SQL Server CLR聚合：序列化字典？ [英] SQL Server CLR Aggregate: Serialize a Dictionary?

问题描述

我正在用C＃编写自定义CLR聚合以在SQL Server 2008中运行。我想将我的大多数 SqlUserDefinedAggregate 实例的状态保存在字典中，所以我可以不要使用内置的序列化。

I'm writing a custom CLR Aggregate in C# to run in SQL Server 2008. I want to hold most of my SqlUserDefinedAggregate instances' states in Dictionaries, so I can't use the builtin serialization.

如何使用 Read（） UserDefined 序列化c $ c>和 Write（）方法？

How can I go about using UserDefined serialization with the Read() and Write() methods?

推荐答案

使用键以及 BinaryReader 和 BinaryWriter 处理的类型的值，将每个Dictionary写入key0，value0中的BinaryWriter中。 keyN-1，值N-1顺序。在字典的 Count 之前添加此序列的前缀。在序列之后，编写一个常数以验证字典的结尾。我个人喜欢 0xDEADBEEF 。

Using keys and values of types that BinaryReader and BinaryWriter handle, Write each Dictionary out to the BinaryWriter in key0,value0...keyN-1,valueN-1 order. Prefix this sequence with Dictionary's Count. After the sequence, write a constant to verify end of dictionary. Personally I like 0xDEADBEEF.

对于反序列化，请执行相反的操作：将循环初始化为运行n次（n为流中的第一个整数），然后在您从键/值对中读取时将其添加到字典中。最后，检查您的常数。

For deserialization, do the reverse: Intialize a loop to run n times (n is the first integer in your stream), and add to Dictionary as you read in key/value pairs. At the end, check for your constant.

这篇关于SQL Server CLR聚合：序列化字典？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！