Python使用PyEnchant打造一个高效拼写检查工具

pip install pyenchant  # 推荐使用清华镜像加速安装
# pip install pyenchant -i https://pypi.tuna.tsinghua.edu.cn/simple

import enchant
d = enchant.Dict("en_US")  # 创建字典实例
print(d.tag)  # 输出: en_US

# 检查单词拼写
print(d.check("Hello"))  # True
print(d.check("Helo"))   # False

# 获取拼写建议
suggestions = d.suggest("Helo")
print(suggestions)  # 输出: ['He lo', 'He-lo', 'Hello', 'Helot', 'Help', 'Halo', 'Hell', 'Held', 'Helm', 'Hero', "He'll"]

# 语言支持检测
print(enchant.dict_exists("zh_CN"))  # False（需额外配置中文词典）
print(enchant.list_languages())  # 输出支持的语言列表，如['en_AU', 'en_GB', 'en_US', 'fr_FR']

# 创建仅包含自定义词汇的字典
with open("custom_words.txt", "w") as f:
    f.write("PyEnchant\nGitHub\nNLP")

pwl_dict = enchant.request_pwl_dict("custom_words.txt")
print(pwl_dict.check("PyEnchant"))  # True

# 合并内置字典与自定义词典
merged_dict = enchant.DictWithPWL("en_US", "custom_words.txt")

from enchant.checker import SpellChecker

text = "Ths is a smple tex with erors."
chkr = SpellChecker("en_US")
chkr.set_text(text)

for error in chkr:
    print(f"错误词: {error.word}, 位置: {error.wordpos}, 建议: {d.suggest(error.word)[:3]}")
    error.replace(d.suggest(error.word)[0])  # 自动替换为首个建议

print("修正后文本:", chkr.get_text())  # 输出: This is a sample text with errors.

from enchant.tokenize import get_tokenizer

tokenizer = get_tokenizer("en_US")
tokens = [token for token in tokenizer("PyEnchant is powerful!")]
print(tokens)  
# 输出: [('PyEnchant', 0), ('is', 10), ('powerful', 13), ('!', 21)]

from concurrent.futures import ThreadPoolExecutor

def check_paragraph(para):
    d = enchant.Dict("en_US")
    return [word for word in para.split() if not d.check(word)]

paragraphs = ["First paragraph...", "Second paragraph..."]
with ThreadPoolExecutor(max_workers=4) as executor:
    results = list(executor.map(check_paragraph, paragraphs))

try:
    d = enchant.Dict("nonexistent_lang")
except enchant.errors.DictNotFoundError:
    print("错误：不支持该语言，请检查语言代码")

try:
    print(d.suggest(""))  # 空字符串检查
except enchant.errors.Error as e:
    print(f"拼写检查失败: {e}")

import nltk
from nltk.tokenize import word_tokenize

text = "PyEnchant's integration with NLTK is seamless."
tokens = word_tokenize(text.lower())  # 转换为小写并分词

d = enchant.Dict("en_US")
errors = [word for word in tokens if d.check(word) is False and word.isalpha()]
print("潜在错误词:", errors)  # 输出: ["'s", "seamless"]（需结合词性标注进一步过滤）

import spacy
from spacy.language import Language

@Language.component("spell_checker")
def spell_check_component(doc):
    d = enchant.Dict("en_US")
    for token in doc:
        if not d.check(token.text):
            token.set_extension("is_misspelled", value=True)
    return doc

nlp = spacy.load("en_core_web_sm")
nlp.add_pipe("spell_checker", last=True)
doc = nlp("Helo World!")
print([token.text for token in doc if token._.is_misspelled])  # 输出: ['Helo']

# 错误方式（每次调用都创建新对象）
def bad_check(word):
    return enchant.Dict("en_US").check(word)

# 正确方式（全局复用）
global_dict = enchant.Dict("en_US")
def good_check(word):
    return global_dict.check(word)