Match text node in element with sub elements

Question

I am having the same problem as described in How to get text from the node in xml file, that contains text and child node? but I need XSLT to match the text.

I have the following simplified XML:

<?xml version="1.0" encoding="UTF-8"?>
<order>2013-09-02T00:00:00COPY
<contentVersion>
    <versionIdentification>2.1.1</versionIdentification>
</contentVersion>
</order>

I have tried the following XPATHs:

<xsl:variable name="vtextPostM" select="text()[self::order]"/>
<xsl:variable name="vtextPostM" select="self::text()[1]"/>
<xsl:variable name="vtextPostM" select="text()[self::*]"/>

I am expecting:

2013-09-02T00:00:00COPY

what's the context? can you post your whole xslt path? what xslt processor are you using? a reproducible example would really help. If you are interested in the 2013... string, why did you include the contentVersion node? — Alex Brown, Jan 15 '14 at 06:48

Joel M. Lamsen · Answer 1 · 2014-01-15T06:40:54.147

2

Whoops, just edited my answer:

<xsl:variable name="vtextPostM" select="order/child::node()[position()=1][self::text()]"/>

edited Jan 15 '14 at 06:40

answered Jan 15 '14 at 05:44

Joel M. Lamsen

7,143
1
12
14

No, you can't. The `child::*` axis does not include text nodes. – Thomas W Jan 15 '14 at 06:18
but why did `` and `` produced the wanted text node if that is not selected? – Joel M. Lamsen Jan 15 '14 at 06:43
I'm sorry, of course `child::*` does include text nodes (unlike `*`). – Thomas W Jan 15 '14 at 07:09
Although it may not be apparent in the given example, there *is* a difference between "the first text node" and "the first node, provided it is text". – michael.hor257k Jan 15 '14 at 07:32
Thank you Joel for your solution - it works perfectly (+1). I will go with the one from Michael though because I do not need the check for position. Thank you, Peter – Peter Jan 15 '14 at 10:08

michael.hor257k · Accepted Answer · 2014-01-14T20:35:50.887

1

Where exactly are you when you define the variable? Without any context, the absolute path:

/order/text()[1]

or just (using your specific example):

/order/text()

should produce the result you want. You will want to wrap this in a normalize-space(), because it includes the trailing line feed before <contentVersion>.

edited Jan 14 '14 at 20:35

answered Jan 14 '14 at 16:24

michael.hor257k

113,275
6
33
51

Hello Michael, thank you. I will go with your solution. My mistake was that I did not consider the current node (order) properly in my XPATHs. Lesson learnt. Best regards, PEter – Peter Jan 15 '14 at 10:09

Match text node in element with sub elements

2 Answers2