利用IPFS构建短视频应用开发经历（二）

本系列文章是针对 https://blog.csdn.net/weixin_43668031/article/details/83962959 内容的实现所编写的。开发经历包括思考过程、重构和推翻重来。

在上一节中Video表（结构体）中含有 uint gratuitySum;字段，此字段冗余，可以通过实时计算获得，可以在结构体中删除。Video 设计如下

     struct Video {
        string title;
        string cover;
        string videoinfo;
        string info;
        uint duration;
        uint timestamp;
        address payable author;
        uint commentsNum;
        uint gratuityNum;
        uint lableNum;
        uint fileNum;
        uint8 permission;
        mapping (uint => Videofile) videofiles;
        mapping (uint => Lable) lables;
        mapping (uint => Gratuity) gratuitys;
        mapping (uint => Comment) comments;
    }
    uint public videoNum;
    mapping(uint => Video) videos;

0x3.智能合约

区块链的智能合约的操作方法更像是mysql dba编写的过程和函数，但是这是远远不够的，我们连传统数据库的insert、delete、update、select都还没有提供，就连最基本的crud也需要自己去实现。这里只要考虑像C语言那样去操作结构体即可，不需要考虑数据同步、分布式等技术，那是以太坊已经给我做好的铺垫。也因为如此，会在我的代码上看到大量set、get方法。

1.用户表

    function setMyInfo (string memory _nickname, string memory _profile, string memory _avatar) public {
        users[msg.sender] = User(_nickname, _profile, _avatar, 0);
    }
    function setMyNickname (string memory _nickname) public {
        users[msg.sender].nickname = _nickname;
    }
    function setMyProfile (string memory _profile) public {
        users[msg.sender].profile = _profile;
    }
    function setMyAvatar (string memory _avatar) public {
        users[msg.sender].avatar = _avatar;
    }
    function getUserInfo (address _userAdd) view public returns (string memory nickname, string memory profile, string memory avatar, uint videoNums) {
        return (users[_userAdd].nickname, users[_userAdd].profile, users[_userAdd].avatar, users[_userAdd].videoNums);
    }

2.视频存储方法

 function publish (string memory _title, string memory _cover, string memory _videoinfo, string memory _info, uint _duration, string memory _filename, string memory _fileinfo, uint _size, uint32 _width, uint32 _height, uint32 _fps) public returns (uint articleId) {
        videos[videoNum++] = Video(_title, _cover, _videoinfo, _info, _duration, now, msg.sender, 0, 0, 0, 0, 0);
        videos[videoNum].videofiles[videos[videoNum].fileNum++] = Videofile(_filename, _fileinfo, _size, _width, _height, _fps, 0);
        return videoNum-1;
	}
	function setVideoPermission(uint _videoId, uint8 _value) public {
	    if( msg.sender != videos[_videoId].author) return;
	    videos[_videoId].permission = _value;
	}
	function setVideo(uint _videoId, string memory _cover, string memory _videoinfo) public {
	    if( msg.sender != videos[_videoId].author) return;
	    videos[_videoId].cover = _cover;
	    videos[_videoId].videoinfo = _videoinfo;
	}
	function setVideoInfo(uint _videoId, string memory _title, string memory _info) public {
	    if( msg.sender != videos[_videoId].author) return;
    	videos[_videoId].title = _title;
    	videos[_videoId].info = _info;
	}
	function addVideofile(uint _videoId, string memory _filename, string memory _fileinfo, uint _size, uint32 _width, uint32 _height, uint32 _fps) public {
	    if( msg.sender != videos[_videoId].author){
	        videos[_videoId].videofiles[videos[_videoId].fileNum++] = Videofile(_filename, _fileinfo, _size, _width, _height, _fps, 1);
	    }else{
	        videos[_videoId].videofiles[videos[_videoId].fileNum++] = Videofile(_filename, _fileinfo, _size, _width, _height, _fps, 0);
	    }
	}
	function setFilePermission(uint _videoId, uint _fileId, uint8 _value) public {
	     if( msg.sender == videos[_videoId].author ){
	         videos[_videoId].videofiles[_fileId].filePermission = _value;
	     }
	}
	function getVideo (uint _videoId) view public returns (string memory title, string memory cover, string memory videoinfo,string memory info, uint timestamp, address author, uint commentsNum, uint videofiles, uint gratuitySum) {
	    if( videos[_videoId].permission != 0 && msg.sender != videos[_videoId].author ){
	         Video storage _video = videos[_videoId];
	         for (uint i = 0; i < _video.gratuityNum; i++)
	            gratuitySum += _video.gratuitys[i].gratuity;
	         return (_video.title, _video.cover,_video.videoinfo,_video.info, _video.timestamp, _video.author,_video.commentsNum, _video.fileNum, gratuitySum);
	    }
	}
	function getVideoFile (uint _videoId, uint _fileId) view public returns (string memory filename, string memory fileinfo,uint size, uint32 width, uint32 height, uint32 fps, uint fileNum){
	    if(videos[_videoId].permission != 0 && msg.sender == videos[_videoId].author || videos[_videoId].permission == 0 && videos[_videoId].videofiles[_fileId].filePermission ==0){
	        Video storage _video = videos[_videoId];
	        Videofile storage _videofile = _video.videofiles[_fileId];
	        return (_videofile.filename, _videofile.fileinfo, _videofile.size, _videofile.width, _videofile.height, _videofile.fps, _video.fileNum);
	    }else{
	         return ("", "", 0, 0, 0, 0, videos[_videoId].fileNum);
	    }
	}

这里会详细说明一下：
getVideoFile 方法中的判断，如果多视频资源有访问权限，并视频的文件也有访问权限，那就是可以访问的，如果视频无访问权限，但是是作者自己上传的视频，也可以访问，还要用于核审呢！

addVideofile方法中如果是作者本人再向视频资源中添加视频文件，那么默认是直接允许访问的，如果是他人向我的视频资源中添加视频文件，默认是不可以访问的，我需要去核审视频并再开放权限，允许这个资源访问。