phoonnx 0.1.0a3__tar.gz → 0.2.0__tar.gz

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: phoonnx
-Version: 0.1.0a3
+Version: 0.2.0
 Home-page: https://github.com/TigreGotico/phoonnx
 Author: JarbasAi
 Author-email: jarbasai@mailfence.com
@@ -41,6 +41,7 @@ Provides-Extra: jam
 Provides-Extra: hsn
 Provides-Extra: fil
 Provides-Extra: ml
+Provides-Extra: mwl
 Provides-Extra: lez
 Provides-Extra: si
 Provides-Extra: kk

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/README.md

@@ -89,6 +89,7 @@ wrap, and the output alphabets they can generate.
 | **Multilingual**    | `MisakiPhonemizer`     | [misaki](https://github.com/hexgrad/misaki)                                                                        | IPA                          | Misaki is a G2P engine designed for Kokoro models.                                                                                          |
 | **Multilingual**    | `TransphonePhonemizer` | [transphone](https://github.com/xinjli/transphone)                                                                 | IPA                          | It provides approximated phoneme tokenizers and G2P model for 7546 languages registered in the Glottolog database.                          |
 | **Multilingual**    | `EpitranPhonemizer`    | [epitran](https://github.com/dmort27/epitran)                                                                      | IPA                          | A tool for transcribing orthographic text as IPA                                                                                            |
+| **Mirandese (mwl)** | `MirandesePhonemizer`  | [mwl_phonemizer](https://github.com/TigreGotico/mwl_phonemizer)                                                                 | IPA                          | A tool for transcribing orthographic text as IPA                                                                                            |
 | **Arabic (ar)**     | `MantoqPhonemizer`     | [mantoq](https://github.com/mush42/mantoq)                                                                         | BUCKWALTER, IPA              | Translates unvoweled Arabic to phonemes, with optional conversion to IPA.                                                                   |
 | **Chinese (zh)**    | `JiebaPhonemizer`      | [jieba](https://github.com/fxsjy/jieba)                                                                            | HANZI                        | Segments Chinese text into words with spaces; useful for pre-processing.                                                                    |
 | **Chinese (zh)**    | `G2pMPhonemizer`       | [g2pC](https://github.com/Kyubyong/g2pC)                                                                           | IPA, Pinyin                  | CRF-based Grapheme-to-Phoneme converter                                                                                                     |

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx/config.py

@@ -45,9 +45,12 @@ class PhonemeType(str, Enum):
     MISAKI = "misaki"
     ESPEAK = "espeak"
     GRUUT = "gruut"
+    GORUUT = "goruut"
     EPITRAN = "epitran"
     BYT5 = "byt5"
     CHARSIU = "charsiu"  # technically same as byt5, but needs special handling for whitespace
+    TRANSPHONE = "transphone"
+    MIRANDESE = "mwl_phonemizer"
 
     DEEPPHONEMIZER = "deepphonemizer" # en
     OPENPHONEMIZER = "openphonemizer" # en
@@ -392,6 +395,7 @@ def get_phonemizer(phoneme_type: PhonemeType,
     from phoonnx.phonemizers import (EpitranPhonemizer, EspeakPhonemizer, OpenPhonemizer, OpenJTaklPhonemizer,
                                      ByT5Phonemizer, CharsiuPhonemizer, DeepPhonemizer, PersianPhonemizer,
                                      G2pCPhonemizer, G2pMPhonemizer, G2PKPhonemizer, G2PEnPhonemizer,
+                                     TransphonePhonemizer, MirandesePhonemizer, GoruutPhonemizer,
                                      GruutPhonemizer, GraphemePhonemizer, MantoqPhonemizer, MisakiPhonemizer,
                                      KoG2PPhonemizer, PypinyinPhonemizer, PyKakasiPhonemizer, CotoviaPhonemizer,
                                      CutletPhonemizer, PhonikudPhonemizer, VIPhonemePhonemizer, XpinyinPhonemizer,
@@ -404,10 +408,16 @@ def get_phonemizer(phoneme_type: PhonemeType,
         phonemizer = CharsiuPhonemizer(model)
     elif phoneme_type == PhonemeType.GRUUT:
         phonemizer = GruutPhonemizer()
+    elif phoneme_type == PhonemeType.GORUUT:
+        phonemizer = GoruutPhonemizer()
     elif phoneme_type == PhonemeType.EPITRAN:
         phonemizer = EpitranPhonemizer()
     elif phoneme_type == PhonemeType.MISAKI:
         phonemizer = MisakiPhonemizer()
+    elif phoneme_type == PhonemeType.TRANSPHONE:
+        phonemizer = TransphonePhonemizer()
+    elif phoneme_type == PhonemeType.MIRANDESE:
+        phonemizer = MirandesePhonemizer()
     elif phoneme_type == PhonemeType.DEEPPHONEMIZER:
         phonemizer = DeepPhonemizer(model)
     elif phoneme_type == PhonemeType.OPENPHONEMIZER:

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx/phonemizers/__init__.py

@@ -11,8 +11,9 @@ from phoonnx.phonemizers.ja import PyKakasiPhonemizer, CutletPhonemizer, OpenJTa
 from phoonnx.phonemizers.ko import KoG2PPhonemizer, G2PKPhonemizer
 from phoonnx.phonemizers.zh import (G2pCPhonemizer, G2pMPhonemizer, PypinyinPhonemizer,
                                     XpinyinPhonemizer, JiebaPhonemizer)
-from phoonnx.phonemizers.mul import (EspeakPhonemizer, EpitranPhonemizer, MisakiPhonemizer,
-                                     GruutPhonemizer, ByT5Phonemizer, CharsiuPhonemizer)
+from phoonnx.phonemizers.mul import (EspeakPhonemizer, EpitranPhonemizer, MisakiPhonemizer, GoruutPhonemizer,
+                                     GruutPhonemizer, ByT5Phonemizer, CharsiuPhonemizer, TransphonePhonemizer)
+from phoonnx.phonemizers.mwl import MirandesePhonemizer
 
 Phonemizer = Union[
     MisakiPhonemizer,
@@ -21,7 +22,10 @@ Phonemizer = Union[
     CharsiuPhonemizer,
     EspeakPhonemizer,
     GruutPhonemizer,
+    GoruutPhonemizer,
     EpitranPhonemizer,
+    TransphonePhonemizer,
+    MirandesePhonemizer,
     OpenJTaklPhonemizer,
     CutletPhonemizer,
     PyKakasiPhonemizer,

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx/phonemizers/mul.py

@@ -436,6 +436,187 @@ class GruutPhonemizer(BasePhonemizer):
         return pho.strip()
 
 
+class GoruutPhonemizer(BasePhonemizer):
+    """
+    A phonemizer class that uses the pygoruut library to convert text into phonemes.
+    https://github.com/neurlang/pygoruut/
+    """
+    GORUUT_LANGS_NON_STD = [
+        'BengaliDhaka', 'BengaliRahr', 'MalayArab', 'VietnameseCentral', 'VietnameseSouthern',
+        'EnglishAmerican', 'EnglishBritish', 'NahuatlClassical', 'Hebrew2', 'Hebrew3',
+        'MinnanTawianese', 'MinnanHokkien', 'MinnanTawianese2', 'MinnanHokkien2']
+    ISO639 = {
+        "af": "Afrikaans",
+        "am": "Amharic",
+        "ar": "Arabic",
+        "az": "Azerbaijani",
+        "be": "Belarusian",
+        "bn": "Bengali",
+        "my": "Burmese",
+        "ceb": "Cebuano",
+        "ce": "Chechen",
+        "zh": "ChineseMandarin",
+        "cs": "Czech",
+        "da": "Danish",
+        "nl": "Dutch",
+        "dz": "Dzongkha",
+        "en": "English",
+        "eo": "Esperanto",
+        "fa": "Farsi",
+        "fi": "Finnish",
+        "fr": "French",
+        "de": "German",
+        "el": "Greek",
+        "gu": "Gujarati",
+        "ha": "Hausa",
+        "he": "Hebrew",
+        "hi": "Hindi",
+        "hu": "Hungarian",
+        "is": "Icelandic",
+        "id": "Indonesian",
+        "tts": "Isan",
+        "it": "Italian",
+        "jam": "Jamaican",
+        "ja": "Japanese",
+        "jv": "Javanese",
+        "kk": "Kazakh",
+        "ko": "Korean",
+        "lb": "Luxembourgish",
+        "mk": "Macedonian",
+        "ml": "Malayalam",
+        "ms": "MalayLatin",
+        "mt": "Maltese",
+        "mr": "Marathi",
+        "mn": "Mongolian",
+        "ne": "Nepali",
+        "no": "Norwegian",
+        "ps": "Pashto",
+        "pl": "Polish",
+        "pt": "Portuguese",
+        "pa": "Punjabi",
+        "ro": "Romanian",
+        "ru": "Russian",
+        "sk": "Slovak",
+        "es": "Spanish",
+        "sw": "Swahili",
+        "sv": "Swedish",
+        "ta": "Tamil",
+        "te": "Telugu",
+        "th": "Thai",
+        "bo": "Tibetan",
+        "tr": "Turkish",
+        "uk": "Ukrainian",
+        "ur": "Urdu",
+        "ug": "Uyghur",
+        "vi": "VietnameseNorthern",
+        "zu": "Zulu",
+        "hy": "Armenian",
+        "eu": "Basque",
+        "bg": "Bulgarian",
+        "ca": "Catalan",
+        "ny": "Chichewa",
+        "hr": "Croatian",
+        "et": "Estonian",
+        "gl": "Galician",
+        "ka": "Georgian",
+        "km": "KhmerCentral",
+        "lo": "Lao",
+        "lv": "Latvian",
+        "lt": "Lithuanian",
+        "sr": "Serbian",
+        "tl": "Tagalog",
+        "yo": "Yoruba",
+        "sq": "Albanian",
+        "an": "Aragonese",
+        "as": "Assamese",
+        "ba": "Bashkir",
+        "bpy": "BishnupriyaManipuri",
+        "bs": "Bosnian",
+        "chr": "Cherokee",
+        "cu": "Chuvash",
+        "gla": "GaelicScottish",
+        "gle": "GaelicIrish",
+        "kl": "Greenlandic",
+        "gn": "Guarani",
+        "ht": "HaitianCreole",
+        "haw": "Hawaiian",
+        "io": "Ido",
+        "ia": "Interlingua",
+        "kn": "Kannada",
+        "quc": "Kiche",
+        "kok": "Konkani",
+        "ku": "Kurdish",
+        "ky": "Kyrgyz",
+        "qdb": "LangBelta",
+        "ltg": "Latgalian",
+        "la": "LatinClassical",
+        "lat": "LatinEcclesiastical",
+        "lfn": "LinguaFrancaNova",
+        "jbo": "Lojban",
+        "smj": "LuleSaami",
+        "mi": "Maori",
+        "nah": "NahuatlCentral",
+        "nci": "NahuatlMecayapan",
+        "ncz": "NahuatlTetelcingo",
+        "nog": "Nogai",
+        "om": "Oromo",
+        "pap": "Papiamento",
+        "qu": "Quechua",
+        "qya": "Quenya",
+        "tn": "Setswana",
+        "shn": "ShanTaiYai",
+        "sjn": "Sindarin",
+        "sd": "Sindhi",
+        "si": "Sinhala",
+        "sl": "Slovenian",
+        "tt": "Tatar",
+        "tk": "Turkmen",
+        "uz": "Uzbek",
+        "cyw": "WelshNorth",
+        "cys": "WelshSouth",
+        "yue": "Cantonese"
+    }
+
+    def __init__(self, remote_url=None):
+        super().__init__(Alphabet.IPA)
+        from pygoruut.pygoruut import Pygoruut
+        from pygoruut.pygoruut_languages import PygoruutLanguages
+
+        self.pygoruut_langs = PygoruutLanguages()
+        if remote_url is not None:
+            # 'https://hashtron.cloud'
+            self.pygoruut = Pygoruut(api=remote_url)
+        else:
+            self.pygoruut = Pygoruut()
+
+    @classmethod
+    def get_lang(cls, target_lang: str) -> str:
+        """
+        Validates and returns the closest supported language code.
+
+        Args:
+            target_lang (str): The language code to validate.
+
+        Returns:
+            str: The validated language code.
+
+        Raises:
+            ValueError: If the language code is unsupported.
+        """
+        if target_lang in cls.GORUUT_LANGS_NON_STD:
+            return target_lang
+        if target_lang.lower() == "en-us":
+            return 'EnglishAmerican'
+        if target_lang.lower() == "en-gb" or target_lang.lower() == "en-uk":
+            return 'EnglishBritish'
+        lang = cls.match_lang(target_lang, list(cls.ISO639))
+        return cls.ISO639[lang]
+
+    def phonemize_string(self, text: str, lang: str) -> str:
+        lang = self.get_lang(lang)
+        return str(self.pygoruut.phonemize(language=lang, sentence=text))
+
+
 class EpitranPhonemizer(BasePhonemizer):
     """
     """
@@ -1178,6 +1359,7 @@ if __name__ == "__main__":
     byt5 = ByT5Phonemizer()
     espeak = EspeakPhonemizer()
     gruut = GruutPhonemizer()
+    goruut = GoruutPhonemizer(remote_url='https://hashtron.cloud')
     epitr = EpitranPhonemizer()
     charsiu = CharsiuPhonemizer()
     misaki = MisakiPhonemizer()
@@ -1194,6 +1376,7 @@ if __name__ == "__main__":
     phonemes1e = charsiu.phonemize(text1, lang)
     phonemes1f = misaki.phonemize(text1, lang)
     phonemes1g = tphone.phonemize(text1, lang)
+    phonemes1h = goruut.phonemize(text1, lang)
     print(f" Espeak         Phonemes: {phonemes1}")
     print(f" Gruut          Phonemes: {phonemes1b}")
     print(f" byt5           Phonemes: {phonemes1c}")
@@ -1201,6 +1384,7 @@ if __name__ == "__main__":
     print(f" Charsiu        Phonemes: {phonemes1e}")
     print(f" Misaki         Phonemes: {phonemes1f}")
     print(f" Transphone     Phonemes: {phonemes1g}")
+    print(f" Goruut         Phonemes: {phonemes1h}")
 
     lang = "nl"
     sentence = "DJ's en bezoekers van Tomorrowland waren woensdagavond dolblij toen het paradepaardje van het festival alsnog opende in Oostenrijk op de Mainstage.\nWant het optreden van Metallica, waar iedereen zo blij mee was, zou hoe dan ook doorgaan, aldus de DJ die het nieuws aankondigde."

phoonnx-0.2.0/phoonnx/phonemizers/mwl.py

@@ -0,0 +1,36 @@
+from phoonnx.phonemizers.base import BasePhonemizer, Alphabet
+
+
+class MirandesePhonemizer(BasePhonemizer):
+    _LANGS = ["mwl"]
+
+    def __init__(self):
+        super().__init__(Alphabet.IPA)
+        from mwl_phonemizer import CRFOrthoCorrector
+        self.pho = CRFOrthoCorrector()
+
+    @classmethod
+    def get_lang(cls, target_lang: str) -> str:
+        """
+        Validates and returns the closest supported language code.
+
+        Args:
+            target_lang (str): The language code to validate.
+
+        Returns:
+            str: The validated language code.
+
+        Raises:
+            ValueError: If the language code is unsupported.
+        """
+        return cls.match_lang(target_lang, cls._LANGS)
+
+    def phonemize_string(self, text: str, lang: str) -> str:
+        # Validate language is supported
+        lang = self.get_lang(lang)
+        return self.pho.phonemize_sentence(text)
+
+
+if __name__ == "__main__":
+    pho = MirandesePhonemizer()
+    print(pho.phonemize_string("ls", "mwl"))

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx/version.py

@@ -1,10 +1,10 @@
 # START_VERSION_BLOCK
 VERSION_MAJOR = 0
-VERSION_MINOR = 1
+VERSION_MINOR = 2
 VERSION_BUILD = 0
-VERSION_ALPHA = 3
+VERSION_ALPHA = 0
 # END_VERSION_BLOCK
 
 VERSION_STR = f"{VERSION_MAJOR}.{VERSION_MINOR}.{VERSION_BUILD}"
 if VERSION_ALPHA:
-    VERSION_STR += f"a{VERSION_ALPHA}"
+    VERSION_STR += f"a{VERSION_ALPHA}"

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: phoonnx
-Version: 0.1.0a3
+Version: 0.2.0
 Home-page: https://github.com/TigreGotico/phoonnx
 Author: JarbasAi
 Author-email: jarbasai@mailfence.com
@@ -41,6 +41,7 @@ Provides-Extra: jam
 Provides-Extra: hsn
 Provides-Extra: fil
 Provides-Extra: ml
+Provides-Extra: mwl
 Provides-Extra: lez
 Provides-Extra: si
 Provides-Extra: kk

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx.egg-info/SOURCES.txt

@@ -25,6 +25,7 @@ phoonnx/phonemizers/he.py
 phoonnx/phonemizers/ja.py
 phoonnx/phonemizers/ko.py
 phoonnx/phonemizers/mul.py
+phoonnx/phonemizers/mwl.py
 phoonnx/phonemizers/vi.py
 phoonnx/phonemizers/zh.py
 phoonnx/thirdparty/__init__.py
@@ -91,6 +92,7 @@ phoonnx_train/../phoonnx/phonemizers/he.py
 phoonnx_train/../phoonnx/phonemizers/ja.py
 phoonnx_train/../phoonnx/phonemizers/ko.py
 phoonnx_train/../phoonnx/phonemizers/mul.py
+phoonnx_train/../phoonnx/phonemizers/mwl.py
 phoonnx_train/../phoonnx/phonemizers/vi.py
 phoonnx_train/../phoonnx/phonemizers/zh.py
 phoonnx_train/../phoonnx/thirdparty/__init__.py

{phoonnx-0.1.0a3 → phoonnx-0.2.0}/phoonnx.egg-info/requires.txt

@@ -206,6 +206,9 @@ epitran
 [mt]
 epitran
 
+[mwl]
+mwl_phonemizer
+
 [my]
 epitran