DocumentDB从另一个存储过程或其本身调用存储过程 [英] DocumentDB call stored procedure from another stored procedure or itself

问题描述

是否有一种方法可以针对DocumentDB文档递归地调用存储过程(甚至可以使用UDF)?

Is there a way to recursively call stored procedures (or even UDFs if that will work) against DocumentDB documents?

我们有一个看起来像这样的文档:

We have a document that looks something like this:

{
  "docID" : "my_id",
  "owner" : "fred",
  "items" : [
    {
      "itemID" : "1",
      "type" : "item",
      "value" : 3
    },
    {
      "itemID" : "2",
      "type" : "group",
      "items" : [
        {
          "itemID" : "2.1",
          "type" : "group",
          "items" : [
            {
              "itemID" : "2.1.1",
              "type" : "item",
              "value" : 2
            },
            {
              "itemID" : "2.1.2",
              "type" : "item",
              "value" : 4
            }
          ]
        },
        {
          "itemID" : "2.2",
          "type" : "item",
          "value" : 1
        }
      ]
    }
  ]
}

任何时候只要有"items"，该"items"数组都可以包含由"type" : "item"和"type" : "group"混合而成的条目. "type" : "item"的条目具有一个简单的"value"字段，需要对其求和. "type" : "group"的条目具有"items"数组...等等.从理论上讲，递归级别没有限制，我承认这是一个问题，但是在实践中，递归级别很少会低于4或5.

Any time we have "items", that "items" array can contain entries that are a mix of "type" : "item" and "type" : "group". Entries that are of "type" : "item" have a simple "value" field that needs to be summed. Entries that are of "type" : "group" have an "items" array... and so on. In theory, there is no limit to the level of recursion, which I acknowledge is a problem, but in practice, the levels will rarely get below 4 or 5 deep.

我要写的伪代码看起来像这样:

Pseudocode for what I'm trying to write looks something like this:

function sumValues(items) {
    int total = 0;
    forEach(item in items) {
        if (item.type == "item") {
            total += item.value;
        } else {
            total += sumValues(item.items);
        }
    }
    return total;
}

function sumAllValues() {
  var ctx = getContext();
  var coll = ctx.getCollection();
  var response = ctx.getResponse();

  // query for docs by owner
  var filterQuery = 'SELECT * FROM Docs d where d.owner = \\\"fred\\\"';
  var done = coll.queryDocuments(coll.getSelfLink(), filterQuery, {},
    function (err, docs, options) {
      if (err) throw new Error ('Error' + err.message);

      var total = 0;
      docs.forEach(function(doc) {
        total += sumTotals(doc.items);
      });

      response.setBody('Total: ' + total);
    });
}

这甚至有可能吗? DocumentDB是否支持从另一个存储过程调用一个存储过程? sproc可以自己调用吗?

Is this even possible? Does DocumentDB support calling a sproc from another sproc? Can a sproc call itself?

我在网上找到了一些DocumentDB存储过程参考，包括此和此和这以及其他许多页面.

I have found a few DocumentDB stored procedure references online, including this and this and this and this plus a host of other pages.

如果可能的话，我想我可能必须以某种方式查询集合以获取要调用的sproc，然后以某种方式引用该sproc，而不是像使用独立语言那样直接调用sumTotals().

If it's possible, I think I might have to somehow query the collection to get the sproc that I want to call, and then somehow reference the sproc, instead of just calling sumTotals() directly as you would with a standalone language.

我们只是开始研究使用DocumentDB进行编程，因此我们还不确定是否可以使用它进行处理.感谢您的帮助或建议.

We're just starting to look at programming with DocumentDB, so we're not completely sure what we can do with it yet. Thanks for any assistance or advice.

推荐答案

我认为您在这里的路线正确.

I think you're on the right track here.

不可能从存储过程中执行存储过程.

It is not possible to execute a stored procedure from within a stored procedure.

但是，您可以在存储过程中定义JS函数，可以在该存储过程中对其进行引用，调用和重用.

However, you can define JS functions inside a stored procedure, which can be can be referenced, called, and re-used from within that stored procedure.

在这种情况下，只需在父sumAllValues()存储过程中定义您的sumValues()函数(就像您提到的swapItems()示例一样).

In this case, simply define your sumValues() function inside the parent sumAllValues() stored procedure (just like the swapItems() example you mentioned).

function sumAllValues() {
    var ctx = getContext();
    var coll = ctx.getCollection();
    var response = ctx.getResponse();

    // query for docs by owner
    var filterQuery = 'SELECT * FROM Docs d where d.owner = \\\"fred\\\"';
    var done = coll.queryDocuments(coll.getSelfLink(), filterQuery, {},
        function (err, docs, options) {
            if (err) throw new Error ('Error' + err.message);

            var total = 0;
            docs.forEach(function(doc) {
                total += sumValues(doc.items);
            });

            response.setBody('Total: ' + total);
        });

    function sumValues(items) {
        int total = 0;
        items.forEach(function(item) {
            if (item.type == "item") {
                total += item.value;
            } else {
                total += sumValues(item.items);
            }
        });
        return total;
    }
}

您还可以为要在多个存储过程和查询之间共享和重用的逻辑定义UDF.

You can also define UDFs for logic you'd like to share and re-use across multiple stored procedures and queries.

这篇关于DocumentDB从另一个存储过程或其本身调用存储过程的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！