1.优酷上传
1）调用优酷的sdk完成优酷视频的上传首先需要将实例化YoukuUpload类实例化，传入的参数为(client_id,access_token,文件地址)
实例化时执行__init__（）
2）实例化完成后得到类的对象，通过对象调用upload方法，传入参数为字典（dict），字典内的必传参数为title，其余可为默认，其中的一些参数是为了控制视频一些信息的，具体参见代码的注释
3）在upload方法中
（1）会先判断upload_token 这个参数是否存在，该参数为优酷返回，存在就继续之前的上传，如果不存在的话就判断为新上传。
（2）在新上传中限制性create方法来在服务端创建上传，在此之前会对传进来的参数和默认的参数会放入新的字典中
def prepare_video_params(self, title=None, tags='Others', description='', copyright_type='original', public_type='all', category=None, watch_password=None, latitude=None, longitude=None, shoot_time=None ): # 准备视频所以需要的一些参数 params = {} if title is None: # 如果没有传title的话，title默认等于文件名 title = self.file_name elif len(title) > 50: # 如果title过长，长度大于50的话，就截取前50个字符作为title title = title[:50] params['title'] = title params['tags'] = tags # tags就等于你传进来的那个tags，默认为others params['description'] = description # 描述默认为空，可以传进来 params['copyright_type'] = copyright_type # 版权所有 original: 原创 reproduced: 转载 params['public_type'] = public_type # 公开类型 all: 公开 friend: 仅好友观看 password: 输入密码观看 private: 私有 if category: params['category'] = category # 视频分类 默认为空，可传 if watch_password: params['watch_password'] = watch_password # 观看密码，默认为空，可传 """ latitude/longitude用户记录地理位置信息 shoot_time用来标记视频中正片的开始时间 """ if latitude: params['latitude'] = latitude # 纬度 if longitude: params['longitude'] = longitude # 经度 if shoot_time: params['shoot_time'] = shoot_time log.debug("prepare_video_params:%s" % params) return params
（3）在create方法中，会用get方法访问url = 'https://api.youku.com/uploads/create.json'，并将参数传过去，创建oss客户端，优酷会返回一些字段
def create(self, params): # prepare file info params['file_name'] = self.file_name params['file_size'] = self.file_size # 在__init__中获取到了，也就是在类的实例化时就已经根据文件的地址获取了文件的大小 params['file_md5'] = self.file_md5 = self.checksum_md5file(self.file) # 将文件信息做md5 校验，根据文件名打开文件，然后每次读取8192B，进行md5更新后再转为十六进制返回 self.logger.info('upload file %s, size: %d bytes' % (self.file_name, self.file_size)) self.logger.info('md5 of %s: %s' % (self.file_name, self.file_md5)) params['client_id'] = self.client_id # client_id params['access_token'] = self.access_token # access_token url = 'https://api.youku.com/uploads/create.json' r = requests.get(url, params=params)# 以get方法将文件信息发送到'https://api.youku.com/uploads/create.json' check_error(r, 201) result = r.json() log.debug("file--->vid:%s,return_result:%s" % (self.v_vid, result)) self.upload_token = result['upload_token'] self.logger.info('upload token of %s: %s' % (self.file_name, self.upload_token)) self.upload_server_ip = socket.gethostbyname( result['upload_server_uri']) self.logger.info('upload_server_ip of %s: %s' % (self.file_name, self.upload_server_ip)) log.debug("file_vid:%s ip:%s" % (self.v_vid, self.upload_server_ip))
(4) 调用create_file方法，将文件的大小、token、每次上传切片大小传到上一步返回的ip所指向的服务器
(5) 调用new_slice方法 告诉服务器才准备上传文件切片，目的在于检查服务器状态和返回服务器中这个新建切片的信息，没有报错就执行_save_slice_state方法
(6)在_save_slice_state方法中，将切片的信息进行更新或者保存，之后判断返回的task_id是否为0，如果为零就直接提交commit完成整个上传，不为0就 调用upload_slice方法上传文件切片
(7)在upload_slice方法中，每次打开文件移动指针位置到上一次上传的地方，然后读取2048kb的数据，之后调用url进行上传。
def upload_slice(self): # 上传文件切片 seek():移动文件读取指针到指定位置 data = None with open(self.file, 'rb') as f: f.seek(self.slice_offset) data = f.read(self.slice_length) # 2048 params = { 'upload_token': self.upload_token, 'slice_task_id': self.slice_task_id, 'offset': self.slice_offset, 'length': self.slice_length, # Byte 'hash': self.checksum_md5data(data)# hash这个字段是为了在服务端进行校验，来验证上传的文件有没有出现丢失或错误 } url = 'http://%s/gupload/upload_slice' % self.upload_server_ip r = requests.post(url, params=params, data=data) check_error(r, 201) self._save_slice_state(r.json())
(8) 循环判断task_id，若task_id为0则结束上传(即调用commit())
整体流程:应该是先本地将文件的信息上传到服务器，在服务器创建同名文件，然后根据文件的size和每次对文件进行切片的大小控制上传，在上传过程中服务端会先返回这次要上传的切片的task_id，如果在服务端的文件大小等于你上传的参数中提交的size，就会返回task_id为0，否则就不为0。本地会根据这个参数来判断是否结束上传。
2. 完整优酷上传sdk代码
"""Youku cloud Python Client doc: http://cloud.youku.com/docs/doc?id=110 """ import os import requests import json import time import hashlib import socket import logging from time import sleep from util import check_error, YoukuError if not os.path.exists('/var/log/youku/'): os.makedirs('/var/log/youku') log = logging.getLogger() log.setLevel(logging.DEBUG) fmt = logging.Formatter("%(asctime)s %(pathname)s %(filename)s %(funcName)s %(lineno)s %(levelname)s - %(message)s", "%Y-%m-%d %H:%M:%S") stream_handler = logging.FileHandler( '/var/log/youku/debug-%s.log' % (time.strftime('%Y-%m-%d', time.localtime(time.time())))) stream_handler.setLevel(logging.DEBUG) stream_handler.setFormatter(fmt) log.addHandler(stream_handler) class YoukuUpload(object): """Youku Upload Client. Upload file to Youku Video. Support resume upload if interrupted. Should use one instance of YoukuUpload for one upload file in one thread, since it has internal state of upload process. doc: http://cloud.youku.com/docs/doc?id=110 """ def __init__(self, client_id, access_token, file, v_vid = None, logger=None): """ Args: file: string, include path and filename for open(). filename must contain video file extension. """ super(YoukuUpload, self).__init__() self.client_id = client_id self.access_token = access_token self.v_vid = v_vid # 上传的视频id号 self.logger = logger or logging.getLogger(__name__) # file info self.file = file self.file_size = os.path.getsize(self.file) # int 获取文件的大小 self.file_dir, self.file_name = os.path.split(self.file) # string 分割路径来获取文件名 if self.file_dir == '': self.file_dir = '.' self.file_ext = self.file_name.rsplit('.', 1)[1], # file extension self.file_md5 = None # string, do later # upload state self.upload_token = None # string self.upload_server_ip = None # string self.slice_task_id = None # int self.slice_offset = None # int self.slice_length = None # int self.transferred = None # int for bytes has uploaded self.finished = False # boolean # resume upload state self._read_upload_state_from_file() def prepare_video_params(self, title=None, tags='Others', description='', copyright_type='original', public_type='all', category=None, watch_password=None, latitude=None, longitude=None, shoot_time=None ): # 准备视频所以需要的一些参数 """ util method for create video params to upload. Only need to provide a minimum of two essential parameters: title and tags, other video params are optional. All params spec see: http://cloud.youku.com/docs?id=110#create . Args: title: string, 2-50 characters. tags: string, 1-10 tags joind with comma. description: string, less than 2000 characters. copyright_type: string, 'original' or 'reproduced' public_type: string, 'all' or 'friend' or 'password' watch_password: string, if public_type is password. latitude: double. longitude: double. shoot_time: datetime. Returns: dict params that upload/create method need. """ params = {} if title is None: # 如果没有传title的话，title默认等于文件名 title = self.file_name elif len(title) > 50: # 如果title过长，长度大于50的话，就截取前50个字符作为title title = title[:50] params['title'] = title params['tags'] = tags # tags就等于你传进来的那个tags，默认为others params['description'] = description # 描述默认为空，可以传进来 params['copyright_type'] = copyright_type # 版权所有 original: 原创 reproduced: 转载 params['public_type'] = public_type # 公开类型 all: 公开 friend: 仅好友观看 password: 输入密码观看 private: 私有 if category: params['category'] = category # 视频分类 默认为空，可传 if watch_password: params['watch_password'] = watch_password # 观看密码，默认为空，可传 """ latitude/longitude用户记录地理位置信息 shoot_time用来标记视频中正片的开始时间 """ if latitude: params['latitude'] = latitude # 纬度 if longitude: params['longitude'] = longitude # 经度 if shoot_time: params['shoot_time'] = shoot_time log.debug("prepare_video_params:%s" % params) return params def create(self, params): # prepare file info params['file_name'] = self.file_name params['file_size'] = self.file_size # 在__init__中获取到了，也就是在类的实例化时就已经根据文件的地址获取了文件的大小 params['file_md5'] = self.file_md5 = self.checksum_md5file(self.file) # 将文件信息做md5 校验，根据文件名打开文件，然后每次读取8192B，进行md5更新后再转为十六进制返回 self.logger.info('upload file %s, size: %d bytes' % (self.file_name, self.file_size)) self.logger.info('md5 of %s: %s' % (self.file_name, self.file_md5)) params['client_id'] = self.client_id # client_id params['access_token'] = self.access_token # access_token url = 'https://api.youku.com/uploads/create.json' r = requests.get(url, params=params)# 以get方法将文件信息发送到'https://api.youku.com/uploads/create.json' check_error(r, 201) result = r.json() log.debug("file--->vid:%s,return_result:%s" % (self.v_vid, result)) self.upload_token = result['upload_token'] self.logger.info('upload token of %s: %s' % (self.file_name, self.upload_token)) self.upload_server_ip = socket.gethostbyname( result['upload_server_uri']) self.logger.info('upload_server_ip of %s: %s' % (self.file_name, self.upload_server_ip)) log.debug("file_vid:%s ip:%s" % (self.v_vid, self.upload_server_ip)) def _save_upload_state_to_file(self): """if create and create_file has execute, save upload state to file for next resume upload if current upload process is interrupted. """ # 保存文件的上传信息，先判断文件是否可写、可读、可执行 # 保存的信息包括文件的上传upload_token.上传到的服务器ip # 保存的文件会在上传完毕之后删除 if os.access(self.file_dir, os.W_OK | os.R_OK | os.X_OK): save_file = '/tmp' + 'youku.upload' data = { 'upload_token': self.upload_token, 'upload_server_ip': self.upload_server_ip } with open(save_file, 'w') as f: json.dump(data, f) def _read_upload_state_from_file(self): save_file = '/tmp' + 'youku.upload' try: with open(save_file) as f: data = json.load(f) self.upload_token = data['upload_token'] self.upload_server_ip = data['upload_server_ip'] # check upload_token expired try: self.check() except YoukuError, e: if e.code == 120010223: # Expired upload token self.upload_token = None self.upload_server_ip = None self._delete_upload_state_file() except: pass def _delete_upload_state_file(self): try: os.remove('/tmp' + 'youku.upload') except: pass def checksum_md5file(self, filename): md5 = hashlib.md5() with open(filename, 'rb') as f: for chunk in iter(lambda: f.read(8192), b''): md5.update(chunk) return md5.hexdigest() def checksum_md5data(self, data): md5 = hashlib.md5() md5.update(data) return md5.hexdigest() def create_file(self): params = { 'upload_token': self.upload_token, 'file_size': self.file_size, # Byte 'ext': self.file_ext, 'slice_length': 2048 # KB } # 上传文件每次传2048KB url = 'http://%s/gupload/create_file' % self.upload_server_ip r = requests.post(url, data=params) check_error(r, 201) # save upload state to resume upload self._save_upload_state_to_file() def new_slice(self): params = { 'upload_token': self.upload_token } url = 'http://%s/gupload/new_slice' % self.upload_server_ip r = requests.get(url, params=params) check_error(r, 201) self._save_slice_state(r.json()) def _save_slice_state(self, result): # 更新切片状态 self.slice_task_id = result['slice_task_id'] self.slice_offset = result['offset'] self.slice_length = result['length'] self.transferred = result['transferred'] self.finished = result['finished'] def upload_slice(self): # 上传文件切片 seek():移动文件读取指针到指定位置 data = None with open(self.file, 'rb') as f: f.seek(self.slice_offset) data = f.read(self.slice_length) # 2048 params = { 'upload_token': self.upload_token, 'slice_task_id': self.slice_task_id, 'offset': self.slice_offset, 'length': self.slice_length, # Byte 'hash': self.checksum_md5data(data) } url = 'http://%s/gupload/upload_slice' % self.upload_server_ip r = requests.post(url, params=params, data=data) check_error(r, 201) self._save_slice_state(r.json()) def check(self): params = { 'upload_token': self.upload_token } url = 'http://%s/gupload/check' % self.upload_server_ip r = requests.get(url, params=params) check_error(r, 200) return r.json() def commit(self): status = self.check()# 检查上传状态 if status['status'] == 4: raise ValueError('upload has not complete, should not commit') while status['status'] != 1: # status is 2 or 3 sleep(10) status = self.check() params = { 'access_token': self.access_token, 'client_id': self.client_id, 'upload_token': self.upload_token, 'upload_server_ip': status['upload_server_ip'] } url = 'https://api.youku.com/uploads/commit.json' r = requests.post(url, data=params) check_error(r, 200) self.finished = True self._delete_upload_state_file()# 删除记录视频上传信息的文件 log.debug("sdk---->vid:%s youku video_id:%s" % (self.v_vid, r.json()['video_id'])) return r.json()['video_id'] def cancel(self): status = self.check() params = { 'access_token': self.access_token, 'client_id': self.client_id, 'upload_token': self.upload_token, 'upload_server_ip': status['upload_server_ip'] } url = 'https://api.youku.com/uploads/cancel.json' r = requests.get(url, params=params) check_error(r, 200) self._delete_upload_state_file() return r.json()['upload_token'] def spec(self): url = 'https://api.youku.com/schemas/upload/spec.json' r = requests.get(url) check_error(r, 200) return r.json() def transferred_percent(self): """return current transferred percent """ return int(self.transferred / self.file_size) def upload(self, params={}): """start uploading the file until upload is complete or error. This is the main method to used, If you do not care about state of process. Args: params: a dict object describe video info, eg title, tags, description, category. all video params see the doc of prepare_video_params. Returns: return video_id if upload successfully """ if self.upload_token is not None: # resume upload status = self.check() if status['status'] != 4: return self.commit() else: self.new_slice() while self.slice_task_id != 0: self.upload_slice() return self.commit() else: # new upload try: log.debug('youku upload params:%s' % params) # 记录上传参数 except Exception as e: pass self.create(self.prepare_video_params(**params)) # 创建上传 self.create_file() self.new_slice() while self.slice_task_id != 0: self.upload_slice() return self.commit()
扫一扫
228
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
