def listen_ss(self, p: str, time_str: str, interval_sec: int = 60):
"""
此函数独立计算等待时间,不与BangumiController.py使用.
预定时间来滚动获取新番的最新弹幕.
在初始化时使用同一季的任意一集的url即可.
如果是已经发布的集数, time_str填写当前时间即可.
填写时间的目的是为了减少不必要的检查"目标集数是否可用"的次数, 减少被ban的概率.
对于未来的集数, 脚本会在到时间之后启动, 即使番剧推迟几分钟公开也不会报错.
在检测到相应的剧集可以观看时开始获取弹幕.
通过计算相邻两次获取的弹幕增量, 动态调整获取弹幕的时间间隔.
:param p: 视频分p,即集数,未发布也可(只要在初始化时是处于同一个系列的就可以"
:param time_str: 视频更新时间,格式为 "yyyy-mm-ddThh:mm",例如 "2020-01-02T03:04"
:param interval_sec: 每次获取间隔的初始时间, 时间 > 10秒
:return:
"""
target_time = Converter.str_to_timestamp(time_str)
interval_sec = max(interval_sec, 11)
sec_wait = max(11, target_time - int(time.time()))
print("wait:", sec_wait, "seconds")
time.sleep(sec_wait - 10)
# 循环监测视频是否可用
while True:
url = "https://www.bilibili.com/bangumi/play/" + self.ssid
response = Spider.get_html(url)
ep_json = self.get_epinfo_in_html(response)
new_series = ep_json['epList']
if len(new_series) >= int(p):
print("符合条件开始获取")
time.sleep(5)
target_ep = new_series[int(p)-1]["id"]
new_url = "https://www.bilibili.com/bangumi/play/ep" + str(target_ep)
self._get_info_ep(new_url)
break
print("未找到相应剧集,等待", interval_sec, "秒")
time.sleep(interval_sec)
previous_danmu = None
while True:
content_bytes = Spider.get_current_danmu(self.cid, self.url)
now = datetime.fromtimestamp(time.time(), timezone(timedelta(hours=8))).strftime('%Y-%m-%d %H:%M:%S')
print(now, "获取了弹幕")
with open(self.fileName + '_latest_' + str(int(time.time())) + '.xml', 'wb') as f:
f.write(content_bytes)
danmu = DanmuFile.init_from_str(content_bytes.decode('utf-8'))
if previous_danmu is not None:
_, inc, _ = DanmuCombinator.diff(previous_danmu, danmu)
ratio = len(inc) / int(danmu.max_limit)
print("时间比例:", ratio, )
if ratio > 0.5:
interval_sec = int(interval_sec / 5)
print("时间间隔修改为:", interval_sec)
if ratio < 0.3:
interval_sec = min(int(interval_sec * 1.5), 1800)
print("时间间隔修改为:", interval_sec)
previous_danmu = danmu
time.sleep(int(interval_sec))
def listen_ss_once(self):
content_bytes = Spider.get_current_danmu(self.cid, self.url)
if content_bytes is None:
return -1
now = datetime.fromtimestamp(time.time(), timezone(
timedelta(hours=8))).strftime('%Y-%m-%d %H:%M:%S')
print('[TASK]', now, "获取:[", self.title, "]")
with open(self.fileName + '_latest_' + str(int(time.time())) + '.xml',
'wb') as f:
f.write(content_bytes)
danmu = DanmuFile.init_from_str(content_bytes.decode('utf-8'))
ratio = -1
if self.danmu_set is not None:
dep, inc, com = DanmuCombinator.diff(self.danmu_set, danmu)
dep_int, inc_int, com_int = len(dep), len(inc), len(com)
print("[TASK] 原有弹幕数[",
dep_int + com_int,
"], 新增弹幕数[",
inc_int,
end=' ], ')
ratio = inc_int / int(danmu.max_limit)
print("算得新增比例:[", format(ratio, '0.5f'), "]")
else:
print("[TASK]首次获取")
self.danmu_set = danmu
self.timeProgress = int(time.time())
return ratio
def _get_current_danmu(self):
content_bytes = Spider.get_current_danmu(self.cid, self.url)
if content_bytes is None:
return False
# 将要与历史弹幕整合的弹幕文件
with open(self.fileName+'.xml', 'wb') as f:
f.write(content_bytes)
# 当前弹幕池(上限数量)的弹幕
with open(self.fileName+'_latest.xml', 'wb') as f:
f.write(content_bytes)
return True
def _get_info_av(self, url: str):
html = Spider.get_html(url)
pattern = re.compile(r'"cid":(\d+),"page":%s' % self.page)
pattern1 = re.compile(r'"title":"(.*?)","pubdate":(\d+)')
self.cid = re.search(pattern, html).group(1)
self.title, timeUnix_str = re.search(pattern1, html).groups()
self.timeUnix = int(timeUnix_str)
folder = "harvest/" + self.no + '_' + DanmuMaster.process_filename(self.title) + "/"
if not os.path.exists(folder):
os.mkdir(folder)
file_name = DanmuMaster.process_filename(self.no + '_' + self.title + '_p' + self.page)
self.fileName = folder + file_name
def _get_history_danmu(self, date: str):
"""
send history danmu request in specific date_str, return xml_str
:param date: date string in 'YYYY-MM-DD' format
:return: xml string in UTF-8 encoding
"""
content_bytes = Spider.get_history_danmu(self.cid, self.url, date, self.cookie_path)
xml_str = content_bytes.decode('utf-8')
with open(self.fileName + '_' + date + '.xml', 'wb') as f:
f.write(content_bytes)
print('data length', len(xml_str))
return xml_str
def check_ep_exist(self):
response = Spider.get_html(self.url)
if response is None:
print("[WARNING] 未获取到:", self.title)
return False
ep_json = self.get_epinfo_in_html(response)
ep_int = int(self.no[2:])
new_series = ep_json['epList']
for ep in new_series:
if ep['id'] == ep_int:
self.init_from_ep_json(ep_json, ep_int, self.cookie_path)
print("[TASK] 新剧集:", ep_json['h1Title'], "已放出")
return True
return False
def _get_info_ep(
self,
url: str,
):
html = None
none_times = 0
while html is None and none_times < 5:
html = Spider.get_html(url)
none_times += 1
if html is None:
print("多次请求番剧信息失败")
exit(1)
ep_json = self.get_epinfo_in_html(html)
self._resolve_ep_json(ep_json)
def _get_current_danmu(self):
content_bytes = None
none_count = 0
while content_bytes is None and none_count < 5:
content_bytes = Spider.get_history_danmu(self.cid, self.url, date,
self.cookie_path)
none_count += 1
if content_bytes is None:
print("多次请求失败.")
return False
# 将要与历史弹幕整合的弹幕文件
with open(self.fileName + '.xml', 'wb') as f:
f.write(content_bytes)
# 当前弹幕池(上限数量)的弹幕
with open(self.fileName + '_latest.xml', 'wb') as f:
f.write(content_bytes)
return True
def _get_info_ep(self, url: str,):
html = Spider.get_html(url)
ep_json = self.get_epinfo_in_html(html)
self._resolve_ep_json(ep_json)
def _get_history_month(self, month: str):
content_bytes = Spider.get_history_month(self.cid, self.url, month, self.cookie_path)
json_str = content_bytes.decode('utf-8')
return json_str