basa 0.1.0__py3-none-any.whl

basa/core/slang.py

@@ -0,0 +1,2071 @@
+"""
+BASA - Slang Normalization Module
+===================================
+Converts informal/colloquial Indonesian text (including Javanese/Sundanese
+particles) into standardized Indonesian.
+
+Supported categories:
+    - Pronouns (aku/kamu shortcuts)
+    - Kinship & address terms (kk, bokap, nyokap, dll)
+    - Negation (ga/gak/nggak → tidak)
+    - Compound negation (gamau, gabisa, gatau → tidak mau, dll)
+    - Conjunctions & prepositions (abbreviations like yg, dgn, krn)
+    - Common verbs (udah, blm, mo, tau, liat, dll)
+    - Adjectives & adverbs (bgt, bener, lumayan, dll)
+    - Question words (knp, gmn, kmn, dll)
+    - Greetings & responses (makasih, ok, sip, tq, dll)
+    - Temporal & location (skrg, ntr, kmrn, sini, dll)
+    - Internet slang → Indonesian equivalent (btw, lol, omg, dll)
+    - E-commerce & finance (ongkir, rekber, tf, dll)
+    - Youth & Gen-Z slang (mager, baper, gabut, dll)
+    - Discourse markers (mksd, intinya, menurutku, dll)
+    - Javanese extended (mangan, turu, apik, dll)
+    - Sundanese extended (abdi, anjeun, geus, dll)
+    - Nouns (temen, hp, laptop, dll)
+    - Health terms (dmm, btk, opname, dll)
+    - Emotions & expressions (galau, baper, kesal, ghosting, dll)
+    - Food & drink (nasgor, kopsu, laper, kenyang, dll)
+    - Clothing & fashion (ootd, thrifting, hoodie, dll)
+    - Transportation (ojol, krl, macet, nebeng, dll)
+    - Religion & culture (alhamdulillah, bismillah, ultah, dll)
+    - Education extended (bimbel, ospek, maba, snbt, dll)
+    - Work & office (wfh, deadline, lembur, resign, dll)
+    - Numbers & quantity (1rb, 5jt, rata2, tiba2, dll)
+    - Compound expressions (ngapain, otw ke, kapan2, dll)
+    - Particles (lah, deh, nih, dong, kok, sih, lho, pun, dll)
+
+Usage:
+    from basa.core.slang import slang
+
+    # Basic usage (convenience singleton)
+    result = slang.normalize("gw gamau pergi krn lg baper bgt")
+    # → "saya tidak mau pergi karena sedang bawa perasaan banget"
+
+    # Inspect all supported slang words
+    words = slang.supported_words()
+    # → {"adek": "adik", "ak": "aku", "ane": "saya", ...}
+
+    # Count total entries
+    len(slang)
+    # → 120+
+
+    # With custom entries
+    from basa.core.slang import SlangNormalizer
+    normalizer = SlangNormalizer(custom_mapping={"jancok": "ekspresi"})
+    result = normalizer.normalize("gw jancok kaget")
+
+    # Batch processing
+    texts = ["gw udah makan", "km blm tidur?"]
+    results = slang.normalize_batch(texts)
+
+Notes:
+    - All replacements are lowercased (standard for NLP preprocessing).
+    - Word boundaries (\\b) prevent partial-word substitution.
+    - Slang keys sorted longest-first to prevent premature partial matches.
+    - Repeated character normalization distinguishes vowels vs consonants
+      to avoid breaking abbreviations like "kk", "dll", "mm".
+"""
+
+import re
+
+# ─────────────────────────────────────────────────────────────────────────────
+# SLANG DICTIONARY (organized by category for easier maintenance)
+# ─────────────────────────────────────────────────────────────────────────────
+
+# First & second person pronouns
+_PRONOUNS: dict[str, str] = {
+    # First person singular
+    "aq": "aku",
+    "ak": "aku",
+    "akuh": "aku",
+    "q": "aku",
+    "sy": "saya",
+    "w": "saya",
+    "gw": "saya",
+    "gue": "saya",
+    "gua": "saya",
+    "ane": "saya",  # Betawi/Jakartan slang
+    "ana": "saya",  # Arabic-influenced
+    "wkwk": "tertawa",  # common filler — override happens in internet
+    # First person possessive
+    "gw nya": "saya punya",
+    "gue nya": "saya punya",
+    # Second person
+    "u": "kamu",
+    "lu": "kamu",
+    "lo": "kamu",
+    "loe": "kamu",
+    "elo": "kamu",
+    "elu": "kamu",
+    "km": "kamu",
+    "ente": "kamu",  # Betawi/Jakartan slang
+    "antum": "kamu",  # Arabic-influenced (plural polite)
+    "sampeyan": "kamu",  # Javanese polite
+    "sampean": "kamu",  # Javanese casual
+    "sampyan": "kamu",  # Javanese casual variant
+    # Third person
+    "dy": "dia",
+    "dya": "dia",
+    "doi": "dia",  # Jakartan slang for girlfriend/boyfriend
+    # Plural
+    "mrk": "mereka",
+    "qt": "kita",
+    "kln": "kalian",
+    # Reflexive
+    "sndiri": "sendiri",
+    "sndr": "sendiri",
+}
+
+# Kinship & address terms
+_KINSHIP: dict[str, str] = {
+    "kk": "kakak",
+    "kak": "kakak",
+    "adk": "adik",
+    "adek": "adik",
+    "bpk": "bapak",
+    "bpak": "bapak",
+    "bapk": "bapak",
+    "ayh": "ayah",
+    "klg": "keluarga",
+    "klrg": "keluarga",
+    "kel": "keluarga",
+    "ortu": "orang tua",
+    "ortuku": "orang tuaku",
+    "bokap": "ayah",
+    "nyokap": "ibu",
+    "pakde": "paman",
+    "bude": "bibi",
+    "om": "paman",
+    "tante": "bibi",
+}
+
+# Standalone negation words
+_NEGATION: dict[str, str] = {
+    "g": "tidak",
+    "ga": "tidak",
+    "gak": "tidak",
+    "gk": "tidak",
+    "ngga": "tidak",
+    "nggak": "tidak",
+    "engga": "tidak",
+    "enggak": "tidak",
+    "kagak": "tidak",
+    "nope": "tidak",
+    "nop": "tidak",
+    "gakk": "tidak",
+    "gaknya": "tidaknya",
+    "bkn": "bukan",
+    "bknnya": "bukannya",
+    "jgn": "jangan",
+    "jgnan": "jangan",
+    "belom": "belum",
+    "belom2": "belum-belum",
+    "blom": "belum",
+    "blm2": "belum-belum",
+    "masih blm": "masih belum",
+}
+
+# Compound negation — written as one word without spaces (very common in chat)
+# e.g. "gamau" = "ga" + "mau" = "tidak mau"
+# NOTE: These MUST be registered before single-char negation in the dict so
+# that the longest-first sort in _compile_pattern() handles them correctly.
+_COMPOUND_NEGATION: dict[str, str] = {
+    "gamau": "tidak mau",
+    "gmau": "tidak mau",
+    "gabisa": "tidak bisa",
+    "gbisa": "tidak bisa",
+    "gakbisa": "tidak bisa",
+    "gapunya": "tidak punya",
+    "gpunya": "tidak punya",
+    "gasuka": "tidak suka",
+    "gsuka": "tidak suka",
+    "gaada": "tidak ada",
+    "gada": "tidak ada",  # gaada → gada after vowel reduction
+    "gatau": "tidak tahu",
+    "gtau": "tidak tahu",
+    "gatahu": "tidak tahu",
+    "gaperlu": "tidak perlu",
+    "gperlu": "tidak perlu",
+    "gaboleh": "tidak boleh",
+    "gboleh": "tidak boleh",
+    "gamakan": "tidak makan",
+    "gaminum": "tidak minum",
+    "gadatang": "tidak datang",
+    "gamasuk": "tidak masuk",
+    "gasempet": "tidak sempat",
+    "gabawa": "tidak bawa",
+    "gabayar": "tidak bayar",
+    "gaberani": "tidak berani",
+    "gabalik": "tidak balik",
+    "gabener": "tidak benar",
+    "gakuat": "tidak kuat",
+    "gamungkin": "tidak mungkin",
+}
+
+# Conjunctions, prepositions, and formal abbreviations
+_CONJUNCTIONS: dict[str, str] = {
+    "tp": "tapi",
+    "tpi": "tapi",
+    "trs": "terus",
+    "trus": "terus",
+    "dr": "dari",
+    "sm": "sama",
+    "ama": "sama",
+    "dpt": "dapat",
+    "tdk": "tidak",
+    "krn": "karena",
+    "krna": "karena",
+    "karna": "karena",
+    "utk": "untuk",
+    "bwt": "buat",
+    "spy": "supaya",
+    "klo": "kalau",
+    "kalo": "kalau",
+    "yg": "yang",
+    "dgn": "dengan",
+    "dg": "dengan",
+    "jg": "juga",
+    "bkn": "bukan",
+    "pd": "pada",
+    "dlm": "dalam",
+    # Formal abbreviations
+    "dll": "dan lain-lain",
+    "dsb": "dan sebagainya",
+    "dst": "dan seterusnya",
+    "tsb": "tersebut",
+    "yth": "yang terhormat",
+    "nb": "catatan",
+    # Time-relation conjunctions
+    "stlh": "setelah",
+    "sblm": "sebelum",
+    "slma": "selama",
+    "tntg": "tentang",
+    "ttng": "tentang",
+    "thd": "terhadap",
+    "antr": "antara",
+    "srt": "serta",
+    "kmdian": "kemudian",
+    "kmdn": "kemudian",
+    "lalu": "kemudian",
+    # Formal/business abbreviations
+    "ttd": "tanda tangan",
+    "an": "atas nama",
+    "up": "untuk perhatian",
+    "re": "balasan",
+    "fwd": "teruskan",
+    "fw": "teruskan",
+    "lmk": "beri tahu saya",
+    "etc": "dan lain-lain",
+    "ps": "catatan tambahan",
+    "no": "nomor",
+}
+
+# Verbs and auxiliary verbs
+_VERBS: dict[str, str] = {
+    "udh": "sudah",
+    "sdh": "sudah",
+    "dah": "sudah",
+    "udah": "sudah",
+    "udeh": "sudah",
+    "blm": "belum",
+    "blum": "belum",
+    "lg": "sedang",
+    "lgi": "sedang",
+    "bs": "bisa",
+    "bsa": "bisa",
+    "jd": "jadi",
+    "mo": "mau",
+    "mw": "mau",
+    "tau": "tahu",
+    "taw": "tahu",
+    "pake": "pakai",
+    "make": "pakai",
+    "mkn": "makan",
+    "mkan": "makan",
+    "liat": "lihat",
+    "liad": "lihat",
+    "ngerti": "mengerti",
+    "ngarti": "mengerti",
+    "nanya": "bertanya",
+    "nany": "bertanya",
+    "nyari": "mencari",
+    "nyoba": "mencoba",
+    "ngomong": "berbicara",
+    "nunggu": "menunggu",
+    "lakuin": "lakukan",
+    "byr": "bayar",
+    "hrs": "harus",
+    "blh": "boleh",
+    "prlu": "perlu",
+    "blg": "bilang",
+    "bilng": "bilang",
+    "sk": "suka",
+    "dtg": "datang",
+    "dateng": "datang",
+    # Extended verbs
+    "bikin": "membuat",
+    "bkin": "membuat",
+    "ambl": "ambil",
+    "simpen": "simpan",
+    "smpen": "simpan",
+    "masukin": "masukkan",
+    "keluarin": "keluarkan",
+    "tambahin": "tambahkan",
+    "kurangin": "kurangi",
+    "hapusin": "hapuskan",
+    "benerin": "betulkan",
+    "gnti": "ganti",
+    "pndh": "pindah",
+    "plng": "pulang",
+    "brngkt": "berangkat",
+    "balik": "kembali",
+    "blk": "kembali",
+    "nyampe": "sampai",
+    "nyampek": "sampai",
+    "smpe": "sampai",
+    "sampe": "sampai",
+    "ketemu": "bertemu",
+    "ktmu": "bertemu",
+    "ngikut": "ikut",
+    "ikutin": "ikuti",
+    "dengerin": "dengarkan",
+    "denger": "dengar",
+    "dnger": "dengar",
+    "lpd": "lupa",
+    "inget": "ingat",
+    "ngelupain": "melupakan",
+    "ngebantu": "membantu",
+    "bntu": "bantu",
+    "cek": "periksa",
+    "ngecek": "memeriksa",
+    "cobaain": "coba",
+    "ngelakuin": "melakukan",
+    "mnta": "minta",
+    "mintain": "mintakan",
+    "krm": "kirim",
+    "kirimin": "kirimkan",
+    "bli": "beli",
+    "beliin": "belikan",
+    "jln": "jalan",
+    "maen": "main",
+    "nemenin": "menemani",
+    "temenin": "menemani",
+    "temenan": "berteman",
+    "kenalan": "berkenalan",
+    "ngobrol": "mengobrol",
+    "ngopi": "minum kopi",
+    "nitip": "titip",
+    "titipin": "titipkan",
+    "bayarin": "bayarkan",
+    "bljr": "belajar",
+    "bljar": "belajar",
+    "ngajar": "mengajar",
+    "ngjar": "mengajar",
+    "pegi": "pergi",
+}
+
+# Adjectives, adverbs, and discourse particles
+_ADJECTIVES_ADVERBS: dict[str, str] = {
+    "bgt": "banget",
+    "bngt": "banget",
+    "bnget": "banget",
+    "bget": "banget",
+    "bet": "banget",
+    "amet": "amat",
+    "bener": "benar",
+    "bnr": "benar",
+    "bner": "benar",
+    "cepet": "cepat",
+    "cpet": "cepat",
+    "cpat": "cepat",
+    "ttp": "tetap",
+    "lbh": "lebih",
+    "krg": "kurang",
+    "sdkt": "sedikit",
+    "dikit": "sedikit",
+    "dkit": "sedikit",
+    "hny": "hanya",
+    "hnya": "hanya",
+    "bbrp": "beberapa",
+    "bbrpa": "beberapa",
+    "emg": "memang",
+    "emang": "memang",
+    "aja": "saja",
+    "doang": "saja",
+    "kek": "seperti",
+    "kayak": "seperti",
+    "kyk": "seperti",
+    "syg": "sayang",
+    "syng": "sayang",
+    "syang": "sayang",
+    "sayng": "sayang",
+    # Common informal particles mapped to nearest standard equivalent
+    "nih": "ini",
+    "tuh": "itu",
+    "ntu": "itu",
+    "yok": "ayo",
+    "yuk": "ayo",
+    "ae": "saja",  # Sundanese/Javanese particle
+    "wae": "saja",  # Sundanese: just/only
+    "tok": "saja",  # Javanese: only
+    # Extended adjectives/adverbs
+    "abis": "habis",
+    "abish": "habis",
+    "hbs": "habis",
+    "pol": "sekali",
+    "beneran": "sungguhan",
+    "lm": "lama",
+    "lame": "lama",
+    "trllu": "terlalu",
+    "ckp": "cukup",
+    "smpt": "sempat",
+    "prnh": "pernah",
+    "pernh": "pernah",
+    "prnah": "pernah",
+    "srg": "sering",
+    "jrng": "jarang",
+    "jrang": "jarang",
+    "sllu": "selalu",
+    "slalu": "selalu",
+    "kdng": "kadang",
+    "mngkin": "mungkin",
+    "mngkn": "mungkin",
+    "mgkn": "mungkin",
+    "pst": "pasti",
+    "trnyta": "ternyata",
+    "trnyata": "ternyata",
+    "lgsg": "langsung",
+    "lngsng": "langsung",
+    "mlh": "malah",
+    "pdhl": "padahal",
+    "pdhal": "padahal",
+    "pdahl": "padahal",
+    "slnya": "soalnya",
+    "gitu": "begitu",
+    "gt": "begitu",
+    "gini": "begini",
+    "gn": "begini",
+    "segini": "sebanyak ini",
+    "sgini": "sebanyak ini",
+    "segitu": "sebanyak itu",
+    "sgitu": "sebanyak itu",
+    "bnyk": "banyak",
+    "byk": "banyak",
+    "bnyak": "banyak",
+    "sdkit": "sedikit",
+    "bgs": "bagus",
+    "bgus": "bagus",
+    "bagoes": "bagus",
+    "jlk": "jelek",
+    "jlek": "jelek",
+    "kcewa": "kecewa",
+    "mhl": "mahal",
+    "mhal": "mahal",
+    "mahl": "mahal",
+    "mrh": "murah",
+    "pinter": "pintar",
+    "pintr": "pintar",
+    "pntar": "pintar",
+    "pntr": "pintar",
+    "hbat": "hebat",
+    "hebt": "hebat",
+    "hbt": "hebat",
+    "lmyn": "lumayan",
+    "lumyn": "lumayan",
+    "lmyan": "lumayan",
+    "mayan": "lumayan",
+    "enk": "enak",
+    "mntap": "mantap",
+    "asik": "asyik",
+    "cpk": "capai",
+    "cape": "capai",
+    "lela": "lelah",
+    "ngantu": "mengantuk",
+    "laper": "lapar",
+    "lper": "lapar",
+    "skt": "sakit",
+    "skit": "sakit",
+    "smbh": "sembuh",
+    "smbuh": "sembuh",
+    "seneng": "senang",
+    "sneng": "senang",
+    "bhgia": "bahagia",
+    "tkt": "takut",
+    "tkut": "takut",
+    "bngun": "bingung",
+    "psng": "pusing",
+    "pusng": "pusing",
+    "pening": "pusing",
+    "pning": "pusing",
+    "mls": "malas",
+    "mlas": "malas",
+    "mles": "malas",
+    "males": "malas",
+    "rjn": "rajin",
+    "rjin": "rajin",
+    "sbnrnya": "sebenarnya",
+    "sbnr": "sebenarnya",
+    "mending": "lebih baik",
+    "mnding": "lebih baik",
+    "kyknya": "sepertinya",
+    "sptnya": "sepertinya",
+    "sprtnya": "sepertinya",
+    "gitu2": "begitu-begitu",
+    "pelan2": "pelan-pelan",
+    "lama2": "lama-lama",
+    "buru2": "terburu-buru",
+    "kwalitas": "kualitas",
+}
+
+# Question words
+_QUESTION_WORDS: dict[str, str] = {
+    "knp": "kenapa",
+    "knpa": "kenapa",
+    "knapa": "kenapa",
+    "gmn": "bagaimana",
+    "gmna": "bagaimana",
+    "gimna": "bagaimana",
+    "gimana": "bagaimana",
+    "kmn": "kemana",
+    "kmna": "kemana",
+    "kmana": "kemana",
+    "kemna": "kemana",
+    "dmn": "di mana",
+    "dmna": "di mana",
+    "dmana": "di mana",
+    "dimna": "di mana",
+    "dimana": "di mana",
+    "drmn": "dari mana",
+    "drmna": "dari mana",
+    "drmana": "dari mana",
+    "kpn": "kapan",
+    "kpan": "kapan",
+    "manaa": "mana",
+    # Sundanese question words
+    "kumaha": "bagaimana",
+    "naon": "apa",
+    "saha": "siapa",
+    "iraha": "kapan",
+    "kamana": "kemana",
+}
+
+# Greetings, affirmations, and common social responses
+_GREETINGS_RESPONSES: dict[str, str] = {
+    # Thanks
+    "makasih": "terima kasih",
+    "makasi": "terima kasih",
+    "mksh": "terima kasih",
+    "mksih": "terima kasih",
+    "trims": "terima kasih",
+    "thx": "terima kasih",
+    "tq": "terima kasih",
+    "ty": "terima kasih",
+    "trimakasih": "terima kasih",
+    "terimakasih": "terima kasih",
+    "mkch": "terima kasih",
+    "tengkyu": "terima kasih",
+    "tengkyuu": "terima kasih",
+    # Requests
+    "pls": "tolong",
+    "pliss": "tolong",
+    "plz": "tolong",
+    "plzz": "tolong",
+    "tlg": "tolong",
+    # Apology
+    "mf": "maaf",
+    "afwan": "maaf",  # Arabic-Indonesian
+    "maap": "maaf",
+    "sori": "maaf",
+    "sory": "maaf",
+    "sorry": "maaf",
+    "sorri": "maaf",
+    "mnt maaf": "minta maaf",
+    # Agreement
+    "ok": "oke",
+    "okk": "oke",
+    "okee": "oke",
+    "okelah": "oke",
+    "sip": "baik",
+    "siapp": "siap",
+    "iy": "iya",
+    "iyap": "iya",
+    "yap": "ya",
+    "yep": "ya",
+    "yup": "ya",
+    "yupp": "ya",
+    "yeah": "ya",
+    "yoi": "iya",  # Jakartan slang
+    "mantul": "mantap betul",
+    # Greetings
+    "hi": "hai",
+    "hey": "hei",
+    "met pagi": "selamat pagi",
+    "met siang": "selamat siang",
+    "met sore": "selamat sore",
+    "met malam": "selamat malam",
+    "slmt dtng": "selamat datang",
+    "smpai jumpa": "sampai jumpa",
+    "dadah": "sampai jumpa",
+    "dah": "sampai jumpa",
+    "bye": "selamat tinggal",
+    "byee": "selamat tinggal",
+    "byebye": "selamat tinggal",
+    "ciao": "selamat tinggal",
+    "papay": "selamat tinggal",
+    # Javanese / Sundanese particles (polite registers)
+    "nggih": "iya",  # Javanese: yes (polite)
+    "inggih": "iya",  # Javanese: yes (formal)
+    "mboten": "tidak",  # Javanese: no/not
+    "monggo": "silakan",  # Javanese: please go ahead
+    "suwun": "terima kasih",  # Javanese: thank you
+    "matur suwun": "terima kasih",  # Javanese: thank you (full form)
+    "nuwun": "terima kasih",  # Javanese: thank you
+    # Sundanese
+    "muhun": "iya",
+    "hapunten": "maaf",
+    "punten": "permisi",
+    "hatur nuhun": "terima kasih",
+    # Arabic-Indonesian
+    "syukron": "terima kasih",
+    # Polite expressions
+    "prmisi": "permisi",
+    "silahkan": "silakan",
+    "samasama": "sama-sama",
+    "samsama": "sama-sama",
+    "sama sama": "sama-sama",
+}
+
+# Temporal expressions and locations
+_TEMPORAL_LOCATION: dict[str, str] = {
+    "skrg": "sekarang",
+    "skrng": "sekarang",
+    "ntr": "nanti",
+    "ntar": "nanti",
+    "kmrn": "kemarin",
+    "kmren": "kemarin",
+    "bsk": "besok",
+    "td": "tadi",
+    "dlu": "dulu",
+    "dlo": "dulu",
+    "hr": "hari",
+    "hri": "hari",
+    "mlm": "malam",
+    "malem": "malam",
+    "pg": "pagi",
+    "sni": "sini",
+    "dsini": "di sini",
+    "dsni": "di sini",
+    "sono": "sana",
+    "mgg": "minggu",
+    "bln": "bulan",
+    "thn": "tahun",
+    "mnt": "menit",
+    "dtk": "detik",
+    "jdwl": "jadwal",
+    "wktu": "waktu",
+}
+
+# Internet slang — mapped to Indonesian equivalents (not English)
+_INTERNET_SLANG: dict[str, str] = {
+    "otw": "dalam perjalanan",
+    "btw": "omong-omong",
+    "fyi": "sebagai informasi",
+    "imo": "menurut saya",
+    "imho": "menurut saya",
+    "cmiiw": "koreksi jika saya salah",
+    "lol": "tertawa",
+    "wkwk": "tertawa",
+    "wkwwk": "tertawa",
+    "wkwkwk": "tertawa",
+    "awokwok": "tertawa",
+    "xixi": "tertawa",  # Chinese laughter, common in Indonesian social media
+    "lmao": "tertawa keras",
+    "omg": "astaga",
+    "idk": "saya tidak tahu",
+    "tbh": "sejujurnya",
+    "ngl": "sejujurnya",
+    "tbf": "sejujurnya",
+    "nvm": "lupakan saja",
+    "jk": "bercanda",
+    "np": "tidak masalah",
+    "afk": "pergi sebentar",
+    "brb": "segera kembali",
+    "dm": "pesan langsung",
+    "tldr": "ringkasnya",
+    "wdyt": "menurutmu",
+    "wdym": "apa maksudmu",
+    "ftr": "akhirnya",
+    # Platforms
+    "wa": "whatsapp",
+    "ig": "instagram",
+    "fb": "facebook",
+    "tt": "tiktok",
+    "yt": "youtube",
+    "twit": "twitter",
+    "linkin": "linkedin",
+    "gform": "google form",
+    "gdrive": "google drive",
+    "gmeet": "google meet",
+    # Social media terms
+    "wag": "grup whatsapp",
+    "gc": "grup chat",
+    "bc": "pesan siaran",
+    "nohp": "nomor handphone",
+    "ootd": "pakaian hari ini",
+    "vibe": "suasana",
+    "vibes": "suasana",
+    "mood": "suasana hati",
+    "hype": "heboh",
+    "trending": "sedang tren",
+    "lowkey": "diam-diam",
+    "highkey": "terang-terangan",
+}
+
+# E-commerce and finance terms
+_ECOMMERCE_FINANCE: dict[str, str] = {
+    "gan": "saudara",
+    "sis": "saudari",
+    "rekber": "rekening bersama",
+    "cod": "bayar di tempat",
+    "ongkir": "ongkos kirim",
+    "tf": "transfer",
+    "dp": "uang muka",
+    "hrg": "harga",
+    "hrga": "harga",
+    "jastip": "jasa titip",
+    "pesen": "pesan",
+    "minat": "berminat",
+    "duit": "uang",
+    "duwit": "uang",
+    "fulus": "uang",
+    "rb": "ribu",
+    "jt": "juta",
+    "mlyr": "miliar",
+    "pcs": "buah",
+    "lbr": "lembar",
+    "btl": "botol",
+    "asap": "sesegera mungkin",
+    "acc": "disetujui",
+    "bkl": "bakal",
+    "bakal": "akan",
+    "mslh": "masalah",
+    "sls": "selesai",
+    "slsai": "selesai",
+    "mlai": "mulai",
+    "mulain": "mulai",
+}
+
+# Youth / Gen-Z Indonesian slang
+_YOUTH_SLANG: dict[str, str] = {
+    "mager": "malas bergerak",
+    "baper": "bawa perasaan",
+    "gabut": "tidak ada kegiatan",
+    "kepo": "ingin tahu",
+    "santuy": "santai",
+    "woles": "santai",
+    "gercep": "gerak cepat",
+    "gass": "ayo",
+    "gaspol": "ayo semangat",
+    "gaskeun": "ayo lakukan",  # Sundanese-influenced
+    "bucin": "budak cinta",
+    "julid": "iri hati",
+    "halu": "halusinasi",
+    "flexing": "pamer",
+    "pansos": "panjat sosial",
+    "curhat": "berbagi cerita",
+    "kekinian": "tren terkini",
+    "kekinan": "tren terkini",
+    "hits": "populer",
+    "relate": "relatable",
+    "selfie": "foto sendiri",
+    "caption": "keterangan foto",
+    "story": "cerita",
+    "feeds": "beranda",
+}
+
+# Discourse markers and meta-commentary
+_DISCOURSE_MARKERS: dict[str, str] = {
+    "mksd": "maksud",
+    "mksdnya": "maksudnya",
+    "pkoknya": "pokoknya",
+    "pokonya": "pokoknya",
+    "jdinya": "jadinya",
+    "hslnya": "hasilnya",
+    "sngktnya": "singkatnya",
+    "menurutku": "menurut saya",
+    "mnrtku": "menurut saya",
+    "menurutgw": "menurut saya",
+    "akhirny": "akhirnya",
+}
+
+# Javanese extended vocabulary
+_JAVANESE_EXTENDED: dict[str, str] = {
+    # Pronouns
+    "awakmu": "kamu",
+    "njenengan": "anda",
+    "kulo": "saya",
+    "kula": "saya",
+    "inyong": "saya",
+    "dewek": "sendiri",
+    # Adjectives
+    "apik": "bagus",
+    "elek": "jelek",
+    "okeh": "banyak",
+    "sithik": "sedikit",
+    "gedhe": "besar",
+    "cilik": "kecil",
+    "adoh": "jauh",
+    "cedhak": "dekat",
+    "alon": "pelan",
+    "anyar": "baru",
+    "lawas": "lama",
+    "sae": "baik",
+    # Verbs
+    "mangkat": "berangkat",
+    "kondur": "pulang",
+    "mangan": "makan",
+    "ngombe": "minum",
+    "turu": "tidur",
+    "tangi": "bangun",
+    "mlaku": "berjalan",
+    "mlayu": "berlari",
+    "lungguh": "duduk",
+    "ngadeg": "berdiri",
+    "weruh": "tahu",
+    "krungu": "dengar",
+    "ndelok": "lihat",
+    "takon": "bertanya",
+    "kandha": "berkata",
+    "tuku": "beli",
+    "adol": "jual",
+}
+
+# Sundanese extended vocabulary
+_SUNDANESE_EXTENDED: dict[str, str] = {
+    # Pronouns
+    "abdi": "saya",
+    "anjeun": "kamu",
+    "maneh": "kamu",
+    # Particles
+    "atuh": "dong",
+    "euy": "ya",
+    "mah": "sih",
+    "teh": "itu",
+    "nya": "ya",
+    # Adjectives
+    "alus": "bagus",
+    "awon": "jelek",
+    "caket": "dekat",
+    "leueur": "lambat",
+    "sae pisan": "sangat bagus",
+    # Verbs/expressions
+    "geus": "sudah",
+    "keur": "sedang",
+    "bade": "akan",
+    "hoyong": "mau",
+    "daek": "mau",
+    "tiasa": "bisa",
+    "aya": "ada",
+    "teu": "tidak",
+    "acan": "belum",
+    "can": "belum",
+    # Quantity
+    "loba": "banyak",
+    "saeutik": "sedikit",
+    "ged\u00e9": "besar",
+    "leutik": "kecil",
+}
+
+# Common nouns (abbreviations and informal forms)
+_NOUNS: dict[str, str] = {
+    # People & relationships
+    "tmn": "teman",
+    "temen": "teman",
+    "tmn2": "teman-teman",
+    "temen2": "teman-teman",
+    "shbt": "sahabat",
+    "pcr": "pacar",
+    "mntm": "mantan",
+    # Work & study
+    "krjn": "pekerjaan",
+    "kntr": "kantor",
+    "sklh": "sekolah",
+    "kmps": "kampus",
+    "kliah": "kuliah",
+    "tgs": "tugas",
+    "pr": "pekerjaan rumah",
+    "ujin": "ujian",
+    "ulngn": "ulangan",
+    "mhsw": "mahasiswa",
+    "mhs": "mahasiswa",
+    "mrd": "murid",
+    "kls": "kelas",
+    "mapel": "mata pelajaran",
+    "matkul": "mata kuliah",
+    "mtkl": "mata kuliah",
+    "skrpsi": "skripsi",
+    "wsd": "wisuda",
+    "nlai": "nilai",
+    "ipk": "indeks prestasi kumulatif",
+    "ip": "indeks prestasi",
+    "dl": "batas waktu",
+    "prsnts": "presentasi",
+    "lprn": "laporan",
+    "rvisi": "revisi",
+    # Things & gadgets
+    "hp": "handphone",
+    "hape": "handphone",
+    "ltop": "laptop",
+    "mtr": "motor",
+    "mbl": "mobil",
+    "rmh": "rumah",
+    "mkn": "makan",
+    "mnum": "minum",
+    # Places
+    "wrg": "warung",
+    "warteg": "warung tegal",
+    "psr": "pasar",
+    "gdg": "gedung",
+    "rs": "rumah sakit",
+    "pskms": "puskesmas",
+}
+
+# Health & wellness vocabulary
+_HEALTH: dict[str, str] = {
+    # Symptoms
+    "dmm": "demam",
+    "dmam": "demam",
+    "btk": "batuk",
+    "btuk": "batuk",
+    "plk": "pilek",
+    "ml": "mual",
+    "mntah": "muntah",
+    "alrgi": "alergi",
+    "sesak": "sesak napas",
+    "sktkpl": "sakit kepala",
+    "mencret": "diare",
+    "mules": "mulas",
+    "lebam": "memar",
+    "kslng": "keseleo",
+    # Medical personnel & places
+    "obt": "obat",
+    "dkter": "dokter",
+    "dr": "dokter",
+    "prwt": "perawat",
+    "rwt": "rawat",
+    "opname": "rawat inap",
+    "ugd": "unit gawat darurat",
+    "igd": "instalasi gawat darurat",
+    "poli": "poliklinik",
+    "apotk": "apotek",
+    "resep": "resep dokter",
+    "rsep": "resep dokter",
+    "rapid test": "tes cepat",
+    # Wellness & fitness
+    "jalan2": "jalan-jalan",
+    "olrga": "olahraga",
+    "gym": "olahraga",
+    "lari": "berlari",
+    "jogging": "joging",
+    "tdr": "tidur",
+    "tdrsiang": "tidur siang",
+    "bgdng": "begadang",
+    "istrhat": "istirahat",
+    "istrht": "istirahat",
+    "shat": "sehat",
+    "smbh": "sembuh",
+    "lekas sembuh": "lekas sembuh",
+    "cpt smbh": "lekas sembuh",
+    "get well soon": "lekas sembuh",
+    "gws": "lekas sembuh",
+    "isoman": "isolasi mandiri",
+}
+
+# Emotions, feelings & expressions
+_EMOTIONS_EXPRESSIONS: dict[str, str] = {
+    # Happiness / excitement
+    "happy": "bahagia",
+    "excited": "bersemangat",
+    "semngt": "semangat",
+    "smngat": "semangat",
+    "smgt": "semangat",
+    "bersemgt": "bersemangat",
+    "yey": "hore",
+    "yeay": "hore",
+    "yay": "hore",
+    "asoy": "asyik",
+    "krn2": "keren-keren",
+    "mantab": "mantap",
+    "yoi": "iya",
+    "gokil": "luar biasa",
+    "gila": "luar biasa",
+    "gile": "luar biasa",
+    "gilak": "luar biasa",
+    "parah": "luar biasa",
+    "lebay": "berlebihan",
+    "lbay": "berlebihan",
+    # Sadness / disappointment
+    "sdh2": "sedih",
+    "kcewa": "kecewa",
+    "kzl": "kesal",
+    "bete": "kesal",
+    "bt": "kesal",
+    "gondok": "kesal",
+    "sewot": "kesal",
+    "jutek": "judes",
+    "nangis": "menangis",
+    "nangiss": "menangis",
+    "mewek": "menangis",
+    "ngedumel": "mengeluh",
+    "ngeluh": "mengeluh",
+    "ngrundel": "mengeluh",
+    "geregetan": "kesal",
+    "gereget": "kesal",
+    "dongkol": "kesal",
+    "badmood": "suasana hati buruk",
+    "bad mood": "suasana hati buruk",
+    "overthink": "terlalu banyak pikiran",
+    "otk": "terlalu banyak pikiran",
+    "down": "sedang terpuruk",
+    "ngedown": "sedang terpuruk",
+    "insecure": "tidak percaya diri",
+    "minder": "rendah diri",
+    # Surprise / shock
+    "shock": "syok",
+    "bnggong": "bengong",
+    "melongo": "terpana",
+    "bnggung": "bingung",
+    "mumet": "pusing",
+    "mupeng": "sangat menginginkan",
+    # Love / affection
+    "syangku": "sayangku",
+    "cintrong": "cinta",
+    "gebetan2": "gebetan-gebetan",
+    "pdkt": "pendekatan",
+    "chat": "pesan",
+    "dm2": "kirim pesan langsung",
+    "ghosting": "menghilang tiba-tiba",
+    "ghosted": "ditinggal tiba-tiba",
+    "friendzone": "zona teman",
+    "fz": "zona teman",
+    "crush": "orang yang disukai",
+    "clbk": "cinta lama bersemi kembali",
+    "php": "pemberi harapan palsu",
+    "playing": "mempermainkan",
+    "dimainin": "dipermainkan",
+    "selingkuh": "berselingkuh",
+    "selingkuhi": "berselingkuhi",
+    "selingkuhan": "perselingkuhan",
+    "putusin": "memutuskan",
+    "balikan": "bersama kembali",
+    # Tired / lazy
+    "capek": "lelah",
+    "capai": "lelah",
+    "kecapean": "kelelahan",
+    "nganggur": "tidak bekerja",
+    "males": "malas",
+    "mager": "malas bergerak",
+    "rebahan": "berbaring santai",
+    "santey": "santai",
+    "ndak": "tidak",
+    "enggak": "tidak",
+    "nyesel": "menyesal",
+    "ngsyel": "menyesal",
+}
+
+# Food & drink vocabulary
+_FOOD_DRINK: dict[str, str] = {
+    # General food terms
+    "mkan": "makan",
+    "mkanan": "makanan",
+    "lauk": "lauk-pauk",
+    "ns": "nasi",
+    "sayur": "sayuran",
+    "syr": "sayuran",
+    "mie": "mi",
+    "mie2": "mie-mie",
+    "indomie": "mi instan",
+    "indmie": "mi instan",
+    "indomie goreng": "mi goreng instan",
+    "migo": "mi goreng",
+    "migor": "mi goreng",
+    "nasgor": "nasi goreng",
+    "nsgr": "nasi goreng",
+    "molen": "kue molen",
+    "grngn": "gorengan",
+    "batagor": "bakso tahu goreng",
+    "cireng": "aci goreng",
+    "cilok": "aci dicolok",
+    "baso": "bakso",
+    "bks": "bakso",
+    "gado2": "gado-gado",
+    "blr": "bubur",
+    "nsuduk": "nasi uduk",
+    "warteg": "warung tegal",
+    "wrtg": "warung tegal",
+    # Drinks
+    "mnum": "minum",
+    "mnuman": "minuman",
+    "kpi": "kopi",
+    "tehmanis": "teh manis",
+    "juice": "jus",
+    "bubble tea": "boba teh",
+    "kopsu": "kopi susu",
+    "americano": "kopi americano",
+    "minol": "minuman beralkohol",
+    # Taste / quality
+    "enakk": "enak",
+    "mantul": "mantap betul",
+    "kriuk": "renyah",
+    "pedes": "pedas",
+    "asem": "asam",
+    "bukan enak": "tidak enak",
+    "gaenak": "tidak enak",
+    "ga enak": "tidak enak",
+    "keabisan": "kehabisan",
+    "khabisan": "kehabisan",
+    # Meal times / hunger
+    "makanmalam": "makan malam",
+    "srpn": "sarapan",
+    "malem": "malam",
+    "laper": "lapar",
+    "laperan": "sedang lapar",
+    "kelaperan": "sangat lapar",
+    "ngidam": "mengidam",
+    "nagih": "ketagihan",
+    "kenyangn": "kekenyangan",
+}
+
+# Clothing & fashion vocabulary
+_CLOTHING_FASHION: dict[str, str] = {
+    # Garments
+    "bj": "baju",
+    "kaos": "kaus",
+    "kos": "kaus",
+    "kmja": "kemeja",
+    "jket": "jaket",
+    "cln": "celana",
+    "dress": "gaun",
+    "jlbb": "jilbab",
+    "hijaber": "pengguna hijab",
+    "krdng": "kerudung",
+    "koko": "baju koko",
+    "pjm": "piyama",
+    "sweter": "sweater",
+    "hudi": "hoodie",
+    "cardigan": "kardigan",
+    "outer": "jaket luar",
+    "outher": "jaket luar",
+    "vest": "rompi",
+    # Footwear
+    "spt2": "sepatu-sepatu",
+    "sdl": "sandal",
+    "sneakers": "sepatu kets",
+    "snakers": "sepatu kets",
+    "boots": "sepatu bot",
+    "heels": "sepatu hak tinggi",
+    "higheels": "sepatu hak tinggi",
+    "flatshoes": "sepatu flat",
+    "slippers": "sandal",
+    # Accessories
+    "tote bag": "tas tote",
+    "totebag": "tas tote",
+    "dmpt": "dompet",
+    "jamtgn": "jam tangan",
+    "kacmt": "kacamata",
+    # Fashion terms
+    "outfit": "pakaian",
+    "ootd": "pakaian hari ini",
+    "mix match": "padupadan",
+    "mixmatch": "padupadan",
+    "casual": "kasual",
+    "stylish": "bergaya",
+    "trendy": "trendi",
+    "vintage": "jadul bergaya",
+    "thrift": "beli baju bekas",
+    "thrifting": "belanja baju bekas",
+    "preloved": "bekas pakai",
+    "ukrn": "ukuran",
+    "size": "ukuran",
+    "kebesaran": "terlalu besar",
+    "kekecilan": "terlalu kecil",
+}
+
+# Transportation vocabulary
+_TRANSPORTATION: dict[str, str] = {
+    # Vehicles
+    "mtr": "motor",
+    "mbl": "mobil",
+    "spd": "sepeda",
+    "angkot": "angkutan kota",
+    "busway": "bus transjakarta",
+    "tj": "transjakarta",
+    "krl": "kereta rel listrik",
+    "mrt": "mass rapid transit",
+    "lrt": "light rail transit",
+    "commuter": "kereta komuter",
+    "komuter": "kereta komuter",
+    "pswt": "pesawat",
+    "bdr": "bandara",
+    "kpl": "kapal",
+    "ojol": "ojek online",
+    "taxi": "taksi",
+    "bentor": "becak motor",
+    "bnsn": "bensin",
+    "bbm": "bahan bakar minyak",
+    # Navigation / travel
+    "jln": "jalan",
+    "mct": "macet",
+    "plkmpng": "pulang kampung",
+    "prjlnan": "perjalanan",
+    "jrk": "jarak",
+    "deket": "dekat",
+    "prk": "parkir",
+    "numpang": "menumpang",
+    "nunut": "menumpang",
+    "nebeng": "menumpang",
+    "nyetir": "mengemudi",
+    "nyupir": "mengemudikan",
+    "setir": "kemudi",
+    "spion": "kaca spion",
+    "gps": "navigasi",
+    "maps": "peta navigasi",
+    "gmaps": "google maps",
+}
+
+# Religion, culture & social life
+_RELIGION_CULTURE: dict[str, str] = {
+    # Islamic expressions (very common in Indonesian daily speech)
+    "insyaallah": "insya allah",
+    "inshallah": "insya allah",
+    "ia": "insya allah",
+    "alhdlh": "alhamdulillah",
+    "alhdlll": "alhamdulillah",
+    "amdlh": "alhamdulillah",
+    "alhamdu": "alhamdulillah",
+    "masyaallah": "masya allah",
+    "subhankl": "subhanallah",
+    "astagfirullah": "astaghfirullah",
+    "astghfr": "astaghfirullah",
+    "bsmllh": "bismillah",
+    "bismillahirrahmanirrahim": "bismillah",
+    "allahuakbar": "allahu akbar",
+    "aamiin": "amin",
+    "aminn": "amin",
+    "jazakallah": "semoga Allah membalas kebaikanmu",
+    "jazakallahu khair": "semoga Allah membalas kebaikanmu",
+    "barakallah": "semoga Allah memberkahimu",
+    "sholat": "salat",
+    "solat": "salat",
+    "shalat": "salat",
+    "bukpus": "buka puasa",
+    "bukber": "buka bersama",
+    "ramadhan": "ramadan",
+    "idulfitri": "idul fitri",
+    "eid": "lebaran",
+    "fitrah": "zakat fitrah",
+    "msdjd": "masjid",
+    "mushola": "musala",
+    "musholla": "musala",
+    "psntrn": "pesantren",
+    "ustadz": "ustaz",
+    "ustad": "ustaz",
+    "kyai": "kiai",
+    # National / cultural
+    "bhineka": "bhinneka",
+    "17an": "tujuh belasan",
+    "hut ri": "hari ulang tahun republik indonesia",
+    "hutri": "hari ulang tahun republik indonesia",
+    # Events / celebrations
+    "ultah": "ulang tahun",
+    "ulangtahun": "ulang tahun",
+    "ultahku": "ulang tahunku",
+    "happy birthday": "selamat ulang tahun",
+    "hbd": "selamat ulang tahun",
+    "hbday": "selamat ulang tahun",
+    "met ultah": "selamat ulang tahun",
+    "met ulth": "selamat ulang tahun",
+    "wedding": "pernikahan",
+    "nikah": "menikah",
+    "nikahin": "menikahkan",
+    "tunang": "tunangan",
+    "akad": "akad nikah",
+    "undngn": "undangan",
+    "kondangan": "menghadiri pernikahan",
+    "kondngan": "kondangan",
+    "ngaji": "mengaji",
+}
+
+# Education extended vocabulary
+_EDUCATION_EXTENDED: dict[str, str] = {
+    # School levels
+    "sd": "sekolah dasar",
+    "smp": "sekolah menengah pertama",
+    "sma": "sekolah menengah atas",
+    "smk": "sekolah menengah kejuruan",
+    "univ": "universitas",
+    "univrsitas": "universitas",
+    "kampus": "kampus",
+    "kuliah": "kuliah",
+    "mhsw": "mahasiswa",
+    "dsn": "dosen",
+    "gr": "guru",
+    "kepsek": "kepala sekolah",
+    "kpsek": "kepala sekolah",
+    # Academic activities
+    "bljar": "belajar",
+    "bljr": "belajar",
+    "ngajar": "mengajar",
+    "ngjar": "mengajar",
+    "ulngn": "ulangan",
+    "tgs": "tugas",
+    "pr": "pekerjaan rumah",
+    "homework": "pekerjaan rumah",
+    "hw": "pekerjaan rumah",
+    "presntsi": "presentasi",
+    "prsnts": "presentasi",
+    "laporan": "laporan",
+    "lprn": "laporan",
+    "mklh": "makalah",
+    "skrpsi": "skripsi",
+    "thesis": "tesis",
+    "wsuda": "wisuda",
+    "wsda": "wisuda",
+    "wsd": "wisuda",
+    "yudisium": "yudisium",
+    "ospek": "orientasi studi",
+    "maba": "mahasiswa baru",
+    "mhs baru": "mahasiswa baru",
+    # Grades & results
+    "nilai": "nilai",
+    "nlai": "nilai",
+    "rapor": "rapor",
+    "raport": "rapor",
+    "ipk": "indeks prestasi kumulatif",
+    "ip": "indeks prestasi",
+    "llus": "lulus",
+    "lls": "lulus",
+    "ggal": "gagal",
+    "ggl": "gagal",
+    "remedial": "remedial",
+    "remidi": "remedial",
+    "ngulang": "mengulang",
+    "nunggak": "tidak naik kelas",
+    "naikin": "menaikkan",
+    "beasiwa": "beasiswa",
+    "bsw": "beasiswa",
+    "beasiswaku": "beasiswaku",
+    "daftar": "daftar",
+    "dft": "daftar",
+    "pendaftran": "pendaftaran",
+    "pendftran": "pendaftaran",
+    "pndftran": "pendaftaran",
+    "sleksi": "seleksi",
+    "slksi": "seleksi",
+    "snbt": "seleksi nasional berdasarkan tes",
+    "utbk": "ujian tulis berbasis komputer",
+    "sbmptn": "seleksi bersama masuk perguruan tinggi negeri",
+    "ppdb": "penerimaan peserta didik baru",
+    "bimbel": "bimbingan belajar",
+    "bmbl": "bimbingan belajar",
+    "tryout": "uji coba",
+    "to": "uji coba",
+}
+
+# Work & office vocabulary
+_WORK_OFFICE: dict[str, str] = {
+    # Job types & positions
+    "krja": "kerja",
+    "kerjaan": "pekerjaan",
+    "krjaan": "pekerjaan",
+    "kntr": "kantor",
+    "prusahaan": "perusahaan",
+    "prshn": "perusahaan",
+    "bos": "atasan",
+    "boss": "atasan",
+    "rekan": "rekan kerja",
+    "krywn": "karyawan",
+    "pgwai": "pegawai",
+    "pns": "pegawai negeri sipil",
+    "asn": "aparatur sipil negara",
+    "honorer": "tenaga honorer",
+    "freelance": "pekerja lepas",
+    "frilens": "pekerja lepas",
+    "wfh": "kerja dari rumah",
+    "wfo": "kerja dari kantor",
+    "wfa": "kerja dari mana saja",
+    "remote": "kerja jarak jauh",
+    "hybrid": "kerja hibrida",
+    "intern": "magang",
+    "mgang": "magang",
+    "kntrak": "kontrak",
+    "resign": "mengundurkan diri",
+    "resignasi": "pengunduran diri",
+    "phk": "pemutusan hubungan kerja",
+    # Work activities
+    "rpt": "rapat",
+    "meeting": "rapat",
+    "mtg": "rapat",
+    "zoom": "rapat video",
+    "gmeet": "google meet",
+    "deadline": "batas waktu",
+    "dl": "batas waktu",
+    "dl nya": "batas waktunya",
+    "submit": "kirimkan",
+    "sbmt": "kirimkan",
+    "rvisi": "revisi",
+    "approval": "persetujuan",
+    "approved": "disetujui",
+    "acc": "disetujui",
+    "reject": "ditolak",
+    "feedback": "masukan",
+    "ftdbk": "masukan",
+    "report": "laporan",
+    "update": "perbarui",
+    "followup": "tindak lanjut",
+    "fu": "tindak lanjut",
+    "asap": "sesegera mungkin",
+    "urgent": "mendesak",
+    "priority": "prioritas",
+    "prio": "prioritas",
+    "kpi": "indikator kinerja utama",
+    "okr": "tujuan dan hasil kunci",
+    "brainstorm": "curah gagasan",
+    "brnstrm": "curah gagasan",
+    # Finance / salary
+    "gji": "gaji",
+    "gajinya": "gajinya",
+    "thr": "tunjangan hari raya",
+    "lmbr": "lembur",
+    "overtime": "lembur",
+    "ot": "lembur",
+    "ct": "cuti",
+    "check in": "masuk kerja",
+    "checkin": "masuk kerja",
+    "check out": "keluar kerja",
+    "checkout": "keluar kerja",
+    # Tools / digital work
+    "email": "surel",
+    "mailing": "mengirim surel",
+    "cc": "tembusan",
+    "bcc": "tembusan tersembunyi",
+    "forward": "teruskan",
+    "teams": "microsoft teams",
+    "ms teams": "microsoft teams",
+    "excel": "microsoft excel",
+    "word": "microsoft word",
+    "ppt": "powerpoint",
+    "powerpoint": "presentasi powerpoint",
+    "drive": "google drive",
+    "cloud": "komputasi awan",
+    "database": "basis data",
+    "app": "aplikasi",
+    "apk": "aplikasi",
+    "software": "perangkat lunak",
+    "hardware": "perangkat keras",
+}
+
+# Numbers, measurements & quantity expressions
+_NUMBERS_QUANTITY: dict[str, str] = {
+    # Informal prefix shortcuts  (sejuta / semiliar are genuine contractions)
+    "sejuta": "satu juta",
+    "semiliar": "satu miliar",
+    # Informal number abbreviations  (rb = ribu, jt = juta)
+    "1rb": "seribu",
+    "2rb": "dua ribu",
+    "5rb": "lima ribu",
+    "10rb": "sepuluh ribu",
+    "20rb": "dua puluh ribu",
+    "50rb": "lima puluh ribu",
+    "100rb": "seratus ribu",
+    "200rb": "dua ratus ribu",
+    "500rb": "lima ratus ribu",
+    "1jt": "satu juta",
+    "2jt": "dua juta",
+    "5jt": "lima juta",
+    "10jt": "sepuluh juta",
+    # Quantity abbreviations  (slang → standard)
+    "slrh": "seluruh",
+    "sktr": "sekitar",
+    "hmpr": "hampir",
+    "hmpir": "hampir",
+    "plng2": "paling-paling",
+    # Reduplications  (written as X2 or concatenated)
+    "rata2": "rata-rata",
+    "ratarata": "rata-rata",
+    "kira2": "kira-kira",
+    "kirakira": "kira-kira",
+    # Time quantities
+    "sebentar": "sebentar",
+    "sbntr": "sebentar",
+    "sbntr lg": "sebentar lagi",
+    "slmanya": "selamanya",
+    "smntara": "sementara",
+    "mndak": "mendadak",
+    "dadakan": "mendadak",
+    "ssgera": "segera",
+    "sgera": "segera",
+    "sgra": "segera",
+    "cpat": "cepat",
+}
+
+# Common compound slang expressions & phrases
+# Only include entries where the COMBINED phrase produces a different result
+# than individual word lookups would give (e.g. slang fusion, negation compounds
+# with particles, or abbreviations that span multiple tokens).
+_COMPOUND_EXPRESSIONS: dict[str, str] = {
+    # Slang word + particle/qualifier — transformation differs from individual lookups
+    "eh iya": "oh iya",  # discourse opener
+    "gitu kan": "begitu kan",
+    "gitu dong": "begitu dong",
+    "beneran ga": "sungguh tidak",
+    "beneran nih": "sungguh ini",
+    "srius": "serius",  # typo of standard word
+    "serius nih": "serius ini",
+    # Question compounds with negation particle
+    "bisa ga": "bisa tidak",
+    "bisa gak": "bisa tidak",
+    "boleh ga": "boleh tidak",
+    "boleh gak": "boleh tidak",
+    "mau ga": "mau tidak",
+    "mau gak": "mau tidak",
+    "tau ga": "tahu tidak",
+    "tau gak": "tahu tidak",
+    # Multi-token slang phrases
+    "gimana caranya": "bagaimana caranya",
+    "gmn caranya": "bagaimana caranya",
+    "gue juga": "saya juga",
+    "gw juga": "saya juga",
+    "kayaknya sih": "sepertinya",
+    "kayaknya iya": "sepertinya iya",
+    "buat apa": "untuk apa",
+    "bwt apa": "untuk apa",
+    "ngapain": "sedang apa",
+    "ngapain aja": "sedang apa saja",
+    "lagi ngapain": "sedang apa",
+    "lagi dimana": "sedang di mana",
+    "lg dmn": "sedang di mana",
+    "otw ke": "dalam perjalanan ke",
+    "nyampe blm": "sudah sampai belum",
+    "sampai blm": "sudah sampai belum",
+    # Compound slang where slang token fuses with standard word
+    "udah mkan": "sudah makan",  # mkan not split correctly if fused
+    "blm makan": "belum makan",
+    "udh lma": "sudah lama",
+    "br aja": "baru saja",
+    "bsk aja": "besok saja",
+    "skrg juga": "sekarang juga",
+    "dr tadi": "dari tadi",
+    "dr dulu": "dari dulu",
+    "tlg bantu": "tolong bantu",
+    # Agreement / disagreement — particle changes meaning of phrase
+    "emang bener": "memang benar",
+    "bener juga": "benar juga",
+    "iya bener": "iya benar",
+    "nggak dong": "tidak dong",
+    "nggak sih": "tidak sih",
+    "ya ga": "ya tidak",
+    "ga juga": "tidak juga",
+    "gak juga": "tidak juga",
+    "bukan gitu": "bukan begitu",
+    "bukan disini": "bukan di sini",
+    "maksudnya gimana": "maksudnya bagaimana",
+    "mksdnya gmn": "maksudnya bagaimana",
+    # Temporal compound abbreviations
+    "kapan2": "kapan-kapan",
+    "kapan2 ya": "kapan-kapan ya",
+    "ntar aja": "nanti saja",
+    "tar aja": "nanti saja",
+    "makasih banyak": "terima kasih banyak",
+}
+
+# ─────────────────────────────────────────────────────────────────────────────
+# PARTICLES
+# Indonesian sentence-final / clause-final pragmatic particles.
+# The 6 canonical forms (lah, deh, nih, dong, kok, sih) are pass-through mapped
+# to prevent them from being stripped. Other particles rely on auto-reduction.
+#
+# Design: _reduce_repeated_chars automatically collapses:
+#   - vowels: 2+ consecutive → 1  (laah → lah, yaa → ya)
+#   - consonants: 3+ consecutive → 1 (lahhhh → lah, yukkk → yuk)
+# Note: 2-consonant repeats (lahh, dehh) are explicitly mapped here as they
+# are not auto-reduced.
+# ─────────────────────────────────────────────────────────────────────────────
+_PARTICLES: dict[str, str] = {
+    # ── lah ──
+    "lah": "lah",
+    "lh": "lah",
+    "lahh": "lah",
+    "la": "lah",
+    # ── deh ──
+    "deh": "deh",
+    "dh": "deh",
+    "dehh": "deh",
+    "de": "deh",
+    # ── nih ──
+    "nih": "nih",
+    "ni": "nih",
+    "nihh": "nih",
+    "neh": "nih",  # Betawi variant
+    "nehh": "nih",
+    # ── dong ──
+    "dong": "dong",
+    "dng": "dong",
+    "dongg": "dong",
+    "don": "dong",
+    # ── kok ──
+    "kok": "kok",
+    "kokk": "kok",
+    # ── sih ──
+    "sih": "sih",
+    "si": "sih",
+    "sihh": "sih",
+    "sich": "sih",
+    # ── Other Particle Variants ──
+    "loh": "lho",
+    "lo": "lho",
+    "punn": "pun",
+    "tohh": "toh",
+    "kahh": "kah",
+    "tahh": "tah",
+    "woy": "woi",
+    "woyy": "woi",
+    "hey": "hei",
+    "yukk": "yuk",
+    "yuk ah": "yuk",
+    "ahh": "ah",
+    "ihh": "ih",
+    "idihh": "idih",
+}
+
+# ─────────────────────────────────────────────────────────────────────────────
+# MERGE ALL CATEGORIES INTO ONE MAPPING
+# Order matters: later dicts override earlier ones on key conflicts.
+# Compound negation must appear so that it sorts BEFORE single-char negation
+# (handled automatically by longest-first sort in _compile_pattern).
+# ─────────────────────────────────────────────────────────────────────────────
+_DEFAULT_SLANG: dict[str, str] = {}
+for _cat in (
+    _PRONOUNS,
+    _KINSHIP,
+    _NEGATION,
+    _COMPOUND_NEGATION,
+    _CONJUNCTIONS,
+    _VERBS,
+    _ADJECTIVES_ADVERBS,
+    _QUESTION_WORDS,
+    _GREETINGS_RESPONSES,
+    _TEMPORAL_LOCATION,
+    _INTERNET_SLANG,
+    _ECOMMERCE_FINANCE,
+    _YOUTH_SLANG,
+    _DISCOURSE_MARKERS,
+    _JAVANESE_EXTENDED,
+    _SUNDANESE_EXTENDED,
+    _NOUNS,
+    _HEALTH,
+    _EMOTIONS_EXPRESSIONS,
+    _FOOD_DRINK,
+    _CLOTHING_FASHION,
+    _TRANSPORTATION,
+    _RELIGION_CULTURE,
+    _EDUCATION_EXTENDED,
+    _WORK_OFFICE,
+    _NUMBERS_QUANTITY,
+    _COMPOUND_EXPRESSIONS,
+    _PARTICLES,
+):
+    _DEFAULT_SLANG.update(_cat)
+
+
+# ─────────────────────────────────────────────────────────────────────────────
+# REPEATED CHARACTER NORMALIZATION PATTERNS
+# ─────────────────────────────────────────────────────────────────────────────
+
+# Vowels: reduce 2+ consecutive identical vowels → 1
+# e.g. "manaa" → "mana", "bangeeeet" → "banget"
+# Rationale: vowel elongation is always expressive, never meaningful.
+_VOWEL_REPEAT = re.compile(r"([aeiouAEIOU])\1+")
+
+# Strict vowel reduction: only collapse 3+ consecutive identical vowels.
+# Used in the global fallback pass so that valid standard Indonesian spellings
+# containing double vowels (e.g. "pekerjaan", "maaf", "laal") are NOT touched,
+# while obvious elongation ("manaaaa", "beneeer") is still reduced.
+_VOWEL_REPEAT_STRICT = re.compile(r"([aeiouAEIOU])\1{2,}")
+
+# Consonants: reduce 3+ consecutive identical consonants → 1
+# e.g. "bgttt" → "bgt", "wkwwwk" → "wkwk"
+# Rationale: keep "kk" (kakak), "ll" (dalam "dll"), "mm" etc. intact;
+#            only collapse truly elongated forms (3+).
+# Note: [^\W\s\d] matches letter-only consonants, excluding punctuation
+# so that "!!!" is NOT reduced here (that is handled by normalize_punctuation).
+_CONSONANT_REPEAT = re.compile(r"([^\W\s\daeiouAEIOU])\1{2,}")
+
+# Whitespace normalization
+_WHITESPACE = re.compile(r"\s+")
+
+
+# ─────────────────────────────────────────────────────────────────────────────
+# MAIN CLASS
+# ─────────────────────────────────────────────────────────────────────────────
+
+
+class SlangNormalizer:
+    """
+    Normalizes informal/colloquial Indonesian text to standard Indonesian.
+
+    Args:
+        custom_mapping: Optional additional slang entries that override or
+                        extend the built-in dictionary.
+
+    Example:
+        >>> normalizer = SlangNormalizer()
+        >>> normalizer.normalize("gw gamau pergi krn lg baper bgt")
+        'saya tidak mau pergi karena sedang bawa perasaan banget'
+
+        >>> normalizer.add("npwp", "nomor pokok wajib pajak")
+        >>> normalizer.normalize("masukin npwp dulu")
+        'masukkan nomor pokok wajib pajak dulu'
+
+        >>> normalizer.normalize_batch(["gw mau makan", "km udah makan?"])
+        ['saya mau makan', 'kamu sudah makan?']
+    """
+
+    def __init__(self, custom_mapping: dict[str, str] | None = None):
+        self.mapping: dict[str, str] = {k.lower(): v for k, v in _DEFAULT_SLANG.items()}
+        if custom_mapping:
+            self.mapping.update({k.lower(): v for k, v in custom_mapping.items()})
+        self._regex: re.Pattern | None = None
+        self._compile_pattern()
+
+    # ─── Internal ────────────────────────────────────────────────────────────
+
+    def _compile_pattern(self) -> None:
+        """
+        Build a compiled regex from the current mapping.
+
+        Keys are sorted longest-first so compound slang like "gamau"
+        matches before its shorter substring "ga".
+        """
+        if not self.mapping:
+            self._regex = None
+            return
+        sorted_keys = sorted(self.mapping.keys(), key=len, reverse=True)
+        escaped = [re.escape(k) for k in sorted_keys]
+        pattern = r"\b(" + "|".join(escaped) + r")\b"
+        self._regex = re.compile(pattern, re.IGNORECASE)
+
+    @staticmethod
+    def _reduce_repeated_chars(text: str) -> str:
+        """
+        Collapse elongated characters to their canonical form.
+
+        Rules:
+            - Vowels (a, e, i, o, u): 2+ consecutive → 1
+              "manaaaa" → "mana",  "bangeeeet" → "banget"
+            - Consonants: 3+ consecutive → 1
+              "bgttttt" → "bgt",   "wkwwwwk" → "wkwk"
+              "kk" → "kk"  (preserved — abbreviation for "kakak")
+              "dll" → "dll" (preserved — "dan lain-lain")
+
+        Vowels are processed first so that a string like "aaakk" is handled
+        correctly in two distinct passes without cross-contamination.
+        """
+        text = _VOWEL_REPEAT.sub(r"\1", text)
+        text = _CONSONANT_REPEAT.sub(r"\1", text)
+        return text
+
+    @staticmethod
+    def _reduce_repeated_chars_strict(text: str) -> str:
+        """
+        Like _reduce_repeated_chars but uses a stricter threshold for vowels:
+        only collapses 3+ consecutive identical vowels (not 2+).
+
+        This is used as the global fallback pass inside normalize() so that
+        standard Indonesian words with legitimate double vowels (e.g.
+        "pekerjaan", "maaf") are not corrupted, while obvious elongation
+        ("manaaaa", "beneeer") is still collapsed.
+        """
+        text = _VOWEL_REPEAT_STRICT.sub(r"\1", text)
+        text = _CONSONANT_REPEAT.sub(r"\1", text)
+        return text
+
+    # ─── Public API ──────────────────────────────────────────────────────────
+
+    @staticmethod
+    def _mirror_case(original: str, replacement: str) -> str:
+        """
+        Mirror the casing pattern of ``original`` onto ``replacement``.
+
+        Rules (applied in priority order):
+            - All-uppercase original  → replacement.upper()
+            - Title-case original     → replacement.capitalize()
+            - Mixed / lowercase       → replacement unchanged (already lowercase)
+
+        This lets ``normalize(lowercase=False)`` preserve casing intent
+        even after slang substitution.
+
+        Examples:
+            >>> SlangNormalizer._mirror_case("GW", "saya")
+            'SAYA'
+            >>> SlangNormalizer._mirror_case("Gw", "saya")
+            'Saya'
+            >>> SlangNormalizer._mirror_case("gw", "saya")
+            'saya'
+        """
+        if original.isupper():
+            return replacement.upper()
+        if original[0].isupper():
+            return replacement.capitalize()
+        return replacement
+
+    def normalize(
+        self,
+        text: str,
+        normalize_whitespace: bool = True,
+        lowercase: bool = True,
+    ) -> str:
+        """
+        Normalize a single text string.
+
+        Pipeline:
+            1. Slang replacement with two-pass per-token lookup
+            2. Global repeated-char reduction for out-of-vocab elongated words
+            3. Second regex pass for words newly recognisable after reduction
+               (e.g. "gwwww" → step2 → "gw" → step3 → "saya")
+            4. (Optionally) collapse multiple whitespace into one
+
+        Args:
+            text: Input text string.
+            normalize_whitespace: If True, collapse multi-space into single
+                                  space and strip leading/trailing whitespace.
+            lowercase: If True (default), replacement values are returned as-is
+                       (all lowercase). If False, the casing pattern of the
+                       original matched token is mirrored onto the replacement
+                       (ALL-CAPS → upper, Title → capitalize, mixed → as-is).
+
+        Returns:
+            Normalized text string.
+        """
+        if not text or not text.strip():
+            return text
+
+        # Step 1: Slang replacement with two-pass per-token lookup.
+        # The regex matches candidates from the slang dict; for each match:
+        #   Pass A – lookup the original token (handles "krjaan" which has a
+        #            meaningful double-a that char-reduction would destroy).
+        #   Pass B – reduce repeated chars first, then lookup (handles elongated
+        #            forms like "gkkkk" → "gk" → "tidak").
+        #   Fallback – return the char-reduced form so elongation is still
+        #              collapsed even when no dict entry exists.
+        if self._regex:
+
+            def _replace(match: re.Match) -> str:
+                original = match.group(1)
+                word = original.lower()
+
+                # Pass A: direct lookup (preserve meaningful repeated vowels)
+                if word in self.mapping:
+                    result = self.mapping[word]
+                    if not lowercase:
+                        return self._mirror_case(original, result)
+                    return result
+
+                # Pass B: reduce chars first, then lookup
+                reduced = self._reduce_repeated_chars(word)
+                if reduced in self.mapping:
+                    result = self.mapping[reduced]
+                    if not lowercase:
+                        return self._mirror_case(original, result)
+                    return result
+
+                # Fallback: return char-reduced form (preserves original case)
+                return self._reduce_repeated_chars(original)
+
+            text = self._regex.sub(_replace, text)
+
+        # Step 2: Global char reduction (strict mode) for words not captured
+        # by the dict regex, e.g. out-of-vocab elongated words ("manaaaa" ->
+        # "mana") or words whose elongated form isn't in the regex ("gwwww" ->
+        # "gw").
+        # STRICT: only collapses 3+ consecutive identical vowels so that valid
+        # standard Indonesian double vowels in replacement outputs (e.g.
+        # "pekerjaan") are NOT corrupted.
+        text = self._reduce_repeated_chars_strict(text)
+
+        # Step 3: Second regex pass — catches words that only became recognisable
+        # after Step 2 (e.g. "gwwww" -> "gw" -> "saya", "gkkkk" -> "gk" -> "tidak").
+        if self._regex:
+            text = self._regex.sub(_replace, text)
+
+        # Step 4: Normalize whitespace
+        if normalize_whitespace:
+            text = _WHITESPACE.sub(" ", text).strip()
+
+        return text
+
+    def normalize_batch(
+        self,
+        texts: list[str],
+        normalize_whitespace: bool = True,
+        lowercase: bool = True,
+    ) -> list[str]:
+        """
+        Normalize a list of text strings efficiently.
+
+        Args:
+            texts: List of input strings.
+            normalize_whitespace: Passed through to normalize().
+            lowercase: Passed through to normalize().
+
+        Returns:
+            List of normalized strings (same order as input).
+
+        Example:
+            >>> slang.normalize_batch(["gw udah makan", "km blm tidur?"])
+            ['saya sudah makan', 'kamu belum tidur?']
+        """
+        return [self.normalize(t, normalize_whitespace, lowercase) for t in texts]
+
+    def add(self, word: str, replacement: str) -> None:
+        """
+        Add a single entry to the slang dictionary.
+
+        Triggers regex recompilation. For adding many entries at once,
+        prefer bulk_add() to avoid repeated recompilation.
+
+        Args:
+            word: Slang word (stored as lowercase).
+            replacement: Standard replacement string.
+
+        Example:
+            >>> normalizer.add("npwp", "nomor pokok wajib pajak")
+        """
+        self.mapping[word.lower()] = replacement
+        self._compile_pattern()
+
+    def bulk_add(self, mapping: dict[str, str]) -> None:
+        """
+        Add multiple entries at once with a single regex recompilation.
+
+        More efficient than calling add() repeatedly when inserting
+        many domain-specific terms.
+
+        Args:
+            mapping: Dict of {slang: replacement} pairs.
+
+        Example:
+            >>> normalizer.bulk_add({
+            ...     "kk": "kakak",
+            ...     "adek": "adik",
+            ...     "bokap": "ayah",
+            ...     "nyokap": "ibu",
+            ... })
+        """
+        self.mapping.update({k.lower(): v for k, v in mapping.items()})
+        self._compile_pattern()
+
+    def remove(self, word: str) -> bool:
+        """
+        Remove a word from the dictionary.
+
+        Args:
+            word: Slang word to remove.
+
+        Returns:
+            True if the word was found and removed, False otherwise.
+        """
+        word_lower = word.lower()
+        if word_lower in self.mapping:
+            del self.mapping[word_lower]
+            self._compile_pattern()
+            return True
+        return False
+
+    def reset(self) -> None:
+        """
+        Reset dictionary to built-in defaults, discarding all custom changes.
+
+        Example:
+            >>> normalizer.add("x", "ekstra")
+            >>> "x" in normalizer
+            True
+            >>> normalizer.reset()
+            >>> "x" in normalizer
+            False
+        """
+        self.mapping = {k.lower(): v for k, v in _DEFAULT_SLANG.items()}
+        self._compile_pattern()
+
+    def export(self) -> dict[str, str]:
+        """
+        Export the current mapping as a plain dictionary.
+
+        Returns:
+            A copy of the current {slang: replacement} mapping.
+        """
+        return self.mapping.copy()
+
+    def supported_words(self) -> dict[str, str]:
+        """
+        Return all supported slang words and their standard replacements.
+
+        Useful for debugging, documentation, or building custom UIs.
+
+        Returns:
+            A copy of the current {slang: standard} mapping, sorted
+            alphabetically by slang key.
+
+        Example:
+            >>> from basa import slang
+            >>> words = slang.supported_words()
+            >>> words["gw"]
+            'saya'
+            >>> len(slang)
+            120
+        """
+        return dict(sorted(self.mapping.items()))
+
+    def lookup(self, word: str) -> str | None:
+        """
+        Look up a single slang word without normalizing a full sentence.
+
+        Args:
+            word: Word to look up (case-insensitive).
+
+        Returns:
+            The replacement string, or None if not found.
+
+        Example:
+            >>> slang.lookup("bgt")
+            'banget'
+            >>> slang.lookup("makan")
+            None
+        """
+        return self.mapping.get(word.lower())
+
+    # ─── Dunder helpers ──────────────────────────────────────────────────────
+
+    def __len__(self) -> int:
+        """Return number of entries in the slang dictionary."""
+        return len(self.mapping)
+
+    def __contains__(self, word: str) -> bool:
+        """Check if a word is in the slang dictionary (case-insensitive)."""
+        return word.lower() in self.mapping
+
+    def __repr__(self) -> str:
+        return f"SlangNormalizer(entries={len(self.mapping)})"
+
+
+# Backward-compatible alias
+SlangDictionary = SlangNormalizer
+
+# ─────────────────────────────────────────────────────────────────────────────
+# MODULE-LEVEL CONVENIENCE SINGLETON
+#
+# Import and use directly without instantiation:
+#   from basa.core.slang import slang
+#   slang.normalize("gw gamau pergi")
+# ─────────────────────────────────────────────────────────────────────────────
+slang = SlangNormalizer()