diff --git a/app/Article.py b/app/Article.py
index e0f006a..df9ac3a 100644
--- a/app/Article.py
+++ b/app/Article.py
@@ -7,7 +7,7 @@ import random, glob
import hashlib
from datetime import datetime
from flask import Flask, request, redirect, render_template, url_for, session, abort, flash, get_flashed_messages
-from difficulty import get_difficulty_level, text_difficulty_level, user_difficulty_level
+from difficulty import get_difficulty_level_for_user, text_difficulty_level, user_difficulty_level
path_prefix = '/var/www/wordfreq/wordfreq/'
@@ -32,17 +32,20 @@ def get_article_body(s):
return '\n'.join(lst)
-def get_today_article(user_word_list, existing_articles):
+def get_today_article(user_word_list, visited_articles):
rq = RecordQuery(path_prefix + 'static/wordfreqapp.db')
- if existing_articles is None:
- existing_articles = {
+ if visited_articles is None:
+ visited_articles = {
"index" : 0, # 为 article_ids 的索引
"article_ids": [] # 之前显示文章的id列表,越后越新
}
- if existing_articles["index"] > len(existing_articles["article_ids"])-1:
+ if visited_articles["index"] > len(visited_articles["article_ids"])-1: # 生成新的文章,因此查找所有的文章
rq.instructions("SELECT * FROM article")
- else:
- rq.instructions('SELECT * FROM article WHERE article_id=%d' % (existing_articles["article_ids"][existing_articles["index"]]))
+ else: # 生成阅读过的文章,因此查询指定 article_id 的文章
+ if visited_articles["article_ids"][visited_articles["index"]] == 'null': # 可能因为直接刷新页面导致直接去查询了'null',因此当刷新的页面的时候,需要直接进行“上一篇”操作
+ visited_articles["index"] -= 1
+ visited_articles["article_ids"].pop()
+ rq.instructions('SELECT * FROM article WHERE article_id=%d' % (visited_articles["article_ids"][visited_articles["index"]]))
rq.do()
result = rq.get_results()
random.shuffle(result)
@@ -50,28 +53,36 @@ def get_today_article(user_word_list, existing_articles):
# Choose article according to reader's level
d1 = load_freq_history(path_prefix + 'static/frequency/frequency.p')
d2 = load_freq_history(path_prefix + 'static/words_and_tests.p')
- d3 = get_difficulty_level(d1, d2)
+ d3 = get_difficulty_level_for_user(d1, d2)
d = None
+ result_of_generate_article = "not found"
d_user = load_freq_history(user_word_list)
user_level = user_difficulty_level(d_user, d3) # more consideration as user's behaviour is dynamic. Time factor should be considered.
text_level = 0
- if existing_articles["index"] > len(existing_articles["article_ids"])-1: # 下一篇
- flag_get_article = False
- for reading in result:
- text_level = text_difficulty_level(reading['text'], d3)
- factor = random.gauss(0.8,
- 0.1) # a number drawn from Gaussian distribution with a mean of 0.8 and a stand deviation of 1
- if reading['article_id'] not in existing_articles["article_ids"] and within_range(text_level, user_level, (8.0 - user_level) * factor): # 新的文章之前没有出现过且符合一定范围的水平
- d = reading
- existing_articles["article_ids"].append(d['article_id']) # 列表添加新的文章id;下面进行
- flag_get_article = True
- break
- if not flag_get_article:
- existing_articles["index"] -= 1
- else: # 上一篇
+ if visited_articles["index"] > len(visited_articles["article_ids"])-1: # 生成新的文章
+ amount_of_visited_articles = len(visited_articles["article_ids"])
+ amount_of_existing_articles = result.__len__()
+ if amount_of_visited_articles == amount_of_existing_articles: # 如果当前阅读过的文章的数量 == 存在的文章的数量,即所有的书本都阅读过了
+ result_of_generate_article = "had read all articles"
+ else:
+ for k in range(3): # 最多尝试3次
+ for reading in result:
+ text_level = text_difficulty_level(reading['text'], d3)
+ factor = random.gauss(0.8, 0.1) # a number drawn from Gaussian distribution with a mean of 0.8 and a stand deviation of 1
+ if reading['article_id'] not in visited_articles["article_ids"] and within_range(text_level, user_level, (8.0 - user_level) * factor): # 新的文章之前没有出现过且符合一定范围的水平
+ d = reading
+ visited_articles["article_ids"].append(d['article_id']) # 列表添加新的文章id;下面进行
+ result_of_generate_article = "found"
+ break
+ if result_of_generate_article == "found": # 用于成功找到文章后及时退出外层循环
+ break
+ if result_of_generate_article != "found": # 阅读完所有文章,或者循环3次没有找到适合的文章,则放入空(“null”)
+ visited_articles["article_ids"].append('null')
+ else: # 生成已经阅读过的文章
d = random.choice(result)
text_level = text_difficulty_level(d['text'], d3)
+ result_of_generate_article = "found"
today_article = None
if d:
@@ -86,7 +97,7 @@ def get_today_article(user_word_list, existing_articles):
"answer": get_answer_part(d['question'])
}
- return existing_articles, today_article
+ return visited_articles, today_article, result_of_generate_article
def load_freq_history(path):
diff --git a/app/Login.py b/app/Login.py
index db4df18..cd750d1 100644
--- a/app/Login.py
+++ b/app/Login.py
@@ -3,6 +3,18 @@ import string
from datetime import datetime, timedelta
from UseSqlite import InsertQuery, RecordQuery
+def md5(s):
+ '''
+ MD5摘要
+ :param str: 字符串
+ :return: 经MD5以后的字符串
+ '''
+ h = hashlib.md5(s.encode(encoding='utf-8'))
+ return h.hexdigest()
+
+# import model.user after the defination of md5(s) to avoid circular import
+from model.user import get_user_by_username, insert_user, update_password_by_username
+
path_prefix = '/var/www/wordfreq/wordfreq/'
path_prefix = './' # comment this line in deployment
@@ -12,13 +24,9 @@ def verify_pass(newpass,oldpass):
def verify_user(username, password):
- rq = RecordQuery(path_prefix + 'static/wordfreqapp.db')
- password = md5(username + password)
- rq.instructions_with_parameters("SELECT * FROM user WHERE name=:username AND password=:password", dict(
- username=username, password=password)) # the named style https://docs.python.org/3/library/sqlite3.html
- rq.do_with_parameters()
- result = rq.get_results()
- return result != []
+ user = get_user_by_username(username)
+ encoded_password = md5(username + password)
+ return user is not None and user.password == encoded_password
def add_user(username, password):
@@ -26,19 +34,12 @@ def add_user(username, password):
expiry_date = (datetime.now() + timedelta(days=30)).strftime('%Y%m%d') # will expire after 30 days
# 将用户名和密码一起加密,以免暴露不同用户的相同密码
password = md5(username + password)
- rq = InsertQuery(path_prefix + 'static/wordfreqapp.db')
- rq.instructions_with_parameters("INSERT INTO user VALUES (:username, :password, :start_date, :expiry_date)", dict(
- username=username, password=password, start_date=start_date, expiry_date=expiry_date))
- rq.do_with_parameters()
+ insert_user(username=username, password=password, start_date=start_date, expiry_date=expiry_date)
def check_username_availability(username):
- rq = RecordQuery(path_prefix + 'static/wordfreqapp.db')
- rq.instructions_with_parameters(
- "SELECT * FROM user WHERE name=:username", dict(username=username))
- rq.do_with_parameters()
- result = rq.get_results()
- return result == []
+ existed_user = get_user_by_username(username)
+ return existed_user is None
def change_password(username, old_password, new_password):
@@ -54,35 +55,16 @@ def change_password(username, old_password, new_password):
# 将用户名和密码一起加密,以免暴露不同用户的相同密码
if verify_pass(new_password,old_password): #新旧密码一致
return False
- password = md5(username + new_password)
- rq = InsertQuery(path_prefix + 'static/wordfreqapp.db')
- rq.instructions_with_parameters("UPDATE user SET password=:password WHERE name=:username", dict(
- password=password, username=username))
- rq.do_with_parameters()
+ update_password_by_username(username, new_password)
return True
def get_expiry_date(username):
- rq = RecordQuery(path_prefix + 'static/wordfreqapp.db')
- rq.instructions_with_parameters(
- "SELECT expiry_date FROM user WHERE name=:username", dict(username=username))
- rq.do_with_parameters()
- result = rq.get_results()
- if len(result) > 0:
- return result[0]['expiry_date']
- else:
+ user = get_user_by_username(username)
+ if user is None:
return '20191024'
-
-
-def md5(s):
- '''
- MD5摘要
- :param str: 字符串
- :return: 经MD5以后的字符串
- '''
- h = hashlib.md5(s.encode(encoding='utf-8'))
- return h.hexdigest()
-
+ else:
+ return user.expiry_date
class UserName:
def __init__(self, username):
diff --git a/app/account_service.py b/app/account_service.py
index c1bd64c..a7ed0c4 100644
--- a/app/account_service.py
+++ b/app/account_service.py
@@ -37,7 +37,7 @@ def signup():
session[username] = username
session['username'] = username
session['expiry_date'] = get_expiry_date(username)
- session['existing_articles'] = None
+ session['visited_articles'] = None
return jsonify({'status': '2'})
else:
return jsonify({'status': '1'})
@@ -66,7 +66,7 @@ def login():
session['username'] = username
user_expiry_date = get_expiry_date(username)
session['expiry_date'] = user_expiry_date
- session['existing_articles'] = None
+ session['visited_articles'] = None
return jsonify({'status': '1'})
else:
return jsonify({'status': '0'})
diff --git a/app/admin_service.py b/app/admin_service.py
index 1d1ba6e..a604b5e 100644
--- a/app/admin_service.py
+++ b/app/admin_service.py
@@ -91,10 +91,7 @@ def article():
question = data.get("question", "")
level = data.get("level", "4")
if content:
- try: # check level
- if level not in ['1', '2', '3', '4']:
- raise ValueError
- except ValueError:
+ if level not in ['1', '2', '3', '4']:
return "Level must be between 1 and 4."
add_article(content, source, level, question)
_update_context()
diff --git a/app/difficulty.py b/app/difficulty.py
index 50aa179..cb93768 100644
--- a/app/difficulty.py
+++ b/app/difficulty.py
@@ -8,6 +8,7 @@
import pickle
import math
from wordfreqCMD import remove_punctuation, freq, sort_in_descending_order, sort_in_ascending_order
+import snowballstemmer
def load_record(pickle_fname):
@@ -17,41 +18,51 @@ def load_record(pickle_fname):
return d
-def difficulty_level_from_frequency(word, d):
- level = 1
- if not word in d:
- return level
-
- if 'what' in d:
- ratio = (d['what']+1)/(d[word]+1) # what is a frequent word
- level = math.log( max(ratio, 1), 2)
+def convert_test_type_to_difficulty_level(d):
+ """
+ 对原本的单词库中的单词进行难度评级
+ :param d: 存储了单词库pickle文件中的单词的字典
+ :return:
+ """
+ result = {}
+ L = list(d.keys()) # in d, we have test types (e.g., CET4,CET6,BBC) for each word
- level = min(level, 8)
- return level
+ for k in L:
+ if 'CET4' in d[k]:
+ result[k] = 4 # CET4 word has level 4
+ elif 'OXFORD3000' in d[k]:
+ result[k] = 5
+ elif 'CET6' in d[k] or 'GRADUATE' in d[k]:
+ result[k] = 6
+ elif 'OXFORD5000' in d[k] or 'IELTS' in d[k]:
+ result[k] = 7
+ elif 'BBC' in d[k]:
+ result[k] = 8
+
+ return result # {'apple': 4, ...}
-def get_difficulty_level(d1, d2):
- d = {}
- L = list(d1.keys()) # in d1, we have freuqence for each word
- L2 = list(d2.keys()) # in d2, we have test types (e.g., CET4,CET6,BBC) for each word
- L.extend(L2)
- L3 = list(set(L)) # L3 contains all words
- for k in L3:
- if k in d2:
- if 'CET4' in d2[k]:
- d[k] = 4 # CET4 word has level 4
- elif 'CET6' in d2[k]:
- d[k] = 6
- elif 'BBC' in d2[k]:
- d[k] = 8
- if k in d1: # BBC could contain easy words that are not in CET4 or CET6. So 4 is not reasonable. Recompute difficulty level.
- d[k] = min(difficulty_level_from_frequency(k, d1), d[k])
- elif k in d1:
- d[k] = difficulty_level_from_frequency(k, d1)
+def get_difficulty_level_for_user(d1, d2):
+ """
+ d2 来自于词库的35511个已标记单词
+ d1 用户不会的词
+ 在d2的后面添加单词,没有新建一个新的字典
+ """
+ # TODO: convert_test_type_to_difficulty_level() should not be called every time. Each word's difficulty level should be pre-computed.
+ d2 = convert_test_type_to_difficulty_level(d2) # 根据d2的标记评级{'apple': 4, 'abandon': 4, ...}
+ stemmer = snowballstemmer.stemmer('english')
- return d
+ for k in d1: # 用户的词
+ if k in d2: # 如果用户的词以原型的形式存在于词库d2中
+ continue # 无需评级,跳过
+ else:
+ stem = stemmer.stemWord(k)
+ if stem in d2: # 如果用户的词的词根存在于词库d2的词根库中
+ d2[k] = d2[stem] # 按照词根进行评级
+ else:
+ d2[k] = 3 # 如果k的词根都不在,那么就当认为是3级
+ return d2
-
def revert_dict(d):
'''
@@ -62,12 +73,13 @@ def revert_dict(d):
for k in d:
if type(d[k]) is list: # d[k] is a list of dates.
lst = d[k]
- elif type(d[k]) is int: # for backward compatibility. d was sth like {'word':1}. The value d[k] is not a list of dates, but a number representing how frequent this word had been added to the new word book.
+ elif type(d[
+ k]) is int: # for backward compatibility. d was sth like {'word':1}. The value d[k] is not a list of dates, but a number representing how frequent this word had been added to the new word book.
freq = d[k]
- lst = freq*['2021082019'] # why choose this date? No particular reasons. I fix the bug in this date.
+ lst = freq * ['2021082019'] # why choose this date? No particular reasons. I fix the bug in this date.
for time_info in lst:
- date = time_info[:10] # until hour
+ date = time_info[:10] # until hour
if not date in d2:
d2[date] = [k]
else:
@@ -76,42 +88,44 @@ def revert_dict(d):
def user_difficulty_level(d_user, d):
- d_user2 = revert_dict(d_user) # key is date, and value is a list of words added in that date
+ d_user2 = revert_dict(d_user) # key is date, and value is a list of words added in that date
count = 0
geometric = 1
- for date in sorted(d_user2.keys(), reverse=True): # most recently added words are more important while determining user's level
- lst = d_user2[date] # a list of words
- lst2 = [] # a list of tuples, (word, difficulty level)
- for word in lst:
+ for date in sorted(d_user2.keys(),
+ reverse=True): # most recently added words are more important while determining user's level
+ lst = d_user2[date] # a list of words
+ lst2 = [] # a list of tuples, (word, difficulty level)
+ for word in lst:
if word in d:
lst2.append((word, d[word]))
- lst3 = sort_in_ascending_order(lst2) # easiest tuple first
- #print(lst3)
+ lst3 = sort_in_ascending_order(lst2) # easiest tuple first
+ # print(lst3)
for t in lst3:
word = t[0]
hard = t[1]
- #print('WORD %s HARD %4.2f' % (word, hard))
+ # print('WORD %s HARD %4.2f' % (word, hard))
geometric = geometric * (hard)
count += 1
if count >= 10:
- return geometric**(1/count)
+ return geometric ** (1 / count)
- return geometric**(1/max(count,1))
+ return geometric ** (1 / max(count, 1))
def text_difficulty_level(s, d):
s = remove_punctuation(s)
L = freq(s)
- lst = [] # a list of tuples, each tuple being (word, difficulty level)
+ lst = [] # a list of tuples, each tuple being (word, difficulty level)
+ stop_words = {'the':1, 'and':1, 'of':1, 'to':1, 'what':1, 'in':1, 'there':1, 'when':1, 'them':1, 'would':1, 'will':1, 'out':1, 'his':1, 'mr':1, 'that':1, 'up':1, 'more':1, 'your':1, 'it':1, 'now':1, 'very':1, 'then':1, 'could':1, 'he':1, 'any':1, 'some':1, 'with':1, 'into':1, 'you':1, 'our':1, 'man':1, 'other':1, 'time':1, 'was':1, 'than':1, 'know':1, 'about':1, 'only':1, 'like':1, 'how':1, 'see':1, 'is':1, 'before':1, 'such':1, 'little':1, 'two':1, 'its':1, 'as':1, 'these':1, 'may':1, 'much':1, 'down':1, 'for':1, 'well':1, 'should':1, 'those':1, 'after':1, 'same':1, 'must':1, 'say':1, 'first':1, 'again':1, 'us':1, 'great':1, 'where':1, 'being':1, 'come':1, 'over':1, 'good':1, 'himself':1, 'am':1, 'never':1, 'on':1, 'old':1, 'here':1, 'way':1, 'at':1, 'go':1, 'upon':1, 'have':1, 'had':1, 'without':1, 'my':1, 'day':1, 'be':1, 'but':1, 'though':1, 'from':1, 'not':1, 'too':1, 'another':1, 'this':1, 'even':1, 'still':1, 'her':1, 'yet':1, 'under':1, 'by':1, 'let':1, 'just':1, 'all':1, 'because':1, 'we':1, 'always':1, 'off':1, 'yes':1, 'so':1, 'while':1, 'why':1, 'which':1, 'me':1, 'are':1, 'or':1, 'no':1, 'if':1, 'an':1, 'also':1, 'thus':1, 'who':1, 'cannot':1, 'she':1, 'whether':1} # ignore these words while computing the artile's difficulty level
for x in L:
word = x[0]
- if word in d:
+ if word not in stop_words and word in d:
lst.append((word, d[word]))
- lst2 = sort_in_descending_order(lst) # most difficult words on top
- #print(lst2)
+ lst2 = sort_in_descending_order(lst) # most difficult words on top
+ # print(lst2)
count = 0
geometric = 1
for t in lst2:
@@ -119,24 +133,20 @@ def text_difficulty_level(s, d):
hard = t[1]
geometric = geometric * (hard)
count += 1
- if count >= 20: # we look for n most difficult words
- return geometric**(1/count)
-
- return geometric**(1/max(count,1))
+ if count >= 20: # we look for n most difficult words
+ return geometric ** (1 / count)
+ return geometric ** (1 / max(count, 1))
if __name__ == '__main__':
-
-
d1 = load_record('frequency.p')
- #print(d1)
+ # print(d1)
d2 = load_record('words_and_tests.p')
- #print(d2)
+ # print(d2)
-
- d3 = get_difficulty_level(d1, d2)
+ d3 = get_difficulty_level_for_user(d1, d2)
s = '''
South Lawn
@@ -197,7 +207,6 @@ Amidst the aftermath of this shocking referendum vote, there is great uncertaint
'''
-
s = '''
British Prime Minister Boris Johnson walks towards a voting station during the Brexit referendum in Britain, June 23, 2016. (Photo: EPA-EFE)
@@ -218,7 +227,6 @@ The prime minister was forced to ask for an extension to Britain's EU departure
Johnson has repeatedly pledged to finalize the first stage, a transition deal, of Britain's EU divorce battle by Oct. 31. A second stage will involve negotiating its future relationship with the EU on trade, security and other salient issues.
'''
-
s = '''
Thank you very much. We have a Cabinet meeting. We’ll have a few questions after grace. And, if you would, Ben, please do the honors.
@@ -233,17 +241,11 @@ We need — for our farmers, our manufacturers, for, frankly, unions and non-uni
'''
-
-
-
- #f = open('bbc-fulltext/bbc/entertainment/001.txt')
+ # f = open('bbc-fulltext/bbc/entertainment/001.txt')
f = open('wordlist.txt')
s = f.read()
f.close()
-
-
-
print(text_difficulty_level(s, d3))
-
+
diff --git a/app/model/user.py b/app/model/user.py
index 28173b9..d684332 100644
--- a/app/model/user.py
+++ b/app/model/user.py
@@ -1,5 +1,6 @@
from model import *
from Login import md5
+from pony import orm
def get_users():
with db_session:
@@ -11,6 +12,11 @@ def get_user_by_username(username):
if user:
return user.first()
+def insert_user(username, password, start_date, expiry_date):
+ with db_session:
+ user = User(name=username, password=password, start_date=start_date, expiry_date=expiry_date)
+ orm.commit()
+
def update_password_by_username(username, password="123456"):
with db_session:
user = User.select(name=username)
diff --git a/app/pickle_idea2.py b/app/pickle_idea2.py
index 4055fc4..0da55bc 100644
--- a/app/pickle_idea2.py
+++ b/app/pickle_idea2.py
@@ -68,7 +68,7 @@ def save_frequency_to_pickle(d, pickle_fname):
d2 = {}
for k in d:
if not k in exclusion_lst and not k.isnumeric() and not len(k) < 2:
- d2[k] = list(sorted(set(d[k])))
+ d2[k] = list(sorted(d[k])) # 原先这里是d2[k] = list(sorted(set(d[k])))
pickle.dump(d2, f)
f.close()
diff --git a/app/static/config.yml b/app/static/config.yml
index 7b2b3eb..285f31f 100644
--- a/app/static/config.yml
+++ b/app/static/config.yml
@@ -1,18 +1,17 @@
# 全局引入的css文件地址
css:
item:
- - static/css/bootstrap.css
+ - ../static/css/bootstrap.css
# 全局引入的js文件地址
js:
head: # 在页面加载之前加载
- - static/js/jquery.js
- - static/js/read.js
- - static/js/word_operation.js
-
+ - ../static/js/jquery.js
+ - ../static/js/read.js
+ - ../static/js/word_operation.js
bottom: # 在页面加载完之后加载
- - static/js/fillword.js
- - static/js/highlight.js
+ - ../static/js/fillword.js
+ - ../static/js/highlight.js
# 高亮样式,目前仅支持修改颜色
highlight:
diff --git a/app/static/js/highlight.js b/app/static/js/highlight.js
index 5ec9663..0cea31a 100644
--- a/app/static/js/highlight.js
+++ b/app/static/js/highlight.js
@@ -38,8 +38,18 @@ function highLight() {
list[i] = list[i].replace('|', "");
list[i] = list[i].replace('?', "");
if (list[i] !== "" && "".indexOf(list[i]) === -1 && "".indexOf(list[i]) === -1) {
- //将文章中所有出现该单词word的地方改为:" " + word + " "。 正则表达式RegExp()中,"\\s"代表单词前后必须要有空格,以防止只对单词中的部分字符高亮的情况出现。
- articleContent = articleContent.replace(new RegExp("\\s"+list[i]+"\\s", "g"), " " + list[i] + " ");
+ //将文章中所有出现该单词word的地方改为:"" + word + ""。 正则表达式RegExp()中,"\\b"代表单词边界匹配。
+
+ //修改代码
+ let articleContent_fb = articleContent; //文章副本
+ while(articleContent_fb.toLowerCase().indexOf(list[i].toLowerCase()) !== -1 && list[i]!=""){
+ //找到副本中和list[i]匹配的第一个单词(第一种匹配情况),并赋值给list[i]。
+ const index = articleContent_fb.toLowerCase().indexOf(list[i].toLowerCase());
+ list[i] = articleContent_fb.substring(index, index + list[i].length);
+
+ articleContent_fb = articleContent_fb.substring(index + list[i].length); // 使用副本中list[i]之后的子串替换掉副本
+ articleContent = articleContent.replace(new RegExp("\\b"+list[i]+"\\b","g"),"" + list[i] + "");
+ }
}
}
document.getElementById("article").innerHTML = articleContent;
diff --git a/app/static/words_and_tests.p b/app/static/words_and_tests.p
index 62afd6d..5c46c5a 100644
Binary files a/app/static/words_and_tests.p and b/app/static/words_and_tests.p differ
diff --git a/app/templates/admin_index.html b/app/templates/admin_index.html
index 68ee68f..f62a137 100644
--- a/app/templates/admin_index.html
+++ b/app/templates/admin_index.html
@@ -30,7 +30,7 @@
diff --git a/app/templates/admin_manage_user.html b/app/templates/admin_manage_user.html
index a3f0ca0..cee4667 100644
--- a/app/templates/admin_manage_user.html
+++ b/app/templates/admin_manage_user.html
@@ -68,9 +68,9 @@
{% endfor %}
{% endif %}
+