XML to text (2)

raoulvb · June 5th, 2003, 02:53 AM

Hi all,

I have the following xml-document.

Code:

<?xml version="1.0" ?>
<datarows>
  <datarow class="a">Line 0</datarow>
  <datarow class="b">Subline 0</datarow>
  <datarow class="b">Subline 1</datarow>
  <datarow class="b">Subline 2</datarow>
  <datarow class="a">Line 1</datarow>
  <datarow class="b">Subline 10</datarow>
  <datarow class="b">Subline 11</datarow>
  <datarow class="b">Subline 12</datarow>
</datarows>

And i want to end up with a textfile like:

"Line0";"Subline 0";"Subline 1";"Subline 2"
"Line1";"Subline 10";"Subline 11";"Subline 12"

I have made the following stylesheet.

Code:

<?xml version="1.0" encoding="iso-8859-1" ?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
  <xsl:output method="text" encoding="utf-8" />
  <xsl:strip-space elements="*"/>

  <xsl:template match="datarows[datarow[@class='b']]">
    <xsl:apply-templates select="preceding-sibling::datarows[datarow[@class='a']]" />
    <xsl:apply-templates select="datarow" />
  </xsl:template>

  <xsl:template match="datarow">
    <xsl:text>"</xsl:text>
    <xsl:value-of select="normalize-space(.)" />
    <xsl:choose>
      <xsl:when test="position() = last()">
        <xsl:text>"#13;</xsl:text>
      </xsl:when>
      <xsl:otherwise>
        <xsl:text>";</xsl:text>
      </xsl:otherwise>
    </xsl:choose>
  </xsl:template>

</xsl:stylesheet>

First, I can't figure out how to get two lines in the outputdocument.
Second, i want to ask you if there is a way 'to speed up the stylesheet'.
When i have a document that is 10Mb, the time to transform the document can be very long.
The main issue here is, that time the parser encounters a datarow-element with class 'b' it has to search back in the tree to find the previous datarow-element 'a'.
If you have a very big document, you can imagine what happens.
Is it a good idea to work with a xsl:variable or maybe someone has a different approach?

Thanks in advance, Raoul.

rushman · June 5th, 2003, 09:41 AM

Chack this out, it does the trick. Regarding large documents, you should probably turn to SAX to parse your documents and generate events to handle the transformation instead of XSLT.

<?xml version="1.0"?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
    <xsl:output method="text"/>
    <xsl:template match="/">
        <xsl:apply-templates/>
    </xsl:template>
    <xsl:template match="datarow">
        <xsl:variable name="class" select="@class"/>
        <xsl:if test="$class='a'">
            <xsl:text>"</xsl:text>
            <xsl:value-of select="."/>
            <xsl:text>"</xsl:text>
            <xsl:if test="boolean(following-sibling::node())">
                <xsl:text>;</xsl:text>
            </xsl:if>
            <xsl:call-template name="subclass"/>
            <xsl:text>#x0d;</xsl:text>
        </xsl:if>
    </xsl:template>
    <xsl:template name="subclass">
        <xsl:variable name="parent" select="."/>
        <xsl:for-each select="following-sibling::*[@class='b' and preceding-sibling::*[@class='a'][1]=$parent]">
            <xsl:text>"</xsl:text>
            <xsl:value-of select="."/>
            <xsl:text>"</xsl:text>
            <xsl:if test="following-sibling::datarow[1]/@class != 'a'">
                <xsl:text>;</xsl:text>
            </xsl:if>
        </xsl:for-each>
    </xsl:template>
</xsl:stylesheet>