EnglishPal/app/vocabulary.py

import os
import pickle
import random
import re
from collections import defaultdict
from datetime import datetime, timedelta

import snowballstemmer
from flask import session


# 词汇表加载
def load_record(pickle_fname):
    with open(pickle_fname, 'rb') as f:
        d = pickle.load(f)
    return d


# 判断是否是英文单词
def is_english_word(word):
    pattern = re.compile(r'^[a-zA-Z]+$')
    return bool(pattern.match(word))


# 判断日期格式是否有效
def is_valid_datetime_string(date_string, format='%Y%m%d%H%M'):
    try:
        datetime.strptime(date_string, format)
        return True
    except ValueError:
        return False


# 去除非单词字符
def remove_non_words(input_string):
    cleaned_string = re.sub(r'[^a-zA-Z\s]', '', input_string)
    words = cleaned_string.split()
    return ' '.join(words)


# 主类：词汇水平估算器
class VocabularyLevelEstimator:
    # 词汇表（单词：【"雅思","高考"...】）
    _test = load_record('static\words_and_tests.p')  # 词汇到测试来源的映射

    @property
    def level(self):
        total = 0.0  # 总评分
        num = 0  # 计算的单词数
        for word in self.word_lst:
            num += 1
            if word in self._test:
                print(f'{word} : {self._test[word]}')  # 输出单词及其来源
            else:
                print(f'{word}')  # 输出没有评分的单词
        return total / num if num else 0.0  # 返回平均值

    def get_word_level(self, word):
        # 常见高频词汇列表
        other = ['went', 'heard', 'i', 'feet', 'got', 'been', 'gone', 'done', 'had', 'said', 'seen', 'made',
                 'taken', 'come', 'gotten', 'got ', 'ran', 'eaten', 'written', 'found', 'given', 'told',
                 'brought', 'kept', 'stood', 'sat', 'won', 'bought', 'caught', 'begun', 'drank', 'rang', 'sang',
                 'swam', 'blew', 'drew', 'flew', 'grew', 'knew', 'threw', 'shown', 'broken', 'chosen', 'forgotten',
                 'spoke', 'woke', 'woken', 'driven', 'fell', 'given', 'risen', 'mistaken', 'ridden', 'lain', 'lied']

        if word not in self._test: return 3  # 如果词汇不在测试数据中，返回3级
        if word in other: return 3  # 如果是常见的高频词汇，返回3级

        k = self._test[word]
        # 根据词汇的来源设置等级
        if 'CET4' in k:
            return 4
        elif 'OXFORD3000' in k:
            return 5
        elif 'CET6' in k or 'GRADUATE' in k:
            return 6
        elif 'OXFORD5000' in k or 'IELTS' in k:
            return 7
        elif 'BBC' in k:
            return 8

# 用户词汇水平类
class UserVocabularyLevel(VocabularyLevelEstimator):
    # 过滤后的用户生词库
    filtered_frequency = []

    def __init__(self, d):
        if d:
            self.d = d  # 用户的生词库
            self.word_lst = list(d.keys())
            self.filter_user_frequency()

    def filter_user_frequency(self):
        # 过滤出最近一周的生词，用于计算用户词汇水平
        stemmer = snowballstemmer.stemmer('english')
        range_datetime = (datetime.now() - timedelta(days=7)).strftime('%Y%m%d%H%M')

        self.filtered_frequency = []

        for word in self.d:
            if is_english_word(word) and is_valid_datetime_string(self.d[word][0]):
                if self.d[word][0] > range_datetime and word not in self.filtered_frequency:
                    self.filtered_frequency.append(stemmer.stemWord(word))

    @property
    def level(self):
        total = 0.0
        num = 0
        if not self.filtered_frequency: return 0.0
        for word in self.filtered_frequency:
            num += 1
            total += self.get_word_level(word)
        return total / num if num else 0.0


# 文章词汇难度类
class ArticleVocabularyLevel(VocabularyLevelEstimator):
    difficulty_word = dict()

    def __init__(self, content):
        if content:
            self.content = remove_non_words(content)
            self.word_lst = self.content.lower().split()
            self.select_difficulty_word()

    def select_difficulty_word(self, n=10):
        self.difficulty_word = {}
        stemmer = snowballstemmer.stemmer('english')
        for word in self.word_lst:
            original_word = stemmer.stemWord(word)
            self.difficulty_word[original_word] = self.get_word_level(original_word)

        if self.difficulty_word:
            sorted_words = sorted(self.difficulty_word.items(), key=lambda item: item[1], reverse=True)
            top_words = sorted_words[:n]
            self.difficulty_word = {word: difficulty for word, difficulty in top_words}

    @property
    def level(self):
        total = 0.0
        num = 0
        if not self.difficulty_word: return 0.0
        for word in self.difficulty_word:
            num += 1
            total += self.difficulty_word[word]
        return total / num if num else 0.0


if __name__ == '__main__':
    d = load_record('static/frequency/frequency_sb.pickle')  # 加载用户词汇数据
    print(d)

    user = UserVocabularyLevel(d)
    print('用户词汇水平：')
    print(user.level)  # 输出用户的词汇水平

    s = """Energetic = haze dynamic = vigorous = animated Such is Love , Plain like Water
    port him to stand up. She scolded him for not having waken her up. He said that he could manage. A serious quarrel was about to burst out again.
    I called them from Zhuhai, the beautiful city of relaxation and exciting views. I wanted to depict to them how pretty a city Zhuhai is."""

    article = ArticleVocabularyLevel(s)
    print('文章词汇难度：')
    print(article.level)  # 输出文章的词汇难度

    # 测试文章保存
    with open('test/article_test.p', 'wb') as file:
        pickle.dump(s, file)

    with open('test/article_test.p', 'rb') as file:
        loaded_data = pickle.load(file)
        print(loaded_data)