基于可选标签的 XML 拆分

Question

I am looking for a simple split of an xml file just based on tags; say 3 tags (optional) always repeat and need split as depicted below:

Input

<?xml version="1.0" encoding="UTF-8"?>
<Test>
    <tag1>A</tag1>
    <tag2>B</tag2>
    <tag3>C</tag3>
    <tag1>1</tag1>
    <tag2>2</tag2>
    <tag3>3</tag3>
    <tag1>apple</tag1>
    <tag2>orange</tag2>
    <tag3>mango</tag3>
</Test>

Expected Output

<Root>
    <Test>
        <tag1>A</tag1>
        <tag2>B</tag2>
        <tag3>C</tag3>
    </Test>
    <Test>
        <tag1>1</tag1>
        <tag2>2</tag2>
        <tag3>3</tag3>
    </Test>
    <Test>
        <tag1>apple</tag1>
        <tag2>orange</tag2>
        <tag3>mango</tag3>
    </Test>
</Root>

The challenge here is that all 3 tags are optional and can or cannot appear in a block. If there weren't optional - my question was already answered here - Split based on just tags

Any help is appreciated

Thanks

Answer 1

If I understand the requirement correctly, you want to do:

XSLT 1.0

<xsl:stylesheet version="1.0" 
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method="xml" version="1.0" encoding="UTF-8" indent="yes"/>

<xsl:template match="/Test">
    <Root>
        <xsl:apply-templates/>
    </Root>
</xsl:template>

<xsl:template match="tag1">
    <Test>
        <xsl:copy-of select=". | following-sibling::*[1][self::tag2 or self::tag3]"/>
        <xsl:if test="following-sibling::*[1][self::tag2]">
            <xsl:copy-of select="following-sibling::*[2][self::tag3] "/>
        </xsl:if>
    </Test>
</xsl:template>

<xsl:template match="tag2[not(preceding-sibling::*[1][self::tag1])]">
    <Test>
        <xsl:copy-of select=". | following-sibling::*[1][self::tag3]"/>
    </Test>
</xsl:template>

<xsl:template match="tag3[not(preceding-sibling::*[1][self::tag2 or self::tag1])]">
    <Test>
        <xsl:copy-of select="."/>
    </Test>
</xsl:template>

<xsl:template match="text()"/>

</xsl:stylesheet>

This will start a new group for each of these:

• a tag1 element;
• a tag2 element that does not immediately follow a tag1 element;
• a tag3 element that does not immediately follow a tag2 or a tag1 element.

A group ends with tag3 , or with the last element before the sequence breaks, whichever comes first.

Applied to the following test input:

XML

<Test>
    <tag2>B</tag2>
    <tag3>C</tag3>
    <tag1>1</tag1>
    <tag3>3</tag3>
    <tag1>apple</tag1>
    <tag2>orange</tag2>
    <tag1>A</tag1>
    <tag2>B</tag2>
    <tag1>1</tag1>
    <tag3>3</tag3>
    <tag3>mango</tag3>
    <tag1>A</tag1>
    <tag2>B</tag2>
    <tag3>C</tag3>
</Test>

the result will be:

<?xml version="1.0" encoding="UTF-8"?>
<Root>
   <Test>
      <tag2>B</tag2>
      <tag3>C</tag3>
   </Test>
   <Test>
      <tag1>1</tag1>
      <tag3>3</tag3>
   </Test>
   <Test>
      <tag1>apple</tag1>
      <tag2>orange</tag2>
   </Test>
   <Test>
      <tag1>A</tag1>
      <tag2>B</tag2>
   </Test>
   <Test>
      <tag1>1</tag1>
      <tag3>3</tag3>
   </Test>
   <Test>
      <tag3>mango</tag3>
   </Test>
   <Test>
      <tag1>A</tag1>
      <tag2>B</tag2>
      <tag3>C</tag3>
   </Test>
</Root>

Answer 2

Here's an alternative solution that can work with groups of any size, provided that at least one element of the group will always be present.

In this example there are up to 5 elements in each group, and element tag3 will be present in each group:

XML

<Test>
    <tag1>A</tag1>
    <tag2>B</tag2>
    <tag3>C</tag3>
    <tag4>D</tag4>
    <tag5>E</tag5>
    <tag2>2</tag2>
    <tag3>3</tag3>
    <tag5>5</tag5>
    <tag3>C</tag3>
    <tag3>3</tag3>
    <tag4>4</tag4>
    <tag5>5</tag5>
    <tag2>B</tag2>
    <tag3>C</tag3>
    <tag4>D</tag4>
    <tag1>1</tag1>
    <tag3>3</tag3>
    <tag5>5</tag5>
</Test>

XSLT 1.0

<xsl:stylesheet version="1.0" 
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method="xml" version="1.0" encoding="UTF-8" indent="yes"/>

<xsl:key name="elems-before" match="tag1 | tag2" use="generate-id(following-sibling::tag3[1])" />
<xsl:key name="elems-after"  match="tag4 | tag5" use="generate-id(preceding-sibling::tag3[1])" />

<xsl:template match="/Test">
    <Root>
        <xsl:for-each select="tag3">
            <Test>
                <xsl:copy-of select="key('elems-before', generate-id())"/>
                <xsl:copy-of select="."/>
                <xsl:copy-of select="key('elems-after', generate-id())"/>
            </Test>
        </xsl:for-each>             
    </Root>
</xsl:template>

</xsl:stylesheet>

Result

<?xml version="1.0" encoding="UTF-8"?>
<Root>
   <Test>
      <tag1>A</tag1>
      <tag2>B</tag2>
      <tag3>C</tag3>
      <tag4>D</tag4>
      <tag5>E</tag5>
   </Test>
   <Test>
      <tag2>2</tag2>
      <tag3>3</tag3>
      <tag5>5</tag5>
   </Test>
   <Test>
      <tag3>C</tag3>
   </Test>
   <Test>
      <tag3>3</tag3>
      <tag4>4</tag4>
      <tag5>5</tag5>
   </Test>
   <Test>
      <tag2>B</tag2>
      <tag3>C</tag3>
      <tag4>D</tag4>
   </Test>
   <Test>
      <tag1>1</tag1>
      <tag3>3</tag3>
      <tag5>5</tag5>
   </Test>
</Root>