如何获取Oracle XPath表达式中父元素的名称? [英] How to get the name of the parent element in an Oracle XPath expression?

问题描述

我有一个查询，该查询从我的Clob中提取了一些XML节点.

I have a query that extract me some the XML nodes from my clob.

select pid, name, xml from (
select d.pid, d.name
, EXTRACT(xmltype(d.data), '//ns1:myId', 'xmlns:ns1="http://acme.com/') xml
from DATA d
order by d.pid desc
)
;

但是我想看看提取的xmlnode的父元素名称实际上是什么.我尝试过

But I'd like to see what the parent elements name of the extracted xmlnode actually is. I tried

, EXTRACT(xmltype(d.data), '//ns1:myId/../name()', ...) xml

和

, EXTRACT(xmltype(d.data), '//ns1:myId/name()', ...) xml

和

, EXTRACT(xmltype(d.data), '//ns1:myId/local-name()', ...) xml

但是Oracle拒绝所有带有无效令牌"错误消息的消息.

but Oracle rejects all f them with "invalid token" error messages.

我的Oracle版本是"11.2.0.3.0".

My Oracle Version is "11.2.0.3.0".

推荐答案

通过提取，您可以向上查找路径，但是(如MOS文档301709.1所述；它适用于9i，但除显示的错误外仍有效) :

Extract lets you look higher up the path, but (as noted in MOS document 301709.1; which is for 9i but still seems valid, apart from the error shown):

无法使用XPATH函数name()返回元素名称，因为方法extract()和extractValue()当前仅支持返回节点集的XPATH操作.

Using the XPATH function name() to return the element name is not possible, because the methods extract() and extractValue() currently only support XPATH operations returning a node-set.

因此，您不能在当前或父节点上使用name()，local-name()等函数.该文档中提供了一种变通方法，可以将其示例稍微简化为:

So you can't use the name(), local-name() etc. functions, on the current or parent node. There is a workaround of sorts in that document, which can be simplified slightly rom their example to:

EXTRACT(xmltype(d.data), '//ns1:myId/..', 
  'xmlns:ns1="http://acme.com/').getRootElement() xml2

或以稍微不同的形式:

xmltype(d.data).extract('//ns1:myId/..', 
  'xmlns:ns1="http://acme.com/').getRootElement()

两者都显示的演示

with data (pid, name, data) as (
  select 42, 'Test', '<?xml version="1.0" encoding="ISO-8859-1"?>
  <ns1:root xmlns:ns1="http://acme.com/"><ns1:parent><ns1:myId>1234</ns1:myId></ns1:parent></ns1:root>' from dual
)
select d.pid, d.name
, EXTRACT(xmltype(d.data), '//ns1:myId', 'xmlns:ns1="http://acme.com/') xml
, EXTRACT(xmltype(d.data), '//ns1:myId/..', 'xmlns:ns1="http://acme.com/').getRootElement() xml2
, xmltype(d.data).extract('//ns1:myId/..', 'xmlns:ns1="http://acme.com/').getRootElement() xml3
from DATA d
order by d.pid desc
;

       PID NAME XML                            XML2       XML3     
---------- ---- ------------------------------ ---------- ----------
        42 Test <ns1:myId xmlns:ns1="http://ac parent     parent    
                me.com/">1234</ns1:myId>                      

但是，不建议使用 extract() .您可以使用 XMLTable :

with data (pid, name, data) as (
  select 42, 'Test', '<?xml version="1.0" encoding="ISO-8859-1"?>
  <ns1:root xmlns:ns1="http://acme.com/"><parent><myId>1234</myId></parent></ns1:root>' from dual
)
select d.pid, d.name, x.*
from data d
cross join xmltable(xmlnamespaces('http://acme.com/' as "ns1"),
  '/ns1:*//myId'
  passing xmltype(d.data)
  columns myId number path '.',
    parent varchar2(20) path './../local-name()'
) x
order by d.pid desc;

       PID NAME       MYID PARENT             
---------- ---- ---------- --------------------
        42 Test       1234 parent              

如果您需要更复杂的东西，则可以从节点级别中获取所需的任何内容，如该答案所示 ；但是根据您的说法，这太过分了.

If you need something more complicated you can pull whatever you need from the node levels, as shown in this answer; but from what you've said that's overkill here.

这篇关于如何获取Oracle XPath表达式中父元素的名称?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！