Merge pull request #1228 from PyThaiNLP/copilot/revert-pep604-union-types

Revert PEP 604 union syntax to Union[]/Optional[] for Python 3.9 compatibility

Author: bact

pythainlp/augment/wordnet.py

@@ -13,6 +13,7 @@
 
 import itertools
 from collections import OrderedDict
+from typing import Optional
 
 from nltk.corpus import wordnet as wn
 
@@ -123,12 +124,12 @@ def __init__(self):
         pass
 
     def find_synonyms(
-        self, word: str, pos: str | None = None, postag_corpus: str = "orchid"
+        self, word: str, pos: Optional[str] = None, postag_corpus: str = "orchid"
     ) -> list[str]:
         """Find synonyms using wordnet
 
         :param str word: word
-        :param str | None pos: part-of-speech type. Default is None.
+        :param Optional[str] pos: part-of-speech type. Default is None.
         :param str postag_corpus: name of POS tag corpus
         :return: list of synonyms
         :rtype: List[str]

pythainlp/benchmarks/word_tokenization.py

@@ -54,7 +54,7 @@ def _flatten_result(my_dict: dict, sep: str = ":") -> dict:
     :param str sep: separator between the two keys (default: ":")
 
     :return: a one-dimension dictionary with keys combined
-    :rtype: dict[str, float | str]
+    :rtype: dict[str, Union[float, str]]
     """
     return {
         f"{k1}{sep}{k2}": v
@@ -146,7 +146,7 @@ def compute_stats(ref_sample: str, raw_sample: str) -> dict:
     :param str samples: samples that we want to evaluate
 
     :return: metrics at character- and word-level and indicators of correctly tokenized words
-    :rtype: dict[str, float | str]
+    :rtype: dict[str, Union[float, str]]
     """
     ref_sample = _binary_representation(ref_sample)
     sample = _binary_representation(raw_sample)

pythainlp/classify/param_free.py

@@ -5,6 +5,7 @@
 
 import gzip
 import json
+from typing import Optional
 
 import numpy as np
 
@@ -14,15 +15,15 @@ class GzipModel:
     “Low-Resource” Text Classification: A Parameter-Free Classification Method
     with Compressors (Jiang et al., Findings 2023)
 
-    :param list | None training_data: list [(text_sample,label)].
+    :param Optional[list] training_data: list [(text_sample,label)].
         Default is None.
     :param str model_path: Path for loading model (if you saved the model).
         Default is empty string.
     """
 
     def __init__(
         self,
-        training_data: list[tuple[str, str]] | None = None,
+        training_data: Optional[list[tuple[str, str]]] = None,
         model_path: str = "",
     ):
         if model_path:

pythainlp/corpus/common.py

@@ -8,6 +8,7 @@
 from __future__ import annotations
 
 import ast
+from typing import Union
 
 __all__ = [
     "countries",
@@ -82,7 +83,7 @@ def countries() -> frozenset[str]:
     return _THAI_COUNTRIES
 
 
-def provinces(details: bool = False) -> frozenset[str] | list[dict]:
+def provinces(details: bool = False) -> Union[frozenset[str], list[dict]]:
     """Return a frozenset of Thailand province names in Thai such as "กระบี่",
     "กรุงเทพมหานคร", "กาญจนบุรี", and "อุบลราชธานี".
     \n(See: `dev/pythainlp/corpus/thailand_provinces_th.txt\

pythainlp/corpus/core.py

@@ -12,6 +12,7 @@
 import tarfile
 import zipfile
 from importlib.resources import files
+from typing import Optional
 
 from pythainlp import __version__
 from pythainlp.corpus import corpus_db_path, corpus_db_url, corpus_path
@@ -206,7 +207,7 @@ def get_corpus_as_is(filename: str) -> list:
     return lines
 
 
-def get_corpus_default_db(name: str, version: str = "") -> str | None:
+def get_corpus_default_db(name: str, version: str = "") -> Optional[str]:
     """Get model path from default_db.json
 
     :param str name: corpus name
@@ -236,7 +237,7 @@ def get_corpus_default_db(name: str, version: str = "") -> str | None:
     return None
 
 
-def get_corpus_path(name: str, version: str = "", force: bool = False) -> str | None:
+def get_corpus_path(name: str, version: str = "", force: bool = False) -> Optional[str]:
     """Get corpus path.
 
     :param str name: corpus name

pythainlp/corpus/wordnet.py

@@ -12,6 +12,8 @@
 
 from __future__ import annotations
 
+from typing import Optional
+
 import nltk
 
 try:
@@ -27,12 +29,12 @@
 from nltk.corpus import wordnet
 
 
-def synsets(word: str, pos: str | None = None, lang: str = "tha"):
+def synsets(word: str, pos: Optional[str] = None, lang: str = "tha"):
     """This function returns the synonym set for all lemmas of the given word
     with an optional argument to constrain the part of speech of the word.
 
     :param str word: word to find synsets of
-    :param str | None pos: constraint of the part of speech (i.e. *n* for Noun, *v*
+    :param Optional[str] pos: constraint of the part of speech (i.e. *n* for Noun, *v*
                     for Verb, *a* for Adjective, *s* for Adjective
                     satellites, and *r* for Adverb). Default is None.
     :param str lang: abbreviation of language (i.e. *eng*, *tha*).
@@ -98,12 +100,12 @@ def synset(name_synsets):
     return wordnet.synset(name_synsets)
 
 
-def all_lemma_names(pos: str | None = None, lang: str = "tha"):
+def all_lemma_names(pos: Optional[str] = None, lang: str = "tha"):
     """This function returns all lemma names for all synsets of the given
     part of speech tag and language. If part of speech tag is not
     specified, all synsets of all parts of speech will be used.
 
-    :param str | None pos: constraint of the part of speech (i.e. *n* for Noun,
+    :param Optional[str] pos: constraint of the part of speech (i.e. *n* for Noun,
                     *v* for Verb, *a* for Adjective, *s* for
                     Adjective satellites, and *r* for Adverb).
                     By default, *pos* is **None**.
@@ -140,11 +142,11 @@ def all_lemma_names(pos: str | None = None, lang: str = "tha"):
     return wordnet.all_lemma_names(pos=pos, lang=lang)
 
 
-def all_synsets(pos: str | None = None):
+def all_synsets(pos: Optional[str] = None):
     """This function iterates over all synsets constrained by the given
     part of speech tag.
 
-    :param str | None pos: part of speech tag. Default is None.
+    :param Optional[str] pos: part of speech tag. Default is None.
 
     :return: list of synsets constrained by the given part of speech tag.
     :rtype: Iterable[:class:`Synset`]
@@ -188,12 +190,12 @@ def langs():
     return wordnet.langs()
 
 
-def lemmas(word: str, pos: str | None = None, lang: str = "tha"):
+def lemmas(word: str, pos: Optional[str] = None, lang: str = "tha"):
     """This function returns all lemmas given the word with an optional
     argument to constrain the part of speech of the word.
 
     :param str word: word to find lemmas of
-    :param str | None pos: constraint of the part of speech (i.e. *n* for Noun,
+    :param Optional[str] pos: constraint of the part of speech (i.e. *n* for Noun,
                     *v* for Verb, *a* for Adjective, *s* for
                     Adjective satellites, and *r* for Adverb). Default is None.
     :param str lang: abbreviation of language (i.e. *eng*, *tha*).
@@ -391,12 +393,12 @@ def wup_similarity(synsets1, synsets2):
     return wordnet.wup_similarity(synsets1, synsets2)
 
 
-def morphy(form, pos: str | None = None):
+def morphy(form, pos: Optional[str] = None):
     """This function finds a possible base form for the given form,
     with the given part of speech.
 
     :param str form: the form to finds the base form of
-    :param str | None pos: part of speech tag of words to be searched.
+    :param Optional[str] pos: part of speech tag of words to be searched.
         Default is None.
 
     :return: base form of the given form

pythainlp/el/core.py

@@ -3,6 +3,8 @@
 # SPDX-License-Identifier: Apache-2.0
 from __future__ import annotations
 
+from typing import Union
+
 
 class EntityLinker:
     def __init__(
@@ -35,7 +37,7 @@ def __init__(
 
         self.model = MultiEL(model_name=self.model_name, device=self.device)
 
-    def get_el(self, list_text: list[str] | str) -> list[dict] | str:
+    def get_el(self, list_text: Union[list[str], str]) -> Union[list[dict], str]:
         """Get Entity Linking from Thai Text
 
         :param str Union[List[str], str]: list of Thai text or text

pythainlp/generate/core.py

@@ -10,6 +10,7 @@
 from __future__ import annotations
 
 import random
+from typing import Union
 
 from pythainlp.corpus.oscar import (
     unigram_word_freqs as oscar_word_freqs_unigram,
@@ -50,7 +51,7 @@ def gen_sentence(
         prob: float = 0.001,
         output_str: bool = True,
         duplicate: bool = False,
-    ) -> list[str] | str:
+    ) -> Union[list[str], str]:
         """:param str start_seq: word to begin sentence with
         :param int N: number of words
         :param bool output_str: output as string
@@ -143,7 +144,7 @@ def gen_sentence(
         prob: float = 0.001,
         output_str: bool = True,
         duplicate: bool = False,
-    ) -> list[str] | str:
+    ) -> Union[list[str], str]:
         """:param str start_seq: word to begin sentence with
         :param int N: number of words
         :param bool output_str: output as string
@@ -234,7 +235,7 @@ def gen_sentence(
         prob: float = 0.001,
         output_str: bool = True,
         duplicate: bool = False,
-    ) -> list[str] | str:
+    ) -> Union[list[str], str]:
         """:param str start_seq: word to begin sentence with
         :param int N: number of words
         :param bool output_str: output as string

pythainlp/generate/thai2fit.py

@@ -13,6 +13,7 @@
 
 import pickle
 import random
+from typing import Union
 
 # fastai
 import fastai
@@ -101,7 +102,7 @@ def gen_sentence(
     N: int = 4,
     prob: float = 0.001,
     output_str: bool = True,
-) -> list[str] | str:
+) -> Union[list[str], str]:
     """Text generator using Thai2fit
 
     :param str start_seq: word to begin sentence with

pythainlp/khavee/core.py

@@ -4,6 +4,8 @@
 # ruff: noqa: C901
 from __future__ import annotations
 
+from typing import List, Union
+
 from pythainlp import thai_consonants
 from pythainlp.tokenize import subword_tokenize
 from pythainlp.util import remove_tonemark, sound_syllable
@@ -381,7 +383,7 @@ def check_karu_lahu(self, text):
         else:
             return "lahu"
 
-    def check_klon(self, text: str, k_type: int = 8) -> list[str] | str:
+    def check_klon(self, text: str, k_type: int = 8) -> Union[List[str], str]:
         """
         Check the suitability of the poem according to Thai principles.
 
@@ -625,8 +627,8 @@ def check_klon(self, text: str, k_type: int = 8) -> list[str] | str:
             return "Something went wrong. Make sure you enter it in the correct form."
 
     def check_aek_too(
-        self, text: list[str] | str, dead_syllable_as_aek: bool = False
-    ) -> list[bool] | list[str] | bool | str:
+        self, text: Union[List[str], str], dead_syllable_as_aek: bool = False
+    ) -> Union[List[bool], List[str], bool, str]:
         """
         Checker of Thai tonal words