def get_config(self):
"""
Get config and its associated model
"""
(config, attention) = MODELS[self.model_type]
config = config(**self.config)
transformers.models.roberta.modeling_roberta.RobertaSelfAttention = attention
return config
def __init__(self, model_path='tmp/model', w2id_path='data/w2id.json'):
with open(w2id_path) as f:
self.w2id = json.load(f)
conf = config()
self.num_steps = conf.num_steps
self.cutModel = lstm_model(conf)
self.sess = tf.Session()
saver = tf.train.Saver()
saver.restore(self.sess, model_path)
#Instanciation object Redis
objRedis = initRedis()
#Va contenir le texte du SMS
dataTextSms = ""
#Défini si l'exécution doit continuer pour l'ensemble des scripts
objRedis.set('flagExecute_Treatment', 0)
objRedis.set('nameExecute_Treatment', 'INIT')
"""
On instancie l'objet gérant les configurations
Et on récupère les informations pour les placer dans Redis
Clé : config_*
"""
objConfig = config()
objConfig.getThis()
dataTextSms = dataTextSms + "Config OK \r\n"
# On definit le path du log
logging.basicConfig(filename=objRedis.get('config_path_log').decode("utf-8") +
time.strftime('%Y%m%d') + '_init_adsb.log',
level=logging.INFO)
"""
On instancie l'objet gérant les squawk
Et on récupère les informations pour les placer dans Redis
Clé : squawk
"""
objSquawk = squawk()
returnSquawk = objSquawk.setDataInRedis()
dataTextSms = dataTextSms + " " + returnSquawk.__str__() + " Squawk \r\n"
doc_file_path = model.generate_path([main_path, "docs", doc_name])
print("文档: \"{name}\" \"{link}\"".format(name=doc_name,
link=doc_link))
doc_list.append((doc_link, doc_file_path))
# 下载
if config.Download:
if config.Download_Method == "Aria2": # 这里是调用aria2的下载
model.aira2_download(info_list + video_list + doc_list)
model.download_queue(session,
srt_list,
queue_length=config.Download_Queue_Length
) # 需要session或者有时间期限的
else: # 默认调用自建下载
model.download_queue(session,
info_list + video_list + srt_list +
doc_list,
queue_length=config.Download_Queue_Length)
else:
print("No course Id,Please check!")
return
if __name__ == '__main__':
course_url = ""
# Loading config
config = model.config("settings.conf", "xuetangx")
main(course_url, config)
print("字幕Eng: \"{name}\" \"{link}\"".format(name=srt_file_name, link=srt_eng_link))
srt_list.append((srt_eng_link, srt_file_path))
video_in_chapter_list[-1] += 1
if lesson_content_type == 3 and config.Download_Docs: # Documentation
doc_link = str(re.search(r'textOrigUrl:"(.+?)"', rdata).group(1))
doc_name = "{0}.pdf".format(lesson_name)
doc_path = model.generate_path([main_path, "Docs", lesson_loc_pattern])
doc_file_path = model.generate_path([doc_path, doc_name])
doc_list.append((doc_link, doc_file_path))
print("文档: \"{name}\" \"{link}\"".format(name=doc_name, link=doc_link))
if config.Download:
if config.Download_Method == "Aria2": # 这里是调用aria2的下载
model.aira2_download(info_list + video_list)
# 需要session或者有时间期限的，仍然使用自建下载
model.download_queue(session, srt_list + doc_list, queue_length=config.Download_Queue_Length)
else: # 默认调用自建下载
model.download_queue(session, info_list + video_list + srt_list + doc_list,
queue_length=config.Download_Queue_Length)
else:
err_message = re.search(r'message:(.+)\}\)', rdata).group(1)
print("Error:{0},Please make sure you login by 163-email "
"and your \"Session-Cookies\" pair is right.".format(err_message))
if __name__ == '__main__':
course_url = ""
config = model.config("settings.conf", "icourse163")
main(course_url, config=config)
def main(course_url):
config = model.config("settings.conf", "icourse163")
session = model.login(site="icourse163", conf=config)
http_session_id = session.cookies["NTESSTUDYSI"]
c_tid = re.search(r"(?:(learn)|(course))/(?P<id>(?P<c_id>[\w:+-]+)(\?tid=(?P<t_id>\d+))?)#?/?", course_url)
# Download cache list
main_list = []
srt_list = []
doc_list = []
# handle the course_url links to Get right courseId and termId
if c_tid:
if c_tid.group("t_id"): # 当使用者提供tid的时候默认使用使用者tid
term_id = c_tid.group("t_id")
info_url = "http://www.icourse163.org/course/{id}#/info".format(id=c_tid.group('id'))
else: # 否则通过info页面重新获取最新tid
term_id = None
print("No termId which you want to download.Will Choose the Lastest term.")
info_url = "http://www.icourse163.org/course/{id}#/info".format(id=c_tid.group('c_id')) # 使用课程默认地址
page_about = session.get(url=info_url)
if page_about.url == page_about.request.url: # 存在该课程
# 当课程不存在的时候会302重定向到http://www.icourse163.org/，通过检查返回、请求地址是否一致判断
page_about_bs = BeautifulSoup(page_about.text, "lxml")
course_info_raw = page_about_bs.find("script", text=re.compile(r"termDto")).string.replace("\n", "")
if term_id is None: # 没有提供tid时候自动寻找最新课程信息
term_id = re.search(r"termId : \"(\d+)\"", course_info_raw).group(1)
# 获取课程信息
course_page_title = re.search(r'(.+?)_(.+?)_(.+?)', page_about_bs.title.string)
course_title = model.clean_filename(course_page_title.group(1))
school = course_page_title.group(2)
teacher = model.sort_teacher(page_about_bs.find_all('h3', class_="f-fc3"))
folder = model.clean_filename('-'.join([course_title, school, teacher]))
print("The Download INFO:\n" # Output download course info
"link:{url}\nCourse:{folder}\nid:{id}\n".format(url=info_url, folder=folder, id=term_id))
main_path = model.generate_path([config.Download_Path, folder])
info_img_link = page_about_bs.find("div", id="j-courseImg").img["src"]
img_file_name = r"课程封面图-{title}.png".format(title=course_title)
img_file_path = model.generate_path([main_path, img_file_name])
print("课程封面图: {link}".format(link=info_img_link))
main_list.append((info_img_link, img_file_path))
# intro_video
video_search = re.search(r"videoId : \"(\d+)\"", course_info_raw)
if video_search:
payload = {
'callCount': 1,
'scriptSessionId': '${scriptSessionId}' + str(random.randint(0, 200)),
'httpSessionId': http_session_id,
'c0-scriptName': 'CourseBean',
'c0-methodName': 'getLessonUnitPreviewVo',
'c0-id': 0,
'c0-param0': video_search.group(1),
'c0-param1': 1,
'batchId': random.randint(1000000000000, 20000000000000)
}
ask_video_url = "http://www.icourse163.org/dwr/call/plaincall/CourseBean.getLessonUnitPreviewVo.dwr"
resp = session.post(url=ask_video_url, data=payload).text
for k in ['mp4ShdUrl', 'mp4HdUrl', 'mp4SdUrl']: # , 'flvShdUrl', 'flvHdUrl', 'flvSdUrl'
video_search_group = re.search(r's\d+.(?P<VideoType>' + str(k) + ')="(?P<dllink>.+?)";', resp)
if video_search_group:
info_video_link = video_search_group.group("dllink")
video_file_name = r"课程简介-{title}.mp4".format(title=course_title)
video_file_path = model.generate_path([main_path, video_file_name])
print("课程简介视频: {link}".format(link=info_video_link))
main_list.append((info_video_link, video_file_path))
break
else:
print("Not found this course in \"icourse163.org\",Check Please")
return
# Get course's chapter
payload = {
'callCount': 1,
'scriptSessionId': '${scriptSessionId}' + str(random.randint(0, 200)),
'httpSessionId': http_session_id,
'c0-scriptName': 'CourseBean',
'c0-methodName': 'getLastLearnedMocTermDto',
'c0-id': 0,
'c0-param0': term_id,
'batchId': random.randint(1000000000000, 20000000000000)
}
cs_url = 'http://www.icourse163.org/dwr/call/plaincall/CourseBean.getLastLearnedMocTermDto.dwr'
rdata = session.post(cs_url, data=payload, timeout=None).text
if re.search(r"var s\d+={}", rdata):
print("Generate Download information.")
# Data cleaning Reg
week_reg = re.compile(r"s\d+.contentId=null;"
r".+s\d+.lessons=(?P<lessons>s\d+)"
r".+s\d+.name=\"(?P<week_name>.+?)\"")
chapter_reg = re.compile(r"s\d+.chapterId=\d+;"
r".+s\d+.name=\"(?P<chapter_name>.+?)\"")
lesson_reg = re.compile(r"s\d+.anchorQuestions=(null|s\d+);"
r".+s\d+.contentId=(?P<contentId>\d+)"
r".+s\d+.contentType=(?P<contentType>\d+)"
r".+s\d+.id=(?P<id>\d+)"
r".+s\d+.name=\"(?P<lesson_name>.+?)\"")
# count_list
week_list = []
chapter_list = []
video_in_chapter_list = []
for line in rdata.splitlines():
if re.match(week_reg, line): # Week
week_re = re.search(week_reg, line)
week_name = model.clean_filename(model.raw_unicode_escape(week_re.group("week_name")))
week_list.append(week_name)
if re.match(chapter_reg, line): # Chapter
chapter_re = re.search(chapter_reg, line)
chapter_name = model.clean_filename(model.raw_unicode_escape(chapter_re.group("chapter_name")))
chapter_list.append(chapter_name)
print("\n", week_list[-1], chapter_list[-1])
video_in_chapter_list.append(0)
if re.match(lesson_reg, line):
lesson_re = re.search(lesson_reg, line)
lesson_loc_pattern = model.generate_path([week_list[-1], chapter_list[-1]])
lesson_name = model.clean_filename(model.raw_unicode_escape(lesson_re.group("lesson_name")))
lesson_content_type = int(lesson_re.group("contentType"))
# prepare data and post
payload = {
'callCount': 1,
'scriptSessionId': '${scriptSessionId}' + str(random.randint(0, 200)),
'httpSessionId': http_session_id,
'c0-scriptName': 'CourseBean',
'c0-methodName': 'getLessonUnitLearnVo',
'c0-id': 1,
'c0-param0': lesson_re.group("contentId"),
'c0-param1': lesson_content_type,
'c0-param2': 0,
'c0-param3': lesson_re.group("id"),
'batchId': random.randint(1000000000000, 20000000000000)
}
cs_url = 'http://www.icourse163.org/dwr/call/plaincall/CourseBean.getLessonUnitLearnVo.dwr'
rdata = session.post(cs_url, data=payload, timeout=None).text
# Sort data depend on it's contentType
# 1 -> Video ,2 -> Test ,3 -> Docs ,4 -> Rich text ,5 -> Examination ,6 -> Discussion
if lesson_content_type == 1: # Video
count = video_in_chapter_list[-1]
count_lesson_name = model.clean_filename("{0} {lesson}".format(count, lesson=lesson_name))
for k in ['mp4ShdUrl', 'mp4HdUrl', 'mp4SdUrl']: # , 'flvShdUrl', 'flvHdUrl', 'flvSdUrl'
if re.search(r's\d+.{0}=".+?";'.format(k), rdata):
k_type = re.search("mp4(.+)Url", k).group(1)
video_file_name = "{0}.mp4".format(count_lesson_name)
if k_type != "Shd":
video_file_name = "{0}_{type}.mp4".format(count_lesson_name, type=k_type)
video_link = re.search(r's\d+.' + str(k) + r'="(.+?\.mp4).+?";', rdata).group(1)
video_file_path = model.generate_path([main_path, lesson_loc_pattern, video_file_name])
main_list.append((video_link, video_file_path))
print("视频: \"{name}\" \"{link}\"".format(name=video_file_name, link=video_link))
break
# Subtitle
if config.Download_Srt:
srt_path = model.generate_path([main_path, "Srt", lesson_loc_pattern])
if re.search(r's\d+.name="\\u4E2D\\u6587";s\d+.url="(.+?)"', rdata): # Chinese
srt_chs_re = re.search(r's\d+.name="\\u4E2D\\u6587";s\d+.url="(?P<url>.+?)"', rdata)
srt_file_name = "{0}.chs.srt".format(count_lesson_name)
srt_file_path = model.generate_path([srt_path, srt_file_name])
srt_chs_link = srt_chs_re.group("url")
print("字幕Chs: \"{name}\" \"{link}\"".format(name=srt_file_name, link=srt_chs_link))
srt_list.append((srt_chs_link, srt_file_path))
if re.search(r's\d+.name="\\u82F1\\u6587";s\d+.url="(.+?)"', rdata): # English
srt_eng_re = re.search(r's\d+.name="\\u82F1\\u6587";s\d+.url="(?P<url>.+?)"', rdata)
srt_file_name = "{0}.eng.srt".format(lesson_name)
srt_file_path = model.generate_path([srt_path, srt_file_name])
srt_eng_link = srt_eng_re.group("url")
print("字幕Eng: \"{name}\" \"{link}\"".format(name=srt_file_name, link=srt_eng_link))
srt_list.append((srt_eng_link, srt_file_path))
video_in_chapter_list[-1] += 1
if lesson_content_type == 3 and config.Download_Docs: # Documentation
doc_link = str(re.search(r'textOrigUrl:"(.+?)"', rdata).group(1))
doc_name = "{0}.pdf".format(lesson_name)
doc_path = model.generate_path([main_path, "Docs", lesson_loc_pattern])
doc_file_path = model.generate_path([doc_path, doc_name])
doc_list.append((doc_link, doc_file_path))
print("文档: \"{name}\" \"{link}\"".format(name=doc_name, link=doc_link))
if config.Download:
if config.Download_Method == "Aria2": # 这里是调用aria2的下载
model.aira2_download(main_list)
# 需要session或者有时间期限的，仍然使用自建下载
model.download_queue(session, srt_list + doc_list, queue_length=config.Download_Queue_Length)
else: # 默认调用自建下载
model.download_queue(session, main_list + srt_list + doc_list,
queue_length=config.Download_Queue_Length)
else:
err_message = re.search(r'message:(.+)\}\)', rdata).group(1)
print("Error:{0},Please make sure you login by 163-email "
"and your \"Session-Cookies\" pair is right.".format(err_message))
else:
print("No course Id,Please check!")
return
def main(course_url):
config = model.config("settings.conf", "xuetangx") # Loading config
session = model.login(site="xuetangx", conf=config)
course_id_search = re.search(r"courses/(?P<id>.+)/(courseware|info|discussion|wiki|progress|about)", course_url)
# Download cache list
main_list = []
srt_list = []
doc_list = []
if course_id_search:
course_id = course_id_search.group("id")
main_page = "http://www.xuetangx.com/courses/{course_id}".format(course_id=course_id)
page_about_url = "{course_host}/about".format(course_host=main_page)
page_about = session.get(url=page_about_url)
if page_about.text.find("页面无法找到") == -1: # if Exist
page_about_bs = BeautifulSoup(page_about.text, "lxml")
# load course info
course_detail_bs = page_about_bs.find("section", class_="courseabout_detail")
course_name_tag = course_detail_bs.find("h3", class_="courseabout_title")
course_title = model.clean_filename(course_name_tag.get_text())
school = course_name_tag.find_next("a").get_text()
teacher = model.sort_teacher(
page_about_bs.find("ul", class_="teacher_info").find_all("span", class_="name"))
folder = model.clean_filename('-'.join([course_title, school, teacher]))
print("The Download INFO:\n" # Output download course info
"link:{url}\nCourse:{folder}\nid:{id}\n".format(url=page_about_url, folder=folder, id=course_id))
main_path = model.generate_path([config.Download_Path, folder])
video_box = course_detail_bs.find('div', class_='video_box')
try:
info_img_link = model.link_check("http://www.xuetangx.com", video_box['data-poster'])
info_video_link = get_video(session, video_box["data-ccid"])
if info_video_link:
video_file_name = r"课程简介-{title}.mp4".format(title=course_title)
video_file_path = model.generate_path([main_path, video_file_name])
print("课程简介视频: {link}".format(link=info_video_link))
main_list.append((info_video_link, video_file_path))
except KeyError:
info_img_link = model.link_check("http://www.xuetangx.com", video_box.img["src"])
if info_img_link:
img_file_name = r"课程封面图-{title}.jpg".format(title=course_title)
img_file_path = model.generate_path([main_path, img_file_name])
print("课程封面图: {link}".format(link=info_img_link))
main_list.append((info_img_link, img_file_path))
else:
print("Not found this course in \"xuetangx.com\",Check Please")
return
# 获取课程参与信息及判断是否已经参加课程
page_courseware = session.get(url="{0}/courseware".format(main_page))
if page_courseware.url.find("about") == -1 and page_courseware.url.find("login") == -1: # 成功获取目录
# 这里根据url判断：
# 1、如果登陆了，但是没有参加该课程，会跳转到 ../about页面
# 2、如果未登录(或密码错误)，会跳转到http://www.xuetangx.com/accounts/login?next=.. 页面
print("Generate Download information.")
# 处理courseware页面
courseware_bs = BeautifulSoup(page_courseware.text, "lxml")
chapter = courseware_bs.find_all("div", class_="chapter")
for week in chapter:
week_name = model.clean_filename(week.h3.a.string.strip())
for lesson in week.ul.find_all("a"):
# 获取课程信息
lesson_name = model.clean_filename(lesson.p.string) # 主标题
lesson_page = session.get(url="http://www.xuetangx.com{href}".format(href=lesson['href']),
timeout=None)
lesson_bs = BeautifulSoup(lesson_page.text, "lxml")
tab_list = {}
for tab in lesson_bs.find_all("a", role="tab"):
tab_list[tab.get('id')] = re.search("(.+)", tab.get('title')).group(1)
seq_contents = lesson_bs.find_all('div', class_="seq_contents")
print("\n", week_name, lesson_name)
seq_video_content_len = 0
for seq in seq_contents:
if re.search(r"data-type=[\'\"]Video[\'\"]", seq.text):
seq_video_content_len += 1
for i, seq in enumerate(seq_contents):
seq_name = lesson_name
seq_path = model.generate_path([main_path, week_name])
srt_path = model.generate_path([main_path, "srt", week_name])
doc_path = model.generate_path([main_path, "docs", week_name])
if seq_video_content_len > 1: # 如果只有一个的话，就不用建立子文件夹了
seq_name_raw = model.clean_filename(tab_list[seq.get("aria-labelledby")])
seq_name = r"{0} {1}".format(i, seq_name_raw)
seq_path = model.generate_path([seq_path, lesson_name])
srt_path = model.generate_path([srt_path, lesson_name])
doc_path = model.generate_path([doc_path, lesson_name])
if re.search(r"data-type=[\'\"]Video[\'\"]", seq.text): # 视频
lesson_ccsource = re.search(r"data-ccsource=[\'\"](.+)[\'\"]", seq.text).group(1)
video_link = get_video(session, lesson_ccsource)
video_file_name = "{0}.mp4".format(seq_name)
if video_link.find == -1:
video_file_name = "{0}_sd.mp4".format(seq_name)
video_file_path = model.generate_path([seq_path, video_file_name])
print("视频: \"{name}\" \"{link}\"".format(name=video_file_name, link=video_link))
main_list.append((video_link, video_file_path))
seq_bs = BeautifulSoup(seq.text, "lxml")
if config.Download_Srt and seq_bs.find("a", text="下载字幕"): # 字幕
raw_link = seq_bs.find("a", text="下载字幕")["href"]
srt_link = model.link_check("http://www.xuetangx.com", raw_link)
srt_file_name = "{0}.srt".format(seq_name)
srt_file_path = model.generate_path([srt_path, srt_file_name])
print("字幕: \"{name}\" \"{link}\"".format(name=srt_file_name, link=srt_link))
srt_list.append((srt_link, srt_file_path))
if config.Download_Docs and seq_bs.find("a", text="下载讲义"): # 讲义
raw_link = seq_bs.find("a", text="下载讲义")["href"]
doc_link = model.link_check("http://www.xuetangx.com", raw_link)
doc_file_name = model.clean_filename(doc_link.split("/")[-1])
doc_file_path = model.generate_path([doc_path, doc_file_name])
print("文档: \"{name}\" \"{link}\"".format(name=doc_file_name, link=doc_link))
doc_list.append((doc_link, doc_file_path))
else: # 未登陆成功或者没参加该课程
print("Something Error,You may not Join this course or Enter the wrong password.")
return
# 处理info页面的课程讲义
page_info = session.get(url="{0}/info".format(main_page))
info_bs = BeautifulSoup(page_info.text, "lxml")
doc_menu = info_bs.find("section", attrs={"aria-label": re.compile("讲义导航")})
for each in doc_menu.find_all("a"):
doc_name = each["href"].split("/")[-1]
doc_link = model.link_check("http://www.xuetangx.com", each["href"])
doc_file_path = model.generate_path([main_path, "docs", doc_name])
print("文档: \"{name}\" \"{link}\"".format(name=doc_name, link=doc_link))
doc_list.append((doc_link, doc_file_path))
# 下载
if config.Download:
if config.Download_Method == "Aria2": # 这里是调用aria2的下载
model.aira2_download(main_list + doc_list)
model.download_queue(session, srt_list, queue_length=config.Download_Queue_Length) # 需要session或者有时间期限的
else: # 默认调用自建下载
model.download_queue(session, main_list + srt_list + doc_list,queue_length=config.Download_Queue_Length)
else:
print("No course Id,Please check!")
return
def main():
conf = config()
conf.batch_size = 50
conf.num_steps = 50
train(conf,1000)
print(' Word 编号: {}'.format([i for i in answer_logits if i != pad]))
print(' Response Words: {}'.format(
" ".join([target_int_to_letter[i] for i in answer_logits if i != pad])))
if __name__ == '__main__':
with open(r'data/letters_source.txt', 'r', encoding='utf-8') as f:
source_data = f.read()
with open(r'data/letters_target.txt', 'r', encoding='utf-8') as f:
target_data = f.read()
print(source_data.split('\n')[:10])
print(target_data.split('\n')[:10])
# 构造映射表
source_int_to_letter, source_letter_to_int = extract_character_vocab(source_data)
target_int_to_letter, target_letter_to_int = extract_character_vocab(target_data)
# 对字母进行转换
source_int = [[source_letter_to_int.get(letter, source_letter_to_int['<UNK>'])
for letter in line] for line in source_data.split('\n')]
target_int = [[target_letter_to_int.get(letter, target_letter_to_int['<UNK>'])
for letter in line] + [target_letter_to_int['<EOS>']] for line in target_data.split('\n')]
config = config()
model = Seq2Seq(config, target_letter_to_int, source_letter_to_int)
train(config, model, source_int, target_int)
