Source code for pythainlp.corpus.oscar

# -*- coding: utf-8 -*-
# SPDX-FileCopyrightText: 2016-2024 PyThaiNLP Project
# SPDX-License-Identifier: Apache-2.0
"""
Thai unigram word frequency from OSCAR Corpus (words tokenized using ICU)

Credit: Korakot Chaovavanich
https://web.facebook.com/groups/colab.thailand/permalink/1524070061101680/
"""

__all__ = ["word_freqs", "unigram_word_freqs"]

from collections import defaultdict
from typing import List, Tuple

from pythainlp.corpus import get_corpus_path

_OSCAR_FILENAME = "oscar_icu"


[docs] def word_freqs() -> List[Tuple[str, int]]: """ Get word frequency from OSCAR Corpus (words tokenized using ICU) """ freqs: list[tuple[str, int]] = [] path = get_corpus_path(_OSCAR_FILENAME) if not path: return freqs path = str(path) with open(path, "r", encoding="utf-8-sig") as f: lines = list(f.readlines()) del lines[0] for line in lines: temp = line.strip().split(",") if len(temp) >= 2: if temp[0] != " " and '"' not in temp[0]: freqs.append((temp[0], int(temp[1]))) elif temp[0] == " ": freqs.append(("<s/>", int(temp[1]))) return freqs
[docs] def unigram_word_freqs() -> dict[str, int]: """ Get unigram word frequency from OSCAR Corpus (words tokenized using ICU) """ freqs: dict[str, int] = defaultdict(int) path = get_corpus_path(_OSCAR_FILENAME) if not path: return freqs path = str(path) with open(path, "r", encoding="utf-8-sig") as fh: lines = list(fh.readlines()) del lines[0] for i in lines: temp = i.strip().split(",") if temp[0] != " " and '"' not in temp[0]: freqs[temp[0]] = int(temp[-1]) elif temp[0] == " ": freqs["<s/>"] = int(temp[-1]) return freqs