提问者:小点点

Mongoose save()在findOne()之前返回


我正在使用Mongoose,Express和NodeJS实现一个审批系统。

每个岗位有2套批文。

我实现了如下模型:

Post Schema:
var PostSchema = new Schema({
UserId: { type: ObjectId, unique: true, default: null },
/..
.. Object related fields
../
    Approval1: Object, //chidschema objects of approver
    Approval2: Object
});

Approval Mapping Schema:
var ApprovalMapSchema = new Schema({
  UserId: { type: ObjectId, unique: true, default: null },
  Approver1Id: { type: ObjectId, default: null },
  Approver2Id: { type: ObjectId, default: null }
});

我试图首先在我的函数中找到映射,然后在保存时更新主post对象。

我需要在保存新帖子之前检索两个审批者ID。

请提出一个好办法来做这件事。我目前正在做:

>

  • 保存新的Post对象并保存//在第二步之后返回,我知道在异步中保存在Mongoose中,但我不希望在此之前调用步骤2
  • 在ApprovalMapSchema上使用findOne获取审批者ID
  • 使用审批者条目更新Post对象。

    post.save(function(err,doc){if(err){}else{console.log({success:true,msg:'successful created new post‘,content:doc});}});

    /*var actId = activation._id, distributorId, ttlId;
    
    var approvalMap = ApprovalMap.findOne({ userId: myUserId }, function(err, appMap) {
        if (err) throw err;
        else{           
            Approver1Id = appMap.Approver1Id;
            Approver2Id = appMap.Approver2Id;
        }
    });
    
    // Create child objects after this and update as subdocument
    

    这对我不起作用。请救命!


  • 共2个答案

    匿名用户

    我很难理解你的问题。但听起来您想要做这样的事情(您需要根据您想要做的事情对其进行修改,但希望它能向您展示一种处理异步行为的方法):

        //This will run a query that returns a document where userId == myUserId
    
        ApprovalMap.findOne({ userId: myUserId }, function(err, appMap) {
    
            if (err) throw err;
    
            //Once the query is complete function(err, appMap) will be called
            //err will contain an error object if there was an error
            //appMap will be the data returned by the query.
            //within this block you can run another query
            //for example
    
            ApprovalMap.findOne({approved:'some variable'},function(err, secondQueryData){
    
                if (err) throw err;
    
                //after this query is complete function(err, secondQueryData) will be called
                //once again the err argument will be an error, and the 
    
                //secondQueyrData argument will contain your return data
    
                //here you can do something with the data
                //for example, log the results
    
                console.log(secondQueryData)
            })
        });
    

    如您所见,可以在其他查询的回调中嵌套其他查询或步骤。这将确保事情以正确的顺序运行。您可能还想签出npm异步库,或者像q或Bluebird这样的promise库。他们也有一些很好的解决方案来处理Node中的异步行为。

    匿名用户

    根据@User2263572的建议,选择异步库。

    我现在就是这样实现的:

    async.waterfall(
        [
            function(callback){
                ApprovalMap.findOne({ userId: myUserId }, function(err, appMap) {
                    if (err) throw err;
                    else{           
                        //..Create child approval elements and assign Ids from map
                        callback(err, Approver1Child, Approver2Child);
                    }
                });
            },
            function(Approver1Child, Approver2Child, callback){
                var post = new Post(req.body);
                post.distApproval = Approver1Child;
                post.ttlApproval = Approver2Child;
                post.userId = myUserId;
                callback(null, post);
            }
        ], 
        function(err, post){
            //creating new order
            post.save(function(err, doc) {
                if (err) {
                }
                else
                {
                    console.log({success: true, msg: 'Successful created new post', content: doc});
                }
            });
        }
    );
    

    非常感谢@user2263572!!干杯!