khipro-python.py

# -*- coding: utf-8 -*-
from typing import Dict, List, Tuple, Generator

# --------------------------
# Mapping groups (exactly as provided)
# --------------------------

SHOR: Dict[str, str] = {
    "o": "অ", "a": "আ", "i": "ই", "ii": "ঈ", "u": "উ", "uu": "ঊ", "q": "ঋ", "e": "এ", "oi": "ঐ", "w": "ও", "ou": "ঔ", "ae": "অ্যা", 
	"wa": "ওয়া", "wae": "ওয়্যা", "we": "ওয়ে", 
	"ooo": "অং",
		
	"oof": "ঽ",
}

FKAR: Dict[str, str] = {
    "fuf": "‌ু", "fuuf": "‌ূ", "fqf": "‌ৃ", 
	"fa": "া",
	"fi": "ি", 
	"fii": "ী", 
	"fu": "ু", 
	"fuu": "ূ", 
	"fq": "ৃ", 
	"fe": "ে", 
	"foi": "ৈ", 
	"fw": "ো", 
	"fou": "ৌ", 
	"fae": "্যা",
	"fwa": "োয়া",
	"fwe": "োয়ে",
	"oo": "ং", 
}

BYANJON: Dict[str, str] = {
    "k": "ক", "kh": "খ", "g": "গ", "gh": "ঘ", "ngf": "ঙ",
	"c": "চ", "ch": "ছ", "j": "জ", "jh": "ঝ", "nff": "ঞ", 
	"tf": "ট", "tff": "ঠ", "df": "ড", "dff": "ঢ", "nf": "ণ", 
	"t": "ত", "th": "থ", "d": "দ", "dh": "ধ", "n": "ন", 
	"p": "প", "ph": "ফ", "b": "ব", "v": "ভ", "m": "ম", 
	"z": "য", "l": "ল", "sh": "শ", "sf": "ষ", "s": "স", "h": "হ", 
	"y": "য়", "rf": "ড়", "rff": "ঢ়",
	",,": "়",
}

JUKTOBORNO: Dict[str, str] = {
    "rz": "র‍্য",
	"kk": "ক্ক", "ktf": "ক্ট", "ktfr": "ক্ট্র", "kt": "ক্ত", "ktr": "ক্ত্র", "kb": "ক্ব", "km": "ক্ম", "kz": "ক্য", "kr": "ক্র", "kl": "ক্ল",
	"kf": "ক্ষ", "ksf": "ক্ষ", "kkh": "ক্ষ", "kfnf": "ক্ষ্ণ", "kfn": "ক্ষ্ণ", "ksfnf": "ক্ষ্ণ", "ksfn": "ক্ষ্ণ", "kkhn": "ক্ষ্ণ", "kkhnf": "ক্ষ্ণ", "kfb": "ক্ষ্ব", "ksfb": "ক্ষ্ব", "kkhb": "ক্ষ্ব", "kfm": "ক্ষ্ম", 
	"kkhm": "ক্ষ্ম", "ksfm": "ক্ষ্ম", "kfz": "ক্ষ্য", "ksfz": "ক্ষ্য", "kkhz": "ক্ষ্য",
	"ks": "ক্স",
	"khz": "খ্য", "khr": "খ্র",
	"ggg": "গ্গ", "gnf": "গ্‌ণ", "gdh": "গ্ধ", "gdhz": "গ্ধ্য", "gdhr": "গ্ধ্র", "gn": "গ্ন", "gnz": "গ্ন্য", "gb": "গ্ব", "gm": "গ্ম", "gz": "গ্য", "gr": "গ্র", "grz": "গ্র্য", "gl": "গ্ল",
	"ghn": "ঘ্ন", "ghr": "ঘ্র",
	"ngk": "ঙ্ক", "ngkt": "ঙ্‌ক্ত", "ngkz": "ঙ্ক্য", "ngkr": "ঙ্ক্র", "ngkf": "ঙ্ক্ষ", "ngkkh": "ঙ্ক্ষ", "ngksf": "ঙ্ক্ষ", "ngkh": "ঙ্খ", "ngg": "ঙ্গ", "nggz": "ঙ্গ্য", "nggh": "ঙ্ঘ", "ngghz": "ঙ্ঘ্য", "ngghr": "ঙ্ঘ্র", "ngm": "ঙ্ম",
	"ngfk": "ঙ্ক", "ngfkt": "ঙ্‌ক্ত", "ngfkz": "ঙ্ক্য", "ngfkr": "ঙ্ক্র", "ngfkf": "ঙ্ক্ষ", "ngfkkh": "ঙ্ক্ষ", "ngfksf": "ঙ্ক্ষ", "ngfkh": "ঙ্খ", "ngfg": "ঙ্গ", "ngfgz": "ঙ্গ্য", "ngfgh": "ঙ্ঘ", "ngfghz": "ঙ্ঘ্য", "ngfghr": "ঙ্ঘ্র", "ngfm": "ঙ্ম",
	"cc": "চ্চ", "cch": "চ্ছ", "cchb": "চ্ছ্ব", "cchr": "চ্ছ্র", "cnff": "চ্ঞ", "cb": "চ্ব", "cz": "চ্য",
	"jj": "জ্জ", "jjb": "জ্জ্ব", "jjh": "জ্ঝ", "gg": "জ্ঞ", "jnff": "জ্ঞ", "jb": "জ্ব", "jz": "জ্য", "jr": "জ্র",
	"nc": "ঞ্চ", "nffc": "ঞ্চ", "nj": "ঞ্জ", "nffj": "ঞ্জ", "njh": "ঞ্ঝ", "nffjh": "ঞ্ঝ", "nch": "ঞ্ছ", "nffch": "ঞ্ছ",
	"ttf": "ট্ট", "tftf": "ট্ট", "tfb": "ট্ব", "tfm": "ট্ম", "tfz": "ট্য", "tfr": "ট্র",
	"ddf": "ড্ড", "dfdf": "ড্ড", "dfb": "ড্ব", "dfz": "ড্য", "dfr": "ড্র", "rfg": "ড়্‌গ",
	"dffz": "ঢ্য", "dffr": "ঢ্র",
	"nftf": "ণ্ট", "nftff": "ণ্ঠ", "nftffz": "ণ্ঠ্য", "nfdf": "ণ্ড", "nfdfz": "ণ্ড্য", "nfdfr": "ণ্ড্র", "nfdff": "ণ্ঢ", "nfnf": "ণ্ণ", "nfn": "ণ্ণ", "nfb": "ণ্ব", "nfm": "ণ্ম", "nfz": "ণ্য",
	"tt": "ত্ত", "ttb": "ত্ত্ব", "ttz": "ত্ত্য", "tth": "ত্থ", "tn": "ত্ন", "tb": "ত্ব", "tm": "ত্ম", "tmz": "ত্ম্য", "tz": "ত্য", "tr": "ত্র", "trz": "ত্র্য",
	"thb": "থ্ব", "thz": "থ্য", "thr": "থ্র",
	"dg": "দ্‌গ", "dgh": "দ্‌ঘ", "dd": "দ্দ", "ddb": "দ্দ্ব", "ddh": "দ্ধ", "db": "দ্ব", "dv": "দ্ভ", "dvr": "দ্ভ্র", "dm": "দ্ম", "dz": "দ্য", "dr": "দ্র", "drz": "দ্র্য",
	"dhn": "ধ্ন", "dhb": "ধ্ব", "dhm": "ধ্ম", "dhz": "ধ্য", "dhr": "ধ্র",
	"ntf": "ন্ট", "ntfr": "ন্ট্র", "ntff": "ন্ঠ", "ndf": "ন্ড", "ndfr": "ন্ড্র", "nt": "ন্ত", "ntb": "ন্ত্ব", "ntr": "ন্ত্র", "ntrz": "ন্ত্র্য", "nth": "ন্থ", "nthr": "ন্থ্র", "nd": "ন্দ", "ndb": "ন্দ্ব", "ndz": "ন্দ্য", 
	"ndr": "ন্দ্র", "ndh": "ন্ধ", "ndhz": "ন্ধ্য", "ndhr": "ন্ধ্র", "nn": "ন্ন", "nb": "ন্ব", "nm": "ন্ম", "nz": "ন্য", "ns": "ন্স",
	"ptf": "প্ট", "pt": "প্ত", "pn": "প্ন", "pp": "প্প", "pz": "প্য", "pr": "প্র", "pl": "প্ল", "ps": "প্স",
	"phr": "ফ্র", "phl": "ফ্ল",
	"bj": "ব্জ", "bd": "ব্দ", "bdh": "ব্ধ", "bb": "ব্ব", "bz": "ব্য", "br": "ব্র", "bl": "ব্ল", "vb": "ভ্ব", "vz": "ভ্য", "vr": "ভ্র", "vl": "ভ্ল",
	"mn": "ম্ন", "mp": "ম্প", "mpr": "ম্প্র", "mph": "ম্ফ", "mb": "ম্ব", "mbr": "ম্ব্র", "mv": "ম্ভ", "mvr": "ম্ভ্র", "mm": "ম্ম", "mz": "ম্য", "mr": "ম্র", "ml": "ম্ল",
	"zz": "য্য",
	"lk": "ল্ক", "lkz": "ল্ক্য", "lg": "ল্গ", "ltf": "ল্ট", "ldf": "ল্ড", "lp": "ল্প", "lph": "ল্ফ", "lb": "ল্ব", "lv": "ল্‌ভ", "lm": "ল্ম", "lz": "ল্য", "ll": "ল্ল",
	"shc": "শ্চ", "shch": "শ্ছ", "shn": "শ্ন", "shb": "শ্ব", "shm": "শ্ম", "shz": "শ্য", "shr": "শ্র", "shl": "শ্ল",
	"sfk": "ষ্ক", "sfkr": "ষ্ক্র", "sftf": "ষ্ট", "sftfz": "ষ্ট্য", "sftfr": "ষ্ট্র", "sftff": "ষ্ঠ", "sftffz": "ষ্ঠ্য", "sfnf": "ষ্ণ", "sfn": "ষ্ণ", "sfp": "ষ্প", "sfpr": "ষ্প্র", "sfph": "ষ্ফ", "sfb": "ষ্ব", 
	"sfm": "ষ্ম", "sfz": "ষ্য",
	"sk": "স্ক", "skr": "স্ক্র", "skh": "স্খ", "stf": "স্ট", "stfr": "স্ট্র", "st": "স্ত", "stb": "স্ত্ব", "stz": "স্ত্য", "str": "স্ত্র", "sth": "স্থ", "sthz": "স্থ্য", "sn": "স্ন", "sp": "স্প", "spr": "স্প্র", 
	"spl": "স্প্ল", "sph": "স্ফ", "sb": "স্ব", "sm": "স্ম", "sz": "স্য", "sr": "স্র", "sl": "স্ল",
	"hn": "হ্ন", "hnf": "হ্ণ", "hb": "হ্ব", "hm": "হ্ম", "hz": "হ্য", "hr": "হ্র", "hl": "হ্ল",

    # oshomvob juktoborno
    "ksh": "কশ", "kks": "কক্স",  "nsh": "নশ", "psh": "পশ", "ld": "লদ", "gd": "গদ", "ngkk": "ঙ্কক", "ngks": "ঙ্কস", "ngfkk": "ঙ্কক", "ngfks": "ঙ্কস",
		
	"cn": "চন", 
		"cngf": "চঙ", "cnz": "চন্য", "cnf": "চণ", "cnm": "চন্ম", 
		"cngk": "চঙ্ক", "cngkt": "চঙ্‌ক্ত", "cngkz": "চঙ্ক্য", "cngkr": "চঙ্ক্র", "cngkf": "চঙ্ক্ষ", "cngkkh": "চঙ্ক্ষ", "cngksf": "চঙ্ক্ষ", "cngkh": "চঙ্খ", "cngg": "চঙ্গ", "cnggz": "চঙ্গ্য", "cnggh": "চঙ্ঘ", "cngghz": "চঙ্ঘ্য", "cngghr": "চঙ্ঘ্র", "cngm": "চঙ্ম",
		"cngfk": "চঙ্ক", "cngfkt": "চঙ্‌ক্ত", "cngfkz": "চঙ্ক্য", "cngfkr": "চঙ্ক্র", "cngfkf": "চঙ্ক্ষ", "cngfkkh": "চঙ্ক্ষ", "cngfksf": "চঙ্ক্ষ", "cngfkh": "চঙ্খ", "cngfg": "চঙ্গ", "cngfgz": "চঙ্গ্য", "cngfgh": "চঙ্ঘ", "cngfghz": "চঙ্ঘ্য", "cngfghr": "চঙ্ঘ্র", "cngfm": "চঙ্ম",
		"cnc": "চঞ্চ", "cnffc": "চঞ্চ", "cnj": "চঞ্জ", "cnffj": "চঞ্জ", "cnjh": "চঞ্ঝ", "cnffjh": "চঞ্ঝ", "cnch": "চঞ্ছ", "cnffch": "চঞ্ছ",
		"cnftf": "চণ্ট", "cnftff": "চণ্ঠ", "cnftffz": "চণ্ঠ্য", "cnfdf": "চণ্ড", "cnfdfz": "চণ্ড্য", "cnfdfr": "চণ্ড্র", "cnfdff": "চণ্ঢ", "cnfnf": "চণ্ণ", "cnfn": "চণ্ণ", "cnfb": "চণ্ব", "cnfm": "চণ্ম", "cnfz": "চণ্য",
		"cntf": "চন্ট", "cntfr": "চন্ট্র", "cntff": "চন্ঠ", "cndf": "চন্ড", "cndfr": "চন্ড্র", "cnt": "চন্ত", "cntb": "চন্ত্ব", "cntr": "চন্ত্র", "cntrz": "চন্ত্র্য", "cnth": "চন্থ", "cnthr": "চন্থ্র", "cnd": "চন্দ", "cndb": "চন্দ্ব", "cndz": "চন্দ্য", 
		"cndr": "চন্দ্র", "cndh": "চন্ধ", "cndhz": "চন্ধ্য", "cndhr": "চন্ধ্র", "cnn": "চন্ন", "cnb": "চন্ব", "cnm": "চন্ম", "cnz": "চন্য", "cns": "চন্স",
		"cnsh": "চনশ",
		"cngkk": "চঙ্কক", "cngks": "চঙ্কস", "cngfkk": "চঙ্কক", "cngfks": "চঙ্কস",
		"cnft": "চণত", "cnfd": "চণদ", "cnfth": "চণথ", "cnfdh": "চণধ",
		"cndff": "চনঢ", 
		"cngkth": "চঙ্কথ", "cngkrf": "চঙ্কড়", "cngkrff": "চঙ্কঢ়", "cngghrf": "চঙ্ঘড়", "cngghrff": "চঙ্ঘঢ়",
																					"cngfkth": "চঙ্কথ", "cngfkrf": "চঙ্কড়", "cngfkrff": "চঙ্কঢ়", "cngfghrf": "চঙ্ঘড়", "cngfghrff": "চঙ্ঘঢ়",
		"cnfdfrf": "চণ্ডড়", "cnfdfrff": "চণ্ডঢ়",
		"cntfrf": "চন্টড়", "cntfrff": "চন্টঢ়", "cndfrf": "চন্ডড়", "cndfrff": "চন্ডঢ়", "cntrf": "চন্তড়", "cntrff": "চন্তঢ়", "cnthrf": "চন্থড়", 
		"cnstf": "চনস্ট", "cnst": "চনস্ত", "cnsk": "চনস্ক", 
		"cnthrff": "চন্থঢ়", "cndrf": "চন্দড়", "cndrff": "চন্দঢ়", "cndhrf": "চন্ধড়", "cndhrff": "চন্ধঢ়",
		"cngksh": "চঙ্কশ", "cngfksh": "চঙ্কশ",


	"jn": "জন", 
		"jngf": "জঙ", "jnf": "জণ", "jnz": "জন্য", "jnm": "জন্ম",
		"jngk": "জঙ্ক", "jngkt": "জঙ্‌ক্ত", "jngkz": "জঙ্ক্য", "jngkr": "জঙ্ক্র", "jngkf": "জঙ্ক্ষ", "jngkkh": "জঙ্ক্ষ", "jngksf": "জঙ্ক্ষ", "jngkh": "জঙ্খ", "jngg": "জঙ্গ", "jnggz": "জঙ্গ্য", "jnggh": "জঙ্ঘ", "jngghz": "জঙ্ঘ্য", "jngghr": "জঙ্ঘ্র", "jngm": "জঙ্ম",
		"jngfk": "জঙ্ক", "jngfkt": "জঙ্‌ক্ত", "jngfkz": "জঙ্ক্য", "jngfkr": "জঙ্ক্র", "jngfkf": "জঙ্ক্ষ", "jngfkkh": "জঙ্ক্ষ", "jngfksf": "জঙ্ক্ষ", "jngfkh": "জঙ্খ", "jngfg": "জঙ্গ", "jngfgz": "জঙ্গ্য", "jngfgh": "জঙ্ঘ", "jngfghz": "জঙ্ঘ্য", "jngfghr": "জঙ্ঘ্র", "jngfm": "জঙ্ম",
		"jnc": "জঞ্চ", "jnffc": "জঞ্চ", "jnj": "জঞ্জ", "jnffj": "জঞ্জ", "jnjh": "জঞ্ঝ", "jnffjh": "জঞ্ঝ", "jnch": "জঞ্ছ", "jnffch": "জঞ্ছ",
		"jnftf": "জণ্ট", "jnftff": "জণ্ঠ", "jnftffz": "জণ্ঠ্য", "jnfdf": "জণ্ড", "jnfdfz": "জণ্ড্য", "jnfdfr": "জণ্ড্র", "jnfdff": "জণ্ঢ", "jnfnf": "জণ্ণ", "jnfn": "জণ্ণ", "jnfb": "জণ্ব", "jnfm": "জণ্ম", "jnfz": "জণ্য",
		"jntf": "জন্ট", "jntfr": "জন্ট্র", "jntff": "জন্ঠ", "jndf": "জন্ড", "jndfr": "জন্ড্র", "jnt": "জন্ত", "jntb": "জন্ত্ব", "jntr": "জন্ত্র", "jntrz": "জন্ত্র্য", "jnth": "জন্থ", "jnthr": "জন্থ্র", "jnd": "জন্দ", "jndb": "জন্দ্ব", "jndz": "জন্দ্য", 
		"jndr": "জন্দ্র", "jndh": "জন্ধ", "jndhz": "জন্ধ্য", "jndhr": "জন্ধ্র", "jnn": "জন্ন", "jnb": "জন্ব", "jnm": "জন্ম", "jnz": "জন্য", "jns": "জন্স",
		"jnsh": "জনশ",
		"jngkk": "জঙ্কক", "jngks": "জঙ্কস", "jngfkk": "জঙ্কক", "jngfks": "জঙ্কস",
		"jnft": "জণত", "jnfd": "জণদ", "jnfth": "জণথ", "jnfdh": "জণধ",
		"jndff": "জনঢ", 
		"jngkth": "জঙ্কথ", "jngkrf": "জঙ্কড়", "jngkrff": "জঙ্কঢ়", "jngghrf": "জঙ্ঘড়", "jngghrff": "জঙ্ঘঢ়",
																					"jngfkth": "জঙ্কথ", "jngfkrf": "জঙ্কড়", "jngfkrff": "জঙ্কঢ়", "jngfghrf": "জঙ্ঘড়", "jngfghrff": "জঙ্ঘঢ়",
		"jnfdfrf": "জণ্ডড়", "jnfdfrff": "জণ্ডঢ়",
		"jntfrf": "জন্টড়", "jntfrff": "জন্টঢ়", "jndfrf": "জন্ডড়", "jndfrff": "জন্ডঢ়", "jntrf": "জন্তড়", "jntrff": "জন্তঢ়", "jnthrf": "জন্থড়", 
		"jnstf": "জনস্ট", "jnst": "জনস্ত", "jnsk": "জনস্ক", 
		"jnthrff": "জন্থঢ়", "jndrf": "জন্দড়", "jndrff": "জন্দঢ়", "jndhrf": "জন্ধড়", "jndhrff": "জন্ধঢ়",
		"jngksh": "জঙ্কশ", "jngfksh": "জঙ্কশ",


	"tft": "টত", "dfd": "ডদ", 
		
	"nft": "ণত", "nfd": "ণদ", 
	"lt": "লত", "sft": "ষত", 
	"nfth": "ণথ", "nfdh": "ণধ", 
	"sfth": "ষথ",
	"ktff": "কঠ", "ptff": "পঠ", "ltff": "লঠ", "stff": "সঠ", 
	"dfdff": "ডঢ", 
	"ndff": "নঢ", 
	"ktfrf": "ক্টড়", "ktfrff": "ক্টঢ়", "kth": "কথ", "ktrf": "ক্তড়", "ktrff": "ক্তঢ়", "krf": "কড়", "krff": "কঢ়", "khrf": "খড়", "khrff": "খঢ়", "gggh": "জ্ঞঘ", "gdff": "গঢ", "gdhrf": "গ্ধড়", 
	"gdhrff": "গ্ধঢ়", "grf": "গড়", "grff": "গঢ়", "ghrf": "ঘড়", "ghrff": "ঘঢ়", 
	"ngkth": "ঙ্কথ", "ngkrf": "ঙ্কড়", "ngkrff": "ঙ্কঢ়", "ngghrf": "ঙ্ঘড়", "ngghrff": "ঙ্ঘঢ়", 
	"cchrf": "চ্ছড়", "cchrff": "চ্ছঢ়", 
	"ngfkth": "ঙ্কথ", "ngfkrf": "ঙ্কড়", "ngfkrff": "ঙ্কঢ়", "ngfghrf": "ঙ্ঘড়", "ngfghrff": "ঙ্ঘঢ়",
	"tfrf": "টড়", "tfrff": "টঢ়", "dfrf": "ডড়", "dfrff": "ডঢ়", "rfgh": "ড়ঘ", "dffrf": "ঢড়", "dffrff": "ঢঢ়", 
	"nfdfrf": "ণ্ডড়", "nfdfrff": "ণ্ডঢ়", 
	"trf": "তড়", "trff": "তঢ়", "thrf": "থড়", "thrff": "থঢ়", 
	"dvrf": "দ্ভড়", "dvrff": "দ্ভঢ়", "drf": "দড়", "drff": "দঢ়", "dhrf": "ধড়", "dhrff": "ধঢ়", 
	"ntfrf": "ন্টড়", "ntfrff": "ন্টঢ়", "ndfrf": "ন্ডড়", "ndfrff": "ন্ডঢ়", "ntrf": "ন্তড়", "ntrff": "ন্তঢ়", "nthrf": "ন্থড়", 
	"nstf": "নস্ট", "nst": "নস্ত", "nsk": "নস্ক", 
	"nthrff": "ন্থঢ়", "ndrf": "ন্দড়", "ndrff": "ন্দঢ়", "ndhrf": "ন্ধড়", "ndhrff": "ন্ধঢ়", 
	"pth": "পথ", "pph": "পফ", "prf": "পড়", "prff": "পঢ়", "phrf": "ফড়", "phrff": "ফঢ়", "bjh": "বঝ", "brf": "বড়", "brff": "বঢ়", 
	"mpl": "মপ্ল", 
	"vrf": "ভড়", "vrff": "ভঢ়", "mprf": "ম্পড়", "mprff": "ম্পঢ়", "mbrf": "ম্বড়", "mbrff": "ম্বঢ়", "mvrf": "ম্ভড়", "mvrff": "ম্ভঢ়", "mrf": "মড়", "mrff": "মঢ়", "lkh": "লখ", "lgh": "লঘ", "shrf": "শড়", "shrff": "শঢ়", "sfkh": "ষখ",
	"sfkrf": "ষ্কড়", "sfkrff": "ষ্কঢ়", "sftfrf": "ষ্টড়", "sftfrff": "ষ্টঢ়", "sfprf": "ষ্পড়", "sfprff": "ষ্পঢ়", "skrf": "স্কড়", "skrff": "স্কঢ়", "stfrf": "স্টড়", "stfrff": "স্টঢ়", "strf": "স্তড়", "strff": "স্তঢ়", "sprf": "স্পড়", "sprff": "স্পঢ়", 
	"srf": "সড়", "srff": "সঢ়", "hrf": "হড়", "hrff": "হঢ়", "ldh": "লধ", 
	"ngksh": "ঙ্কশ", "tfth": "টথ", "dfdh": "ডধ", "lth": "লথ",
	"ngfksh": "ঙ্কশ",
	"lks": "ল্কস", 
	"kkf": "কক্ষ", "lkf": "লক্ষ", "sfkf": "ষক্ষ", "skf": "সক্ষ", "kkkh": "কক্ষ", "lkkh": "লক্ষ", "sfkkh": "ষক্ষ", "skkh": "সক্ষ", "kksf": "কক্ষ", "lksf": "লক্ষ", "sfksf": "ষক্ষ", "sksf": "সক্ষ",
	"yr": "য়র",
    "gnj": "গঞ্জ", "pnj": "পঞ্জ", "mnj": "মঞ্জ", "snj": "সঞ্জ",
    "gndf": "গন্ড", "mndf": "মন্ড",
    "tnt": "তন্ত", "tntr": "তন্ত্র", "mnt": "মন্ত", "mntr": "মন্ত্র", "snt": "সন্ত", "sntr": "সন্ত্র", "hnt": "হন্ত",
    "tnd": "তন্দ", "nnd": "নন্দ", "mnd": "মন্দ", "snd": "সন্দ",
    "gndh": "গন্ধ", "gndhz": "গন্ধ্য", "sndh": "সন্ধ", "sndhz": "সন্ধ্য",
}

NG: Dict[str, str] = {
    "ng": "ং", "gng": "গং", "ghng": "ঘং", "cng": "চং", "jng": "জং", "nng": "নং", "nfng": "ণং", "tng": "তং", "dhng": "ধং", "png": "পং", "mng": "মং", "shng": "শং", "sfng": "ষং", "sng": "সং", "hng": "হং", "kfng": "ক্ষং", "kkhng": "ক্ষং", "ksfng": "ক্ষং",
	"ngo": "ঙ", "nga": "ঙা", "ngi": "ঙি", "ngii": "ঙী", "ngu": "ঙু", "nguff": "ঙ‌ু", "nguu": "ঙূ", "nguuff": "ঙ‌ূ", "ngq": "ঙৃ", "ngqff": "ঙ‌ৃ", "nge": "ঙে", "ngoi": "ঙৈ",     
	"ngw": "ঙো", "ngou": "ঙৌ", "ngae": "ঙ্যা",
	"ngof": "ঙঅ", "ngaf": "ঙআ", "ngif": "ঙই", "ngiif": "ঙঈ", "nguf": "ঙউ", "nguuf": "ঙঊ", "ngqf": "ঙঋ", "ngef": "ঙএ", "ngoif": "ঙই",     
	"ngwf": "ঙও", "ngouf": "ঙউ", "ngaef": "ঙঅ্যা",

	"gngo": "গঙ", "gnga": "গঙা", "gngi": "গঙি", "gngii": "গঙী", "gngu": "গঙু", "gnguff": "গঙ‌ু", "gnguu": "গঙূ", "gnguuff": "গঙ‌ূ", "gngq": "গঙৃ", "gngqff": "গঙ‌ৃ", "gnge": "গঙে", "gngoi": "গঙৈ",     
	"gngw": "গঙো", "gngou": "গঙৌ", "gngae": "গঙ্যা",
	"gngof": "গঙঅ", "gngaf": "গঙআ", "gngif": "গঙই", "gngiif": "গঙঈ", "gnguf": "গঙউ", "gnguuf": "গঙঊ", "gngqf": "গঙঋ", "gngef": "গঙএ", "gngoif": "গঙই",     
	"gngwf": "গঙও", "gngouf": "গঙউ", "gngaef": "গঙঅ্যা",

	"ghngo": "ঘঙ", "ghnga": "ঘঙা", "ghngi": "ঘঙি", "ghngii": "ঘঙী", "ghngu": "ঘঙু", "ghnguff": "ঘঙ‌ু", "ghnguu": "ঘঙূ", "ghnguuff": "ঘঙ‌ূ", "ghngq": "ঘঙৃ", "ghngqff": "ঘঙ‌ৃ", "ghnge": "ঘঙে", "ghngoi": "ঘঙৈ",     
	"ghngw": "ঘঙো", "ghngou": "ঘঙৌ", "ghngae": "ঘঙ্যা",
	"ghngof": "ঘঙঅ", "ghngaf": "ঘঙআ", "ghngif": "ঘঙই", "ghngiif": "ঘঙঈ", "ghnguf": "ঘঙউ", "ghnguuf": "ঘঙঊ", "ghngqf": "ঘঙঋ", "ghngef": "ঘঙএ", "ghngoif": "ঘঙই",     
	"ghngwf": "ঘঙও", "ghngouf": "ঘঙউ", "ghngaef": "ঘঙঅ্যা",

	"cngo": "চঙ", "cnga": "চঙা", "cngi": "চঙি", "cngii": "চঙী", "cngu": "চঙু", "cnguff": "চঙ‌ু", "cnguu": "চঙূ", "cnguuff": "চঙ‌ূ", "cngq": "চঙৃ", "cngqff": "চঙ‌ৃ", "cnge": "চঙে", "cngoi": "চঙৈ",     
	"cngw": "চঙো", "cngou": "চঙৌ", "cngae": "চঙ্যা",
	"cngof": "চঙঅ", "cngaf": "চঙআ", "cngif": "চঙই", "cngiif": "চঙঈ", "cnguf": "চঙউ", "cnguuf": "চঙঊ", "cngqf": "চঙঋ", "cngef": "চঙএ", "cngoif": "চঙই",     
	"cngwf": "চঙও", "cngouf": "চঙউ", "cngaef": "চঙঅ্যা",

	"jngo": "জঙ", "jnga": "জঙা", "jngi": "জঙি", "jngii": "জঙী", "jngu": "জঙু", "jnguff": "জঙ‌ু", "jnguu": "জঙূ", "jnguuff": "জঙ‌ূ", "jngq": "জঙৃ", "jngqff": "জঙ‌ৃ", "jnge": "জঙে", "jngoi": "জঙৈ",     
	"jngw": "জঙো", "jngou": "জঙৌ", "jngae": "জঙ্যা",
	"jngof": "জঙঅ", "jngaf": "জঙআ", "jngif": "জঙই", "jngiif": "জঙঈ", "jnguf": "জঙউ", "jnguuf": "জঙঊ", "jngqf": "জঙঋ", "jngef": "জঙএ", "jngoif": "জঙই",     
	"jngwf": "জঙও", "jngouf": "জঙউ", "jngaef": "জঙঅ্যা",

	"nngo": "নঙ", "nnga": "নঙা", "nngi": "নঙি", "nngii": "নঙী", "nngu": "নঙু", "nnguff": "নঙ‌ু", "nnguu": "নঙূ", "nnguuff": "নঙ‌ূ", "nngq": "নঙৃ", "nngqff": "নঙ‌ৃ", "nnge": "নঙে", "nngoi": "নঙৈ",     
	"nngw": "নঙো", "nngou": "নঙৌ", "nngae": "নঙ্যা",
	"nngof": "নঙঅ", "nngaf": "নঙআ", "nngif": "নঙই", "nngiif": "নঙঈ", "nnguf": "নঙউ", "nnguuf": "নঙঊ", "nngqf": "নঙঋ", "nngef": "নঙএ", "nngoif": "নঙই",     
	"nngwf": "নঙও", "nngouf": "নঙউ", "nngaef": "নঙঅ্যা",

	"nfngo": "ণঙ", "nfnga": "ণঙা", "nfngi": "ণঙি", "nfngii": "ণঙী", "nfngu": "ণঙু", "nfnguff": "ণঙ‌ু", "nfnguu": "ণঙূ", "nfnguuff": "ণঙ‌ূ", "nfngq": "ণঙৃ", "nfngqff": "ণঙ‌ৃ", "nfnge": "ণঙে", "nfngoi": "ণঙৈ",     
	"nfngw": "ণঙো", "nfngou": "ণঙৌ", "nfngae": "ণঙ্যা",
	"nfngof": "ণঙঅ", "nfngaf": "ণঙআ", "nfngif": "ণঙই", "nfngiif": "ণঙঈ", "nfnguf": "ণঙউ", "nfnguuf": "ণঙঊ", "nfngqf": "ণঙঋ", "nfngef": "ণঙএ", "nfngoif": "ণঙই",     
	"nfngwf": "ণঙও", "nfngouf": "ণঙউ", "nfngaef": "ণঙঅ্যা",

	"tngo": "তঙ", "tnga": "তঙা", "tngi": "তঙি", "tngii": "তঙী", "tngu": "তঙু", "tnguff": "তঙ‌ু", "tnguu": "তঙূ", "tnguuff": "তঙ‌ূ", "tngq": "তঙৃ", "tngqff": "তঙ‌ৃ", "tnge": "তঙে", "tngoi": "তঙৈ",     
	"tngw": "তঙো", "tngou": "তঙৌ", "tngae": "তঙ্যা",
	"tngof": "তঙঅ", "tngaf": "তঙআ", "tngif": "তঙই", "tngiif": "তঙঈ", "tnguf": "তঙউ", "tnguuf": "তঙঊ", "tngqf": "তঙঋ", "tngef": "তঙএ", "tngoif": "তঙই",     
	"tngwf": "তঙও", "tngouf": "তঙউ", "tngaef": "তঙঅ্যা",

	"dhngo": "ধঙ", "dhnga": "ধঙা", "dhngi": "ধঙি", "dhngii": "ধঙী", "dhngu": "ধঙু", "dhnguff": "ধঙ‌ু", "dhnguu": "ধঙূ", "dhnguuff": "ধঙ‌ূ", "dhngq": "ধঙৃ", "dhngqff": "ধঙ‌ৃ", "dhnge": "ধঙে", "dhngoi": "ধঙৈ",     
	"dhngw": "ধঙো", "dhngou": "ধঙৌ", "dhngae": "ধঙ্যা",
	"dhngof": "ধঙঅ", "dhngaf": "ধঙআ", "dhngif": "ধঙই", "dhngiif": "ধঙঈ", "dhnguf": "ধঙউ", "dhnguuf": "ধঙঊ", "dhngqf": "ধঙঋ", "dhngef": "ধঙএ", "dhngoif": "ধঙই",     
	"dhngwf": "ধঙও", "dhngouf": "ধঙউ", "dhngaef": "ধঙঅ্যা",

	"pngo": "পঙ", "pnga": "পঙা", "pngi": "পঙি", "pngii": "পঙী", "pngu": "পঙু", "pnguff": "পঙ‌ু", "pnguu": "পঙূ", "pnguuff": "পঙ‌ূ", "pngq": "পঙৃ", "pngqff": "পঙ‌ৃ", "pnge": "পঙে", "pngoi": "পঙৈ",     
	"pngw": "পঙো", "pngou": "পঙৌ", "pngae": "পঙ্যা",
	"pngof": "পঙঅ", "pngaf": "পঙআ", "pngif": "পঙই", "pngiif": "পঙঈ", "pnguf": "পঙউ", "pnguuf": "পঙঊ", "pngqf": "পঙঋ", "pngef": "পঙএ", "pngoif": "পঙই",     
	"pngwf": "পঙও", "pngouf": "পঙউ", "pngaef": "পঙঅ্যা",

	"mngo": "মঙ", "mnga": "মঙা", "mngi": "মঙি", "mngii": "মঙী", "mngu": "মঙু", "mnguff": "মঙ‌ু", "mnguu": "মঙূ", "mnguuff": "মঙ‌ূ", "mngq": "মঙৃ", "mngqff": "মঙ‌ৃ", "mnge": "মঙে", "mngoi": "মঙৈ",     
	"mngw": "মঙো", "mngou": "মঙৌ", "mngae": "মঙ্যা",
	"mngof": "মঙঅ", "mngaf": "মঙআ", "mngif": "মঙই", "mngiif": "মঙঈ", "mnguf": "মঙউ", "mnguuf": "মঙঊ", "mngqf": "মঙঋ", "mngef": "মঙএ", "mngoif": "মঙই",     
	"mngwf": "মঙও", "mngouf": "মঙউ", "mngaef": "মঙঅ্যা",

	"shngo": "শঙ", "shnga": "শঙা", "shngi": "শঙি", "shngii": "শঙী", "shngu": "শঙু", "shnguff": "শঙ‌ু", "shnguu": "শঙূ", "shnguuff": "শঙ‌ূ", "shngq": "শঙৃ", "shngqff": "শঙ‌ৃ", "shnge": "শঙে", "shngoi": "শঙৈ",     
	"shngw": "শঙো", "shngou": "শঙৌ", "shngae": "শঙ্যা",
	"shngof": "শঙঅ", "shngaf": "শঙআ", "shngif": "শঙই", "shngiif": "শঙঈ", "shnguf": "শঙউ", "shnguuf": "শঙঊ", "shngqf": "শঙঋ", "shngef": "শঙএ", "shngoif": "শঙই",     
	"shngwf": "শঙও", "shngouf": "শঙউ", "shngaef": "শঙঅ্যা",

	"sfngo": "ষঙ", "sfnga": "ষঙা", "sfngi": "ষঙি", "sfngii": "ষঙী", "sfngu": "ষঙু", "sfnguff": "ষঙ‌ু", "sfnguu": "ষঙূ", "sfnguuff": "ষঙ‌ূ", "sfngq": "ষঙৃ", "sfngqff": "ষঙ‌ৃ", "sfnge": "ষঙে", "sfngoi": "ষঙৈ",     
	"sfngw": "ষঙো", "sfngou": "ষঙৌ", "sfngae": "ষঙ্যা",
	"sfngof": "ষঙঅ", "sfngaf": "ষঙআ", "sfngif": "ষঙই", "sfngiif": "ষঙঈ", "sfnguf": "ষঙউ", "sfnguuf": "ষঙঊ", "sfngqf": "ষঙঋ", "sfngef": "ষঙএ", "sfngoif": "ষঙই",     
	"sfngwf": "ষঙও", "sfngouf": "ষঙউ", "sfngaef": "ষঙঅ্যা",

	"sngo": "সঙ", "snga": "সঙা", "sngi": "সঙি", "sngii": "সঙী", "sngu": "সঙু", "snguff": "সঙ‌ু", "snguu": "সঙূ", "snguuff": "সঙ‌ূ", "sngq": "সঙৃ", "sngqff": "সঙ‌ৃ", "snge": "সঙে", "sngoi": "সঙৈ",     
	"sngw": "সঙো", "sngou": "সঙৌ", "sngae": "সঙ্যা",
	"sngof": "সঙঅ", "sngaf": "সঙআ", "sngif": "সঙই", "sngiif": "সঙঈ", "snguf": "সঙউ", "snguuf": "সঙঊ", "sngqf": "সঙঋ", "sngef": "সঙএ", "sngoif": "সঙই",     
	"sngwf": "সঙও", "sngouf": "সঙউ", "sngaef": "সঙঅ্যা",

	"hngo": "হঙ", "hnga": "হঙা", "hngi": "হঙি", "hngii": "হঙী", "hngu": "হঙু", "hnguff": "হঙ‌ু", "hnguu": "হঙূ", "hnguuff": "হঙ‌ূ", "hngq": "হঙৃ", "hngqff": "হঙ‌ৃ", "hnge": "হঙে", "hngoi": "হঙৈ",     
	"hngw": "হঙো", "hngou": "হঙৌ", "hngae": "হঙ্যা",
	"hngof": "হঙঅ", "hngaf": "হঙআ", "hngif": "হঙই", "hngiif": "হঙঈ", "hnguf": "হঙউ", "hnguuf": "হঙঊ", "hngqf": "হঙঋ", "hngef": "হঙএ", "hngoif": "হঙই",     
	"hngwf": "হঙও", "hngouf": "হঙউ", "hngaef": "হঙঅ্যা",

	"kfngo": "ক্ষঙ", "kfnga": "ক্ষঙা", "kfngi": "ক্ষঙি", "kfngii": "ক্ষঙী", "kfngu": "ক্ষঙু", "kfnguff": "ক্ষঙ‌ু", "kfnguu": "ক্ষঙূ", "kfnguuff": "ক্ষঙ‌ূ", "kfngq": "ক্ষঙৃ", "kfngqff": "ক্ষঙ‌ৃ", "kfnge": "ক্ষঙে", "kfngoi": "ক্ষঙৈ",     
	"kfngw": "ক্ষঙো", "kfngou": "ক্ষঙৌ", "kfngae": "ক্ষঙ্যা",
	"kfngof": "ক্ষঙঅ", "kfngaf": "ক্ষঙআ", "kfngif": "ক্ষঙই", "kfngiif": "ক্ষঙঈ", "kfnguf": "ক্ষঙউ", "kfnguuf": "ক্ষঙঊ", "kfngqf": "ক্ষঙঋ", "kfngef": "ক্ষঙএ", "kfngoif": "ক্ষঙই",     
	"kfngwf": "ক্ষঙও", "kfngouf": "ক্ষঙউ", "kfngaef": "ক্ষঙঅ্যা",

	"kkhngo": "ক্ষঙ", "kkhnga": "ক্ষঙা", "kkhngi": "ক্ষঙি", "kkhngii": "ক্ষঙী", "kkhngu": "ক্ষঙু", "kkhnguff": "ক্ষঙ‌ু", "kkhnguu": "ক্ষঙূ", "kkhnguuff": "ক্ষঙ‌ূ", "kkhngq": "ক্ষঙৃ", "kkhngqff": "ক্ষঙ‌ৃ", "kkhnge": "ক্ষঙে", "kkhngoi": "ক্ষঙৈ",     
	"kkhngw": "ক্ষঙো", "kkhngou": "ক্ষঙৌ", "kkhngae": "ক্ষঙ্যা",
	"kkhngof": "ক্ষঙঅ", "kkhngaf": "ক্ষঙআ", "kkhngif": "ক্ষঙই", "kkhngiif": "ক্ষঙঈ", "kkhnguf": "ক্ষঙউ", "kkhnguuf": "ক্ষঙঊ", "kkhngqf": "ক্ষঙঋ", "kkhngef": "ক্ষঙএ", "kkhngoif": "ক্ষঙই",     
	"kkhngwf": "ক্ষঙও", "kkhngouf": "ক্ষঙউ", "kkhngaef": "ক্ষঙঅ্যা",

	"ksfngo": "ক্ষঙ", "ksfnga": "ক্ষঙা", "ksfngi": "ক্ষঙি", "ksfngii": "ক্ষঙী", "ksfngu": "ক্ষঙু", "ksfnguff": "ক্ষঙ‌ু", "ksfnguu": "ক্ষঙূ", "ksfnguuff": "ক্ষঙ‌ূ", "ksfngq": "ক্ষঙৃ", "ksfngqff": "ক্ষঙ‌ৃ", "ksfnge": "ক্ষঙে", "ksfngoi": "ক্ষঙৈ",     
	"ksfngw": "ক্ষঙো", "ksfngou": "ক্ষঙৌ", "ksfngae": "ক্ষঙ্যা",
	"ksfngof": "ক্ষঙঅ", "ksfngaf": "ক্ষঙআ", "ksfngif": "ক্ষঙই", "ksfngiif": "ক্ষঙঈ", "ksfnguf": "ক্ষঙউ", "ksfnguuf": "ক্ষঙঊ", "ksfngqf": "ক্ষঙঋ", "ksfngef": "ক্ষঙএ", "ksfngoif": "ক্ষঙই",     
	"ksfngwf": "ক্ষঙও", "ksfngouf": "ক্ষঙউ", "ksfngaef": "ক্ষঙঅ্যা",
}

REPH: Dict[str, str] = {
    "rr": "র্", "r": "র",
}

PHOLA: Dict[str, str] = {
    "r": "র", "z": "য",
}

KAR: Dict[str, str] = {
    "o": "", "of": "অ",
	"oof": "ঽ",
	"a": "া", "af": "আ",
	"i": "ি", "if": "ই",
	"ii": "ী", "iif": "ঈ",
	"u": "ু", "uf": "উ",
	"uu": "ূ", "uuf": "ঊ",
	"q": "ৃ", "qf": "ঋ",
	"e": "ে", "ef": "এ",
	"oi": "ৈ", "oif": "ই",
	"w": "ো", "wf": "ও",     
	"ou": "ৌ", "ouf": "উ",
	"ae": "্যা", "aef": "অ্যা",
	"uff": "‌ু", "uuff": "‌ূ", "qff": "‌ৃ", 

	"we": "োয়ে", "wef": "ওয়ে",
	"waf": "ওয়া", "wa": "োয়া",
	"wae": "ওয়্যা",

	"oo": "ং",
}

ONGKO: Dict[str, str] = {
    ".1": ".১", ".2": ".২", ".3": ".৩", ".4": ".৪", ".5": ".৫", ".6": ".৬", ".7": ".৭", ".8": ".৮", ".9": ".৯", ".0": ".০",
    "1": "১", "2": "২", "3": "৩", "4": "৪", "5": "৫", "6": "৬", "7": "৭", "8": "৮", "9": "৯", "0": "০",
	"A": "a", "B": "b", "C": "c", "D": "d", "E": "e", "F": "f", "G": "g", "H": "h", "I": "i", "J": "j", "K": "k", "L": "l", "M": "m",
    "N": "n", "O": "o", "P": "p", "Q": "q", "R": "r", "S": "s", "T": "t", "U": "u", "V": "v", "W": "w", "X": "x", "Y": "y", "Z": "z",
}

DIACRITIC: Dict[str, str] = {
    "qq": "্", "xx": "্‌", "t/": "ৎ", "x": "ঃ", "/": "ঁ", "//": "/", 
	"`": "`", "``": "‌", "```": "``",
	"~": "~", "~~": "‍", "~~~": "~~",
}

BIRAM: Dict[str, str] = {
    ".": "।", "...": "...", "..": ".", "$": "৳", "$f": "₹", ",,,": ",,", ".f": "॥", ".ff": "৺", 
	"+": "+", "-": "-", "=": "=", "+f": "×", "-f": "÷", "=f": "≠", "$$": "$",
}

PRITHAYOK: Dict[str, str] = {
    ";": "", ";;": ";",
}

# The following AE has an AE kar with a preceding joiner
AE: Dict[str, str] = {
    "ae": "‍্যা",
}

# --------------------------
# State machine configuration
# --------------------------

INIT = "init"
SHOR_STATE = "shor-state"
REPH_STATE = "reph-state"
BYANJON_STATE = "byanjon-state"

GROUP_MAPS: Dict[str, Dict[str, str]] = {
    "shor": SHOR,
    "fkar": FKAR,
    "byanjon": BYANJON,
    "juktoborno": JUKTOBORNO,
    "ng": NG,
    "reph": REPH,
    "phola": PHOLA,
    "kar": KAR,
    "ongko": ONGKO,
    "diacritic": DIACRITIC,
    "biram": BIRAM,
    "prithayok": PRITHAYOK,
    "ae": AE,
}

# Group order per state (priority used when same-length matches)
STATE_GROUP_ORDER: Dict[str, List[str]] = {
    INIT: ["diacritic", "ng", "shor", "fkar", "prithayok", "ongko", "biram", "reph", "juktoborno", "byanjon"],
    SHOR_STATE: ["diacritic", "ng", "shor", "fkar", "biram", "prithayok", "ongko", "biram", "reph", "juktoborno", "byanjon"],
    REPH_STATE: ["prithayok", "diacritic", "ng", "ae", "juktoborno", "byanjon", "kar"],
    BYANJON_STATE: ["diacritic", "ng", "prithayok", "ongko", "biram", "kar", "juktoborno", "phola", "byanjon"],
}

# Precompute max key length per group for greedy matching
MAXLEN_PER_GROUP: Dict[str, int] = {g: (max((len(k) for k in m.keys()), default=0)) for g, m in GROUP_MAPS.items()}


def _find_longest(state: str, text: str, i: int) -> Tuple[str, str, str]:
    """Return (group, key, value) for the longest match allowed in current state. If none, return ("", "", "")."""
    allowed = STATE_GROUP_ORDER[state]
    # Determine the max lookahead we need
    maxlen = 0
    for g in allowed:
        maxlen = max(maxlen, MAXLEN_PER_GROUP[g])
    end = min(len(text), i + maxlen)
    best_group = ""
    best_key = ""
    best_val = ""
    best_len = 0

    # Try lengths from longest to shortest to implement greedy matching
    for L in range(end - i, 0, -1):
        chunk = text[i:i + L]
        # Check groups by priority
        for g in allowed:
            m = GROUP_MAPS[g]
            if chunk in m:
                # First match at this length wins due to priority order
                return (g, chunk, m[chunk])
    return ("", "", "")


def _apply_transition(state: str, group: str) -> str:
    # ---------------------------
    # INIT state
    # ---------------------------
    if state == INIT:
        if group in ("diacritic", "ng", "shor", "fkar"):
            return SHOR_STATE
        if group in ("prithayok", "ongko", "biram"):
            return SHOR_STATE
        if group == "reph":
            return REPH_STATE
        if group in ("juktoborno", "byanjon"):
            return BYANJON_STATE
        return INIT

    # ---------------------------
    # SHOR state
    # ---------------------------
    if state == SHOR_STATE:
        if group in ("diacritic", "ng", "fkar"):
            return SHOR_STATE
        if group == "shor":
            return SHOR_STATE
        if group in ("biram", "prithayok", "ongko"):
            return SHOR_STATE
        if group == "reph":
            return REPH_STATE
        if group in ("juktoborno", "byanjon"):
            return BYANJON_STATE
        return SHOR_STATE

    # ---------------------------
    # REPH state
    # ---------------------------
    if state == REPH_STATE:
        if group == "prithayok":
            return SHOR_STATE
        if group in ("diacritic", "ng"):
            return SHOR_STATE
        if group == "ae":
            return SHOR_STATE
        if group in ("juktoborno", "byanjon"):
            return BYANJON_STATE
        if group in ("kar", "nil"):
            return SHOR_STATE
        return REPH_STATE

    # ---------------------------
    # BYANJON state
    # ---------------------------
    if state == BYANJON_STATE:
        if group in ("diacritic", "ng", "prithayok", "ongko", "biram", "kar"):
            return SHOR_STATE
        if group in ("juktoborno", "phola", "byanjon"):
            return BYANJON_STATE
        return BYANJON_STATE

    return INIT


def convert(text: str) -> str:
    """Convert an ASCII input string to Bengali output using the bn-khipro state machine."""
    i = 0
    n = len(text)
    state = INIT
    out: List[str] = []

    while i < n:
        group, key, val = _find_longest(state, text, i)
        if not group:
            # No mapping: pass through this char and reset to INIT
            out.append(text[i])
            i += 1
            state = INIT
            continue

        # Special handling: PHOLA in BYANJON_STATE inserts virama before mapped char
        if state == BYANJON_STATE and group == "phola":
            out.append("্")
            out.append(val)
        else:
            out.append(val)

        i += len(key)
        state = _apply_transition(state, group)

    return "".join(out)


def type_stream(text: str) -> Generator[str, None, None]:
    """Generator: yields the converted text after each keystroke (1..len(text))."""
    for k in range(1, len(text) + 1):
        yield convert(text[:k])


# --------------------------
# Demo
# --------------------------
if __name__ == "__main__":
    import sys

    print("🔡 Khipro Typing Preview (Press Enter to quit)\n")

    while True:
        user_input = input("Type in khipro syntax: ").strip()
        if not user_input:
            break

        print("\nLive Typing:")
        for step, out in enumerate(type_stream(user_input), 1):
            print(f"{user_input[:step]!r} → {out}")
        print("-" * 40)