可以传入NO_INCOMPAT_WARNING=True以关闭此警告。",
+                    IncompatibleWarning,
+                )
+            return f"{base_str[:pos+1]}{TONE2STR.get(self.tone, '')}{base_str[pos+1:]}"
+
+        return base_str
+
+    def __str__(self):
+        return self.to_str()
+
+    def __bool__(self):
+        return bool(self.initial or self.final or self.tone)
+
+    def __eq__(self, other):
+        return self.initial == other.initial and self.final == other.final and self.tone == other.tone
+
+    __hash__ = __int__
+
+    def is_complete(self):
+        return bool(self.initial and self.final and self.tone)
+
+    def need_sep(self, prev: Syllable) -> bool:
+        if self.initial and (self.initial not in {Initial.nul, Initial.unspec, Initial.M}):
+            return False  # 有有效声母隔着则必不需要
+
+        if prev.final in SYLL1SEP or self.final in SYLL2SEP:
+            return True
+
+        return prev.final in SYLLSEP and self.final in SYLLSEP[prev.final]
+
+    def is_valid(self):
+        return _check_syllable_valid(int(self))
+
+    def copy(self):
+        return Syllable(self.initial, self.final, self.tone)
+
+    def _mreject(self, other: Syllable, check_initial: bool = False):
+        return bool(
+            (self.initial and other.initial)
+            or (self.tone and other.tone)
+            or ((self.final or self.tone) and (other.initial or other.final))
+            or (
+                check_initial
+                and (self.final or other.final or ((self.initial or other.initial) and (self.tone or other.tone)))
+                and ((self.initial == Initial.nul) or (not self.initial))
+                and ((other.initial == Initial.nul) or (not other.initial))
+            )
+        )
+        # 化简自(s1 and o1)or(s2 and o2)or(s3 and o3)or(s2 and o1)or(s3 and o1)or(s3 and o2)or(check_initial and(s1 or o1 or s2 or o2)and(s2 or o2 or s3 or o3)and((self.initial==Initial.nul)or(not s1))and((other.initial==Initial.nul)or(not o1)))
+        # 其中s1=self.initial!=Initial.missing s2=self.final!=Final.missing s3=self.tone!=Tone.missing o1=other.initial!=Initial.missing o2=other.final!=Final.missing o3=other.tone!=Tone.missing
+
+    def _merge(self, other: Syllable):
+        if other.initial != Initial.missing:
+            self.initial = other.initial
+        if other.final != Final.missing:
+            self.final = other.final
+        if other.tone != Tone.missing:
+            self.tone = other.tone
+
+
+TOKENS = {
+    "iang": [0x1200],
+    "iāng": [0x1260],  # 江
+    "iáng": [0x1280],  # 凉
+    "iǎng": [0x12A0],  # 抢
+    "iàng": [0x12C0],  # 呛
+    "iong": [0x1800],
+    "iōng": [0x1860],  # 凶
+    "ióng": [0x1880],  # 穷
+    "iǒng": [0x18A0],  # 涌
+    "iòng": [0x18C0],  # 用
+    "uang": [0x2300],
+    "uāng": [0x2360],  # 光
+    "uáng": [0x2380],  # 狂
+    "uǎng": [0x23A0],  # 广
+    "uàng": [0x23C0],  # 旷
+    "ueng": [0x2600],
+    "uēng": [0x2660],  # 翁
+    "uéng": [0x2680],  # 不存在，然而拒绝解析有违直觉，因而保留，下同
+    "uěng": [0x26A0],  # 塕
+    "uèng": [0x26C0],  # 瓮
+    "juan": [0x290C],
+    "juān": [0x296C],  # 捐
+    "juán": [0x298C],  # 不存在
+    "juǎn": [0x29AC],  # 卷
+    "juàn": [0x29CC],  # 倦
+    "quan": [0x2912],
+    "quān": [0x2972],  # 圈
+    "quán": [0x2992],  # 全
+    "quǎn": [0x29B2],  # 犬
+    "quàn": [0x29D2],  # 劝
+    "xuan": [0x2918],
+    "xuān": [0x2978],  # 宣
+    "xuán": [0x2998],  # 悬
+    "xuǎn": [0x29B8],  # 选
+    "xuàn": [0x29D8],  # 炫
+    "yang": [0x1219],
+    "yāng": [0x1279],  # 央
+    "yáng": [0x1299],  # 阳
+    "yǎng": [0x12B9],  # 养
+    "yàng": [0x12D9],  # 样
+    "ying": [0x1719],
+    "yīng": [0x1779],  # 英
+    "yíng": [0x1799],  # 赢
+    "yǐng": [0x17B9],  # 影
+    "yìng": [0x17D9],  # 映
+    "yong": [0x1819],
+    "yōng": [0x1879],  # 拥
+    "yóng": [0x1899],  # 颙
+    "yǒng": [0x18B9],  # 泳
+    "yòng": [0x18D9],  # 用
+    "wang": [0x2317],
+    "wāng": [0x2377],  # 汪
+    "wáng": [0x2397],  # 王
+    "wǎng": [0x23B7],  # 网
+    "wàng": [0x23D7],  # 忘
+    "weng": [0x2617],
+    "wēng": [0x2677],  # 翁
+    "wéng": [0x2697],  # 不存在
+    "wěng": [0x26B7],  # 塕
+    "wèng": [0x26D7],  # 瓮
+    "yuan": [0x2919],
+    "yuān": [0x2979],  # 冤
+    "yuán": [0x2999],  # 圆
+    "yuǎn": [0x29B9],  # 远
+    "yuàn": [0x29D9],  # 怨
+    "ang": [0x0602, 0x0600],
+    "āng": [0x0662, 0x0660],  # 刚 肮
+    "áng": [0x0682, 0x0680],  # 扛 昂
+    "ǎng": [0x06A2, 0x06A0],  # 莽 䇦
+    "àng": [0x06C2, 0x06C0],  # 抗 盎
+    "eng": [0x0B02, 0x0B00],
+    "ēng": [0x0B62, 0x0B60],  # 庚 鞥
+    "éng": [0x0B82, 0x0B80],  # 横 不存在
+    "ěng": [0x0BA2, 0x0BA0],  # 冷 不存在
+    "èng": [0x0BC2, 0x0BC0],  # 赠 不存在
+    "ian": [0x1100],
+    "iān": [0x1160],  # 先
+    "ián": [0x1180],  # 咸
+    "iǎn": [0x11A0],  # 显
+    "iàn": [0x11C0],  # 现
+    "iao": [0x1300],
+    "iāo": [0x1360],  # 交
+    "iáo": [0x1380],  # 嚼
+    "iǎo": [0x13A0],  # 缴
+    "iào": [0x13C0],  # 叫
+    "ing": [0x1700],
+    "īng": [0x1760],  # 星
+    "íng": [0x1780],  # 形
+    "ǐng": [0x17A0],  # 醒
+    "ìng": [0x17C0],  # 幸
+    "ong": [0x1C00],
+    "ōng": [0x1C60],  # 东
+    "óng": [0x1C80],  # 龙
+    "ǒng": [0x1CA0],  # 拢
+    "òng": [0x1CC0],  # 冻
+    "uai": [0x2100],
+    "uāi": [0x2160],  # 乖
+    "uái": [0x2180],  # 淮
+    "uǎi": [0x21A0],  # 拐
+    "uài": [0x21C0],  # 坏
+    "uan": [0x2200],
+    "uān": [0x2260],  # 欢
+    "uán": [0x2280],  # 环
+    "uǎn": [0x22A0],  # 缓
+    "uàn": [0x22C0],  # 换
+    "van": [0x2900],
+    "vān": [0x2960],  # 捐
+    "ván": [0x2980],  # 全
+    "vǎn": [0x29A0],  # 犬
+    "vàn": [0x29C0],  # 倦
+    "üan": [0x2900],
+    "üān": [0x2960],
+    "üán": [0x2980],
+    "üǎn": [0x29A0],
+    "üàn": [0x29C0],
+    "hng": [0x0EE3],  # 哼
+    "zhi": [0x1E1B],  # [ʅ]
+    "zhī": [0x1E7B],  # 支
+    "zhí": [0x1E9B],  # 直
+    "zhǐ": [0x1EBB],  # 纸
+    "zhì": [0x1EDB],  # 至
+    "chi": [0x1E07],  # [ʅ]
+    "chī": [0x1E67],  # 吃
+    "chí": [0x1E87],  # 持
+    "chǐ": [0x1EA7],  # 尺
+    "chì": [0x1EC7],  # 赤
+    "shi": [0x1E15],  # [ʅ]
+    "shī": [0x1E75],  # 诗
+    "shí": [0x1E95],  # 石
+    "shǐ": [0x1EB5],  # 史
+    "shì": [0x1ED5],  # 事
+    "jue": [0x2A0C],
+    "juē": [0x2A6C],  # 撅
+    "jué": [0x2A8C],  # 绝
+    "juě": [0x2AAC],  # 蹶
+    "juè": [0x2ACC],  # 倔
+    "que": [0x2A12],
+    "quē": [0x2A72],  # 缺
+    "qué": [0x2A92],  # 瘸
+    "quě": [0x2AB2],  # 不存在
+    "què": [0x2AD2],  # 雀
+    "xue": [0x2A18],
+    "xuē": [0x2A78],  # 薛
+    "xué": [0x2A98],  # 学
+    "xuě": [0x2AB8],  # 雪
+    "xuè": [0x2AD8],  # 谑
+    "jun": [0x2B0C],
+    "jūn": [0x2B6C],  # 君
+    "jún": [0x2B8C],  # 不存在
+    "jǔn": [0x2BAC],  # 𢉦（RD广军）
+    "jùn": [0x2BCC],  # 郡
+    "qun": [0x2B12],
+    "qūn": [0x2B72],  # 逡
+    "qún": [0x2B92],  # 群
+    "qǔn": [0x2BB2],  # 䊎
+    "qùn": [0x2BD2],  # 不存在
+    "xun": [0x2B18],
+    "xūn": [0x2B78],  # 勋
+    "xún": [0x2B98],  # 寻
+    "xǔn": [0x2BB8],  # 不存在
+    "xùn": [0x2BD8],  # 巽
+    "yao": [0x1319],
+    "yāo": [0x1379],  # 邀
+    "yáo": [0x1399],  # 摇
+    "yǎo": [0x13B9],  # 咬
+    "yào": [0x13D9],  # 药
+    "you": [0x1919],
+    "yōu": [0x1979],  # 优
+    "yóu": [0x1999],  # 游
+    "yǒu": [0x19B9],  # 有
+    "yòu": [0x19D9],  # 右
+    "yan": [0x1119],
+    "yān": [0x1179],  # 烟
+    "yán": [0x1199],  # 盐
+    "yǎn": [0x11B9],  # 眼
+    "yàn": [0x11D9],  # 验
+    "yin": [0x1519],
+    "yīn": [0x1579],  # 阴
+    "yín": [0x1599],  # 银
+    "yǐn": [0x15B9],  # 饮
+    "yìn": [0x15D9],  # 印
+    "wai": [0x2117],
+    "wāi": [0x2177],  # 歪
+    "wái": [0x2197],  # 不存在
+    "wǎi": [0x21B7],  # 𨂿
+    "wài": [0x21D7],  # 外
+    "wei": [0x2417],
+    "wēi": [0x2477],  # 威
+    "wéi": [0x2497],  # 维
+    "wěi": [0x24B7],  # 尾
+    "wèi": [0x24D7],  # 味
+    "wan": [0x2217],
+    "wān": [0x2277],  # 弯
+    "wán": [0x2297],  # 完
+    "wǎn": [0x22B7],  # 碗
+    "wàn": [0x22D7],  # 万
+    "wen": [0x2517],
+    "wēn": [0x2577],  # 温
+    "wén": [0x2597],  # 文
+    "wěn": [0x25B7],  # 稳
+    "wèn": [0x25D7],  # 问
+    "yue": [0x2A19],
+    "yuē": [0x2A79],  # 约
+    "yué": [0x2A99],  # 块（音yué义不详，但字统有记载因而算进来了）
+    "yuě": [0x2AB9],  # 哕
+    "yuè": [0x2AD9],  # 月
+    "yun": [0x2B19],
+    "yūn": [0x2B79],  # 晕
+    "yún": [0x2B99],  # 云
+    "yǔn": [0x2BB9],  # 允
+    "yùn": [0x2BD9],  # 韵
+    "zi": [0x161A],
+    "zī": [0x167A],  # 兹
+    "zí": [0x169A],  # 不存在
+    "zǐ": [0x16BA],  # 紫
+    "zì": [0x16DA],  # 字
+    "ci": [0x1606],
+    "cī": [0x1666],  # 呲
+    "cí": [0x1686],  # 词
+    "cǐ": [0x16A6],  # 此
+    "cì": [0x16C6],  # 次
+    "si": [0x1614],
+    "sī": [0x1674],  # 丝
+    "sí": [0x1694],  # 不存在
+    "sǐ": [0x16B4],  # 死
+    "sì": [0x16D4],  # 四
+    "ri": [0x1E04, 0x1E00],  # [ʐ]/[ʅ]
+    "rī": [0x1E64, 0x1E60],  # 痴 不存在
+    "rí": [0x1E84, 0x1E80],  # 迟 不存在
+    "rǐ": [0x1EA4, 0x1EA0],  # 齿 不存在
+    "rì": [0x1EC4, 0x1EC0],  # 斥 日
+    "hm": [0x0DE3],
+    "ai": [0x0402, 0x0400],
+    "āi": [0x0462, 0x0460],  # 该 挨
+    "ái": [0x0482, 0x0480],  # 孩 皑
+    "ǎi": [0x04A2, 0x04A0],  # 改 矮
+    "ài": [0x04C2, 0x04C0],  # 骇 爱
+    "an": [0x0502, 0x0500],
+    "ān": [0x0562, 0x0560],  # 潘 安
+    "án": [0x0582, 0x0580],  # 盘 儑
+    "ǎn": [0x05A2, 0x05A0],  # 懒 俺
+    "àn": [0x05C2, 0x05C0],  # 烂 暗
+    "ao": [0x0702, 0x0700],
+    "āo": [0x0762, 0x0760],  # 高 凹
+    "áo": [0x0782, 0x0780],  # 豪 熬
+    "ǎo": [0x07A2, 0x07A0],  # 好 拗
+    "ào": [0x07C2, 0x07C0],  # 告 傲
+    "ei": [0x0902, 0x0900],
+    "ēi": [0x0962, 0x0960],  # 飞 不存在（欸等字在eh）
+    "éi": [0x0982, 0x0980],  # 肥 不存在
+    "ěi": [0x09A2, 0x09A0],  # 匪 不存在
+    "èi": [0x09C2, 0x09C0],  # 费 不存在
+    "en": [0x0A02, 0x0A00],
+    "ēn": [0x0A62, 0x0A60],  # 奔 恩
+    "én": [0x0A82, 0x0A80],  # 盆 不存在
+    "ěn": [0x0AA2, 0x0AA0],  # 本 不存在
+    "èn": [0x0AC2, 0x0AC0],  # 笨 摁
+    "er": [0x0C02, 0x0C00],
+    "ēr": [0x0C62, 0x0C60],  # 不存在 不存在
+    "ér": [0x0C82, 0x0C80],  # 不存在 儿
+    "ěr": [0x0CA2, 0x0CA0],  # 不存在 尔
+    "èr": [0x0CC2, 0x0CC0],  # 不存在 佴
+    "ia": [0x1000],
+    "iā": [0x1060],  # 家
+    "iá": [0x1080],  # 夹
+    "iǎ": [0x10A0],  # 贾
+    "ià": [0x10C0],  # 架
+    "ie": [0x1400],
+    "iē": [0x1460],  # 街
+    "ié": [0x1480],  # 截
+    "iě": [0x14A0],  # 解
+    "iè": [0x14C0],  # 借
+    "ii": [0x1600],  # [z]/[ɿ]
+    "in": [0x1500],
+    "īn": [0x1560],  # 侵
+    "ín": [0x1580],  # 琴
+    "ǐn": [0x15A0],  # 寝
+    "ìn": [0x15C0],  # 沁
+    "iu": [0x1900],
+    "iū": [0x1960],  # 秋
+    "iú": [0x1980],  # 求
+    "iǔ": [0x19A0],  # 朽
+    "iù": [0x19C0],  # 锈
+    "ng": [0x1A1C],  # [ŋ̊]，仅见于唔、嗯二字
+    "n̄g": [0x1A7C],  # 然而还是为了不违反直觉，在这插一个n1g
+    "ńg": [0x1A9C],
+    "ňg": [0x1ABC],
+    "ǹg": [0x1ADC],  # n+macron没有单字符表示，且n1g不存在，
+    "ou": [0x1D02, 0x1D00],
+    "ōu": [0x1D62, 0x1D60],  # 沟 欧
+    "óu": [0x1D82, 0x1D80],  # 楼 吽
+    "ǒu": [0x1DA2, 0x1DA0],  # 篓 偶
+    "òu": [0x1DC2, 0x1DC0],  # 够 沤
+    "ua": [0x2000],
+    "uā": [0x2060],  # 花
+    "uá": [0x2080],  # 滑
+    "uǎ": [0x20A0],  # 垮
+    "uà": [0x20C0],  # 跨
+    "ui": [0x2400],
+    "uī": [0x2460],  # 灰
+    "uí": [0x2480],  # 回
+    "uǐ": [0x24A0],  # 毁
+    "uì": [0x24C0],  # 会
+    "un": [0x2500],
+    "ūn": [0x2560],  # 昆
+    "ún": [0x2580],  # 仑
+    "ǔn": [0x25A0],  # 捆
+    "ùn": [0x25C0],  # 论
+    "uo": [0x2700],
+    "uō": [0x2760],  # 锅
+    "uó": [0x2780],  # 活
+    "uǒ": [0x27A0],  # 火
+    "uò": [0x27C0],  # 过
+    "ve": [0x2A00],
+    "vē": [0x2A60],  # 薛
+    "vé": [0x2A80],  # 学
+    "vě": [0x2AA0],  # 雪
+    "vè": [0x2AC0],  # 谑
+    "üe": [0x2A00],
+    "üē": [0x2A60],
+    "üé": [0x2A80],
+    "üě": [0x2AA0],
+    "üè": [0x2AC0],
+    "vn": [0x2B00],
+    "ün": [0x2B00],
+    "ǖn": [0x2B60],  # 逡
+    "ǘn": [0x2B80],  # 群
+    "ǚn": [0x2BA0],  # 允
+    "ǜn": [0x2BC0],  # 孕
+    "ju": [0x280C],
+    "jū": [0x286C],  # 居
+    "jú": [0x288C],  # 局
+    "jǔ": [0x28AC],  # 举
+    "jù": [0x28CC],  # 句
+    "qu": [0x2812],
+    "qū": [0x2872],  # 区
+    "qú": [0x2892],  # 渠
+    "qǔ": [0x28B2],  # 取
+    "qù": [0x28D2],  # 去
+    "xu": [0x2818],
+    "xū": [0x2878],  # 需
+    "xú": [0x2898],  # 徐
+    "xǔ": [0x28B8],  # 许
+    "xù": [0x28D8],  # 序
+    "yi": [0x0F19],
+    "yī": [0x0F79],  # 一
+    "yí": [0x0F99],  # 疑
+    "yǐ": [0x0FB9],  # 以
+    "yì": [0x0FD9],  # 忆
+    "ya": [0x1019],
+    "yā": [0x1079],  # 压
+    "yá": [0x1099],  # 牙
+    "yǎ": [0x10B9],  # 雅
+    "yà": [0x10D9],  # 亚
+    "ye": [0x1419],
+    "yē": [0x1479],  # 噎
+    "yé": [0x1499],  # 爷
+    "yě": [0x14B9],  # 野
+    "yè": [0x14D9],  # 页
+    "wu": [0x1F17],
+    "wū": [0x1F77],  # 屋
+    "wú": [0x1F97],  # 无
+    "wǔ": [0x1FB7],  # 舞
+    "wù": [0x1FD7],  # 物
+    "wa": [0x2017],
+    "wā": [0x2077],  # 洼
+    "wá": [0x2097],  # 娃
+    "wǎ": [0x20B7],  # 瓦
+    "wà": [0x20D7],  # 袜
+    "wo": [0x2717],
+    "wō": [0x2777],  # 窝
+    "wó": [0x2797],  # 不存在
+    "wǒ": [0x27B7],  # 我
+    "wò": [0x27D7],  # 卧
+    "yu": [0x2819],
+    "yū": [0x2879],  # 淤
+    "yú": [0x2899],  # 于
+    "yǔ": [0x28B9],  # 与
+    "yù": [0x28D9],  # 玉
+    "zh": [0x001B],
+    "ch": [0x0007],
+    "sh": [0x0015],
+    "ê": [0x2E00],
+    "ê̄": [0x2E60],
+    "ế": [0x2E80],
+    "ê̌": [0x2EA0],
+    "ề": [0x2EC0],  # U+1EC1 一、二、三声没有结合形式，只能用组合字符；四声有结合形式
+    "a": [0x0302, 0x0300],
+    "ā": [0x0362, 0x0360],  # 妈 啊
+    "á": [0x0382, 0x0380],  # 麻 啊
+    "ǎ": [0x03A2, 0x03A0],  # 马 啊
+    "à": [0x03C2, 0x03C0],  # 骂 啊
+    "e": [0x0802, 0x0800],
+    "ē": [0x0862, 0x0860],  # 歌 婀
+    "é": [0x0882, 0x0880],  # 隔 俄
+    "ě": [0x08A2, 0x08A0],  # 舸 𫫇
+    "è": [0x08C2, 0x08C0],  # 各 恶
+    "i": [0x0F00],  # [i]
+    "ī": [0x0F60],  # 机
+    "í": [0x0F80],  # 急
+    "ǐ": [0x0FA0],  # 挤
+    "ì": [0x0FC0],  # 记
+    "o": [0x1B02, 0x1B00],  # 咯、哦
+    "ō": [0x1B62, 0x1B60],  # 此四音是否统合到uo尚有待商榷，暂定为不统合
+    "ó": [0x1B82, 0x1B80],
+    "ǒ": [0x1BA2, 0x1BA0],
+    "ò": [0x1BC2, 0x1BC0],
+    "u": [0x1F00],
+    "ū": [0x1F60],  # 孤
+    "ú": [0x1F80],  # 湖
+    "ǔ": [0x1FA0],  # 虎
+    "ù": [0x1FC0],  # 固
+    "v": [0x2800],
+    "ü": [0x2800],
+    "ǖ": [0x2860],  # 屈
+    "ǘ": [0x2880],  # 渠
+    "ǚ": [0x28A0],  # 取
+    "ǜ": [0x28C0],  # 去
+    "b": [0x0005],
+    "p": [0x0011],
+    "m": [0x000F],  # 也可为韵母m[m̥]，仅见于呒、呣二字
+    "f": [0x0009],
+    "d": [0x0008],
+    "t": [0x0016],
+    "n": [0x0010],  # 也可为韵母n[n̥]/[ɰ̃]，仅见于唔、嗯二字
+    "l": [0x000E],
+    "g": [0x000A],
+    "h": [0x000B],
+    "j": [0x000C],
+    "k": [0x000D],
+    "q": [0x0012],
+    "x": [0x0018],
+    "r": [0x0013],
+    "z": [0x001A],
+    "c": [0x0006],
+    "s": [0x0014],
+    "y": [0x0019],  # 伪声母y
+    "w": [0x0017],  # 伪声母w
+    "1": [0x0060],
+    "2": [0x0080],
+    "3": [0x00A0],
+    "4": [0x00C0],
+    "5": [0x00E0],  # 轻声
+    "̄": [0x0060],  # ISO 7098:2015 7.1节
+    "́": [0x0080],
+    "̌": [0x00A0],
+    "̀": [0x00C0],
+    # 以下按需启用
+    # "?": [0x0001, 0x0100, 0x0020],
+    # ".": [0x0121],  # 通配
+    # "*": [0x0001, 0x0100],  # 声母韵母通配
+    # "0": [0x0020],  # 声调通配
+    # "/": [0x0002],  # 零声母
+    # "&": [0x0004],  # 伪声母R
+    "n̄": [0x2D7C],  # 多字符，不存在
+    "ń": [0x2D9C],
+    "ň": [0x2DBC],
+    "ǹ": [0x2DDC],
+    "n1": [0x2D7C],  # 不存在
+    "n2": [0x2D9C],
+    "n3": [0x2DBC],
+    "n4": [0x2DDC],
+    "m̄": [0x2C7C],  # 多字符，不存在
+    "ḿ": [0x2C9C],  # 只有ḿ有单字符表示
+    "m̌": [0x2CBC],  # 多字符，不存在
+    "m̀": [0x2CDC],  # 多字符
+    "m1": [0x2C7C],  # 不存在
+    "m2": [0x2C9C],
+    "m3": [0x2CBC],  # 不存在
+    "m4": [0x2CDC],
+}
+VALID_CHARS = set(chain.from_iterable(TOKENS.keys()))
+VALID_CHARS_RE_DEFAULT = re.compile(f"[{re.escape("".join(VALID_CHARS))}]*")
+VALID_SYLLABLES = base64.a85decode(
+    b'q>2-6l15_oq>2-6q=>R.a`&4$zzzH&EXE3V!^_H2AQ`H2AQ`84!jfzzzq<K"&Z18e8q>238q>238#Qb,0zzzq>238Z2,(5q>238q>238i(sa^zzzH2AQ`H1DjUH2AWbH2AWb!"],3zzz@64l!@7pq0+U\\VY0j>\\$6$X"DzzzadNb@Ja`pGamf?2RIC1X#S6t9zzz\\Z6,@YbMO,WJ_5gl2(ei+<UXbzzzq7@[M\\FLO<kkPJdq"#O/#Tsrazzzz!<<*"!<<*"!<<*"!<<*"zzzzzzz!WW3#zzzzzzz!WW3#zzz8G(:08G(:08G(:08G(:05S1a3zzz!(=X\'!!L+<5SV$7!!L+<!!IfPzzz8ArmU5k*/$8G(:08G(:0#R!Enzzz!!L+<#Z/>;!*$c7#ZSV?!"aY\\zzz8@6bE!4:,R8G(:0aRmj[!(9W`zzz8G(:089E5Z8CZ#e8G(:0!&T3+zzz#aE.*!:[f)#kYq58<gpj#[kFJzzz&-)h6&-)h6&-)h6&-)h6!!!-&zzz8<hL%!:\\A98G(:08G(:0!&QtAzzz!!L+<!!\'h8!!L(;!!L+<zzzz5ZGQ"!*$c7!-H$W!-H$W!!IfPzzzz!!!!%!!!!%!!!!%zzzz$(ueKM4ahK$(q7u$(q7u$*\\p[zzzE#TPp0N&YdE*F([GZtpc!!*-%zzzn[fbCZ18e7nbX@0EVgdZ!>$(Jzzz+92ZK+92ZK+92ZK,QJ)O+92ZKzzzpmXAJq"m5Wq"m5Wq"m5WW08t:zzz+:BS!!!=DF+:0Ft+:BRt!!",Azzz+9<kl+9DNC+:0Ft+:BS!!!48CzzzE!nc10N\'Y-?s>MkE)Sq&zzzz+:BS!+:94k+:BS!+:BS!!!<3$zzzE!nK)+:::5E!ni3E!ni3&-=6\\zzzE#LP80N\'Y,E#UtC:g6*c!!j\\Izzz!!",Az!!",A!!",AzzzzE"b>9:fAD;:fB1OE*GL.5SF8&zzz!!L+<!*$c7!*$c7!*$c7!!L+<zzz!!L+<!!\'h8!!L+<!!L+<!!!$"zzz!!L+<!!L+<!#33K!*$c7zzzz!!L+<!!\'h8!!IiQ!!L(;!!%NLzzz!!!!%!!!!%z!!!!%zzzzz!!!!%!!!!%!!!!%zzzz!<<*"!<<*"!<<*"!<'
+)
+# 请自行忽略这个雷霆大位图，0人知道为什么我要把位图直接就内联到代码里
+CHRMAP = str.maketrans("ˉˊˇˋ", "̄́̌̀")
+
+
+@cache
+def _check_syllable_valid(i: int) -> bool:
+    def _check_full(val: int) -> bool:
+        off = val - 866
+        return (0 <= off < 11105) and bool(VALID_SYLLABLES[(off >> 3)] & (1 << (off & 7)))
+
+    initial = i & 0x001F
+    final = i & 0x3F00
+    tone = i & 0x00E0
+
+    i_empty = initial in (Initial.unspec, Initial.missing)
+    f_empty = final in (Final.unspec, Final.missing)
+    t_empty = tone in (Tone.unspec, Tone.missing)
+
+    filled = (not i_empty) + (not f_empty) + (not t_empty)
+
+    if filled <= 1:
+        return True
+    elif filled == 2:
+        if i_empty:
+            for x in Initial:
+                if x not in (Initial.unspec, Initial.missing) and _check_full(x | final | tone):
+                    return True
+        elif f_empty:
+            for x in Final:
+                if x not in (Final.unspec, Final.missing) and _check_full(initial | x | tone):
+                    return True
+        elif t_empty:
+            for x in Tone:
+                if x not in (Tone.unspec, Tone.missing) and _check_full(initial | final | x):
+                    return True
+        return False
+    else:
+        return _check_full(i)
+
+
+def _check_input_valid(s: str, VRE: re.Pattern[str] | str = VALID_CHARS_RE_DEFAULT) -> bool:
+    return bool(re.fullmatch(VRE, s))
+
+
+def __parse(s: str, stack: list[Syllable], force_initial: bool = True, force_valid_syllable: bool = False) -> list[Syllable] | None:
+    # 我知道DFS还不剪枝会导致这个函数性能极差而且有爆递归风险，但是我无能优化了
+    if not s:
+        if force_valid_syllable and (not stack[-1].is_valid()):
+            return None
+        return stack
+    valid_heads = [s[:n] for n in range(min(4, len(s)), 0, -1) if s[:n] in TOKENS]
+    if not valid_heads:
+        return None
+
+    dont_try_again: set[Syllable] = set()
+
+    for head in valid_heads:
+        next_force_initial = not (
+            (head[-1] not in TOKENS) or (not any(((r & 0x001F) and (r & 0x001F) != Initial.nul) for r in TOKENS[head[-1]]))
+        )  # 当前head末位字符不能做声母，那没必要再设置声母回退了。
+        # 不回退但是也不能直接continue（那样会导致更短的head先被尝试然后抢了），只能扔到dont_try_again里防止冗余计算
+        # 虽然但是很明显这块是把原来的for in [True,False]展开了。嘛虽然更长了但至少缩进少了而且效率或许可能会高一点？
+
+        for role in (Syllable(v) for v in TOKENS[head]):
+            if stack[-1]._mreject(role, force_initial):
+                continue
+
+            current_stack = stack.copy()
+            current_stack[-1] = current_stack[-1].copy()
+            current_stack[-1]._merge(role)
+
+            for start_new_syll in [True] if role.tone else [False, True]:  # 声调后必须新开音节
+                if force_valid_syllable and start_new_syll and (not current_stack[-1].is_valid()):
+                    continue
+                if not next_force_initial:
+                    dont_try_again.add(role)
+                next_new_stack = current_stack.copy()
+                if start_new_syll:
+                    next_new_stack.append(Syllable())
+                if ret_stack := __parse(
+                    s=s[len(head) :],
+                    stack=next_new_stack,
+                    force_initial=next_force_initial and start_new_syll,
+                    force_valid_syllable=force_valid_syllable,
+                ):  # 不新开音节就不检测声母
+                    return ret_stack
+
+    for head in valid_heads:
+        for role in (Syllable(v) for v in TOKENS[head]):
+            if role in dont_try_again:
+                continue
+
+            if stack[-1]._mreject(role, force_initial):
+                continue
+
+            current_stack = stack.copy()
+            current_stack[-1] = current_stack[-1].copy()
+            current_stack[-1]._merge(role)
+
+            for start_new_syll in [True] if role.tone else [False, True]:
+                if force_valid_syllable and start_new_syll and (not current_stack[-1].is_valid()):
+                    continue
+                next_new_stack = current_stack.copy()
+                if start_new_syll:
+                    next_new_stack.append(Syllable())
+                if ret_stack := __parse(s[len(head) :], next_new_stack, False, force_valid_syllable):
+                    return ret_stack
+    return None
+
+
+_recompile = cache(re.compile)
+
+
+def parse(s: str, sep: str = "' -", default_tone_neutral=False, force_valid_syllable=False, missing_as_nul: bool = False) -> list[Syllable]:
+    s = normalize("NFKC", s).lower().translate(CHRMAP)
+    if not _check_input_valid(s, _recompile(f"[{re.escape(''.join(VALID_CHARS|set(sep)))}]*")):
+        raise ValueError("无效的输入字符")
+
+    ret = [
+        __parse(s=seg, stack=[Syllable()], force_initial=False, force_valid_syllable=force_valid_syllable)
+        for seg in re.split(f"[{re.escape(sep)}]", s)
+        if seg
+    ]
+    if not all(ret):
+        raise ValueError(f"无法解析 {s}")
+
+    ret = cast(list[list[Syllable]], ret)  # 沟槽的pylance不会用all收窄，怒哩
+
+    for r in ret:
+        if not r[-1]:
+            del r[-1]
+
+    retl = list(chain.from_iterable(ret))
+    for syl in retl:
+        if default_tone_neutral and syl.tone == Tone.missing:
+            syl.tone = Tone.t5
+        if missing_as_nul:
+            syl.initial = syl.initial or Initial.nul
+            syl.final = syl.final or Final.nul
+            syl.tone = syl.tone or Tone.nul
+
+    return retl
+
+
+@cache
+def parse_single(s: str, force_valid_syllable=False) -> Syllable:
+    s = normalize("NFKC", s).lower().translate(CHRMAP)
+    if not _check_input_valid(s):
+        raise ValueError("无效的输入字符")
+
+    ret = __parse(s=s, stack=[Syllable()], force_initial=False, force_valid_syllable=force_valid_syllable)
+
+    if ret and not ret[-1]:
+        del ret[-1]
+
+    if not ret or len(ret) != 1:
+        raise ValueError(f"无法解析 {s}")
+
+    rets = ret[0]
+
+    rets.initial = rets.initial or Initial.nul
+    rets.final = rets.final or Final.nul
+    rets.tone = rets.tone or Tone.t5
+
+    return rets
+
+
+def syllables_to_str(sylls: Iterable[Syllable], sep: str = "'") -> str:
+    ret = []
+    for prev, curr in pairwise(filter(None, sylls)):
+        if not ret:
+            ret.append(str(prev))
+        if curr.need_sep(prev):
+            ret.append(sep)
+        ret.append(str(curr))
+    return "".join(ret)
+
+
+__all__ = ["Initial", "Final", "Tone", "Syllable", "parse_single", "syllables_to_str", "parse"]