开发者

Which is the better approach to remove the redundant white space in XML [strip-space or indent="no"]?

开发者 https://www.devze.com 2022-12-21 14:07 出处:网络
I want to print my output xml in a single line[when viewed in notepad or other simple text-editor], so as to remove the redundant white-space in my xml file. So which is the better method to follow fo

I want to print my output xml in a single line[when viewed in notepad or other simple text-editor], so as to remove the redundant white-space in my xml file. So which is the better method to follow for that ??

I think there are two options,

1) To use

  <xsl:output method="xml" indent="no"/>

2) or to use

  <xsl:strip-space elements="*"/>

Which is more efficient, and why?

some people suggest me to use indent="no",

I believed that strip-space is best suited, but not sure because of suggestions given by others.

To be more elaborated let me take an example:

Input XML:

<root>
 <node>
   <chi开发者_StackOverflow社区ld1/>
   <child2/>
 </node>
</root>

and the output required is:

<root><node><child1/><child2/></node></root>


In order to eliminate anything that looks like "indentation" it may be necessary (that means there are cases when you need) to use both <xsl:strip-space> and ``indent="no"`.

Take the simplest example: you have the identity transformation. Without any of the two methods specified, the transformation will reproduce the white-space-only text nodes from the source XML document. That is, if the source XML document is indented, the transformation will produce indented result, too.

Now, add to this transformation <xsl:output indent="no" />. This instructs the XSLT processor not to perform "pretty-printing" of its own. However, the whitespace-only nodes from the source XML document are still copied to the output and the result document looks still indented (because the source document is indented).

Now, as a last step, add <xsl:strip-space elements="*"/>. You have specified both methods of preventing white-space-only nodes in the output. What happens? No white-space-only nodes are processed at all by the XSLT processor, and it does not indent the output -- you get your desired one-line dense output.

Finally, make a regression, change the <xsl:output indent="no" /> to <xsl:output indent="yes" />. The <xsl:strip-space elements="*"/> is still there, so no whitespace-only nodes are reproduced in the output. But the XSLT processor obeys the <xsl:output indent="yes" /> directive and adds whitespace-only text nodes of its own.

So, from the four possible combinations, only specifying both <xsl:strip-space elements="*"/> and <xsl:output indent="no" /> guarantees that no indentation will be caused either from whitespace-only nodes from the source XML document or from the XSLT processors initiative.

Even this last case, of course, doesn't completely guarantee that the output won't be indented -- if the XSLT programmer intentionally puts there indentation code such as

<xsl:text>

</xsl:text>

the output will contain this indentation.


Perfomance differences are best measured. XSLT processor implementations differ, and you should make the test for yourself (though I suspect that worrying over the performance of the one or the other might fall into the "premature optimization" category in this case).

<xsl:output indent="no" /> might not have the effect you want unless accompanied by

<xsl:template match="text(normalize-space()='')" />

because if whitespace nodes (the ones between your tags) are not removed, then they will appear in the output at some point, regardless of the "output" setting.

0

精彩评论

暂无评论...
验证码 换一张
取 消

关注公众号