您何时选择将XML存储在关系数据库中？[重复]-WhendoyouopttostoreXMLinarelationaldatabase?[duplicate]

3  

Never.

决不。

Storing data to which there is some kind of recognisable structure as XML, instead of as relational data, means abandoning the power of the relational algebra to operate on that data.

将存在某种可识别结构的数据存储为XML而不是关系数据意味着放弃关系代数的功能来操作该数据。

If that is your very intention, fine, but then be plain honest about it and just don't bother to use a relational DBMS at all, just dump your XML in some local/pseudolocal file.

如果这是你的意图，那很好，但是然后对它真诚，​​只是根本不打算使用关系DBMS，只需将XML转储到某个本地/伪局部文件中。

(Observe that anything stored in a filesystem constitutes a database too, so it's not like that solution means that you are "not using a database". You're just not using one that is managed relationally, which was your very intention to boot.)

（观察文件系统中存储的任何东西也构成了数据库，所以它并不像解决方案意味着你“不使用数据库”。你只是没有使用一个关系管理的，这是你打算引导的。 ）

2  

My experience with storing XML in a relational database is usually storing serialized objects for either historical purposes or as part of a persistence strategy with the assumption that I will later retrieve this XML and rehydrate it into an object.

我在关系数据库中存储XML的经验通常是为了历史目的或作为持久性策略的一部分存储序列化对象，并假设我稍后将检索此XML并将其重新水合为对象。

2  

The only reason, IMHO, to put XML in a database is if you need to interact, edit, or change the data with your project. If you are simply reading and using the XML Data, just throw it in a file. But if that data is being modified or added to with the application, its better to use a DB for that and you don't really need XML at all, since the table will describe the structure of the data.

恕我直言，将XML放入数据库的唯一原因是，您需要与项目进行交互，编辑或更改数据。如果您只是阅读和使用XML数据，只需将其放入文件即可。但是，如果正在修改该数据或将其添加到应用程序中，那么最好使用数据库，而您根本不需要XML，因为该表将描述数据的结构。

1  

Your code may want to process XML data in the following ways:

您的代码可能希望以下列方式处理XML数据：

1. Only persist it in the database, do all the processing in the business layer, e.g. a queue or audit trail of SOAP messages to/from a web service. For this I would use a simple text column.
2. 只在数据库中保留它，在业务层中进行所有处理，例如， SOAP消息到/来自Web服务的队列或审计跟踪。为此，我将使用一个简单的文本列。
3. Persist it in the database and run occasional queries in the database layer, e.g. an XHTML document that users could search. For this I would use the xml data type in SQL Server 2005 or above, or the equivalent in other DBMSs.
4. 将其保留在数据库中并在数据库层中偶尔运行查询，例如用户可以搜索的XHTML文档。为此，我将使用SQL Server 2005或更高版本中的xml数据类型，或其他DBMS中的等效类型。
5. Full-scale complex relational queries in the database layer, e.g. a blob of relational data that was merely serialised to XML and must now be stored in its proper form. This would obviously be stored in the set of tables that matched the deserialised data structure.
6. 数据库层中的全尺寸复杂关系查询，例如，一小段关系数据，仅仅序列化为XML，现在必须以适当的形式存储。这显然存储在与反序列化数据结构匹配的表集中。

0  

I tend to store xml blurbs (or other opaque data such as json serialized) when the data itself does not need to have relationships.

当数据本身不需要关系时，我倾向于存储xml blurbs（或其他不透明数据，如json序列化）。

0  

I use these conditions for storing XML data in a database.

我使用这些条件将XML数据存储在数据库中。

1. I don't have the time to develop the relational tables (really, I sometimes start off this way just to get a prototype up and running).

   我没有时间开发关系表（实际上，我有时会以这种方式开始，只是为了让原型启动并运行）。

2. The data is much to large and complex in structure to break into tables without considerable effort

   数据非常庞大且结构复杂，无需付出相当大的努力即可打入表格

In general I like to store relational data in my database, but imagine I have an object that represents all the data required to render a web page (I mean everything, fonts, images the lot), Its going to be incredibly complex, storing it in a table structure will serve to really only casue me maintenace issues - the structure is fluid. it would also cause me query issues - imagine the number joins I would need to do, and the time it would take.

一般来说，我喜欢在我的数据库中存储关系数据，但想象一下，我有一个对象代表渲染网页所需的所有数据（我的意思是一切，字体，图像），它会非常复杂，存储它在一个表格结构中将真正只是为了维护我的维护问题 - 结构是流动的。它也会引起我的查询问题 - 想象一下我需要做的数字连接，以及它需要的时间。