async def old_version_download_callback(self, response):
log.debug("Old version download callback")
file_path = response.request.file_path
metadata = response.request.metadata
if file_path.endswith("video.flv"):
order = file_path.split("/")[-1][0]
log.info("视频片段[{}]下载完毕: AV<{}>".format(order, metadata["id"]))
self.old_version_records[metadata["id"]]["fragments"].append(
file_path)
if len(self.old_version_records[metadata["id"]]["fragments"]) == len(
metadata["video_info"]):
fragments_path = self.old_version_records[
metadata["id"]]["fragments"]
fragments_path = sorted(fragments_path,
key=lambda x: int(x.split("/")[-1][0]))
ts_path_list = []
for fragment_path in fragments_path:
ts_path = fragment_path[:-3] + "ts"
os.system(
"ffmpeg -i {flv_path} -y -loglevel warning -c copy -bsf:v h264_mp4toannexb -f mpegts {ts_path}"
.format(flv_path=fragment_path, ts_path=ts_path))
ts_path_list.append(ts_path)
output_path = os.path.join(metadata["storage_dir"],
"{}.mp4".format(metadata["video_name"]))
os.system(
'ffmpeg -i "concat:{concat_param}" -y -loglevel warning -c copy -bsf:a aac_adtstoasc -movflags +faststart {output_path}'
.format(concat_param="|".join(ts_path_list),
output_path=output_path))
for temp in ts_path_list + fragments_path:
os.remove(temp)
log.info("视频分片合并完毕: AV<{}>".format(metadata["id"]))
async def handle_request(self, request):
log.debug("Handle request {}".format(request))
try:
if isinstance(request, FileRequest):
response = await self.download(request)
else:
response = await self.fetch(request)
# 注意：callback方法必须是一个awaitable对象
response_processed_result = request.callback(response)
if isinstance(response_processed_result, typing.AsyncGenerator):
async for result in response_processed_result:
if isinstance(result, Request):
await self.add_request(result)
else:
result = await response_processed_result
if isinstance(result, Request):
await self.add_request(result)
except Exception as e:
log.exception(e)
log.error("Catch exception: {} {}".format(e, request))
self.failed_request_count += 1
else:
log.debug("Fetched response from {}".format(request))
self.finished_request_count += 1
def _register_all_service():
from media_downloader.application.manager import CrawlerManager
from media_downloader.application.manager import DownloaderManager
from media_downloader.application.service import DownloadingService
Mapper.register(DownloadingService, CrawlerManager(),
DownloaderManager())
log.debug("Inversion Of Controller Registered DownloadingService")
def _register_all_manager():
from media_downloader.application.manager import CrawlerManager
from media_downloader.application.manager import DownloaderManager
Mapper.register(CrawlerManager)
log.debug("Inversion Of Controller Registered CrawlerManager")
Mapper.register(DownloaderManager)
log.debug("Inversion Of Controller Registered DownloaderManager")
async def fetch(self, request: Request):
'''根据指定的请求对象发起HTTP请求'''
log.debug("Send request {}".format(request))
try:
response = await self.http_client.fetch(request)
except HTTPClientError as e:
return Response(e.response, request)
else:
return Response(response, request)
async def status_print_loop(self):
while True:
running_time = int(time.time()) - self.start_time
if running_time > 0 and running_time % 60 == 0:
log.info(self.get_running_messages())
await asyncio.sleep(1)
if self.QUIT is True:
log.info(self.get_running_messages())
break
log.debug("Quit status print loop")
async def run(self, downloader_name=None):
downloader_name = downloader_name if downloader_name else ""
log.debug("Run downloader {}".format(downloader_name))
while self.is_running:
try:
request = self.request_queue.get_nowait()
except QueueEmpty:
await asyncio.sleep(self.__polling_interval)
else:
await self.handle_request(request)
if self.pushed_request_count == self.finished_request_count + self.failed_request_count:
self.is_running = False
await asyncio.sleep(self.delay)
log.debug("Quit downloader {}".format(downloader_name))
async def task_scheduling_loop(self):
while True:
try:
crawler_name, crawler_params = self.schedule_queue.get_nowait()
except QueueEmpty:
await asyncio.sleep(1)
else:
crawler = self.crawler_manager.create_crawler(
crawler_name, crawler_params)
downloader = self.downloader_manager.get_or_create_downloader(
crawler)
await self.downloader_manager.schedule(crawler, downloader)
finally:
if self.QUIT and self.schedule_queue.qsize() == 0:
break
log.debug("Quit task scheduling loop")
async def download(self, request: FileRequest):
log.debug("Downloading '{}'".format(request.file_path))
actual_length = 0
file = open(request.file_path, "wb")
def write_data(data):
nonlocal actual_length
actual_length += len(data)
file.write(data)
tornado_request = Request(
url=request.url,
callback=None,
headers=request.headers,
streaming_callback=write_data,
request_timeout=60*10,
connect_timeout=60*10
)
tornado_response = await self.http_client.fetch(tornado_request)
file.close()
if actual_length == int(tornado_response.headers["Content-Length"]):
log.debug("Success download '{}'".format(request.file_path))
else:
raise Exception("Error download '{}': content-length error".format(request.file_path))
return FileResponse(tornado_response, request, file_size=actual_length, file_path=request.file_path)
def callback(_):
log.debug("Downloading Mission Done!")
self.QUIT = True
async def detail_callback(self, response):
media_headers = deepcopy(self.media_headers)
media_headers[
"Referer"] = response.request.url # 设置media下载请求头的Referer地址
metadata = response.request.metadata
if "DIRECTLY_PARSING" in metadata:
content = json.loads(response.body)
log.info("详情页解析成功：[{}]".format(response.request.url))
log.debug(metadata)
async for request in self.content_parser(content, metadata,
media_headers):
yield request
else:
# _content = re.search(r"<script>window.__playinfo__=(.*?)</script>", response.body.decode())
# if _content is None:
_videoData = re.search(r"\"videoData\":(.*?),\"upData\"",
response.body.decode())
if _videoData is None:
# 如果没有匹配到任何数据，记录日志，并直接退出
log.warning("详情页解析失败: <{}>[{}]".format(response.code,
response.request.url))
else:
videoData = json.loads(_videoData.group(1))
storage_dir = os.path.join(
metadata["base_storage_dir"], "av{}-{}".format(
metadata["av_number"],
videoData["title"].replace("/", "-").replace(" ", "")))
if not os.path.exists(storage_dir):
os.mkdir(storage_dir)
metadata["storage_dir"] = storage_dir
content_base_url = "https://api.bilibili.com/x/player/playurl?avid={aid}&cid={cid}&qn=0&type=&otype=json&fnver=0&fnval=16"
pages = videoData["pages"]
if len(pages) == 1:
url = content_base_url.format(aid=videoData["aid"],
cid=videoData["cid"])
metadata[
"DIRECTLY_PARSING"] = "标示该请求是通过提取 videoData 中的 aid 和 cid 来发起单独的请求解析到 JSON Content 的"
metadata["video_name"] = videoData["title"].replace(
"/", "-").replace(" ", "")
metadata["id"] = "{}-{}".format(videoData["aid"],
videoData["cid"])
yield Request(
url,
headers=media_headers,
callback=self.detail_callback,
metadata=metadata,
# proxy_host=PROXY_HOST,
# proxy_port=PROXY_PORT,
# proxy_username=PROXY_USERNAME,
# proxy_password=PROXY_PASSWORD,
)
else:
for page in pages:
url = content_base_url.format(aid=videoData["aid"],
cid=page["cid"])
metadata[
"DIRECTLY_PARSING"] = "标示该请求是通过提取 videoData 中的 aid 和 cid 来发起单独的请求解析到 JSON Content 的"
metadata["video_name"] = "{}-{}".format(
page["page"],
page["part"].replace("/", "-").replace(" ", ""))
metadata["id"] = "{}-{}".format(
videoData["aid"], page["cid"])
yield Request(
url,
headers=media_headers,
callback=self.detail_callback,
metadata=deepcopy(metadata),
# proxy_host=PROXY_HOST,
# proxy_port=PROXY_PORT,
# proxy_username=PROXY_USERNAME,
# proxy_password=PROXY_PASSWORD,
)
