XSL to filter duplicates [duplicate] - xslt

Question

This question already has answers here:
XSLT Removing duplicates
(4 answers)
Closed 9 years ago.
I am trying generate a PDF using XSL. But I dont want duplicates(it should not be displayed back to back) , only for value=Started.
Below is the XSL snippet where I display the query result (string_1 column).
<xsl:for-each select="root/query1/row">
<fo:table-row height="0.9cm">
<xsl:if test="event_name = 'Started'">
<fo:table-cell border-style="solid" border-width="0.5pt"
number-columns-spanned="5">
<fo:block font-family="Courier" color="Blue" font-size="10pt" font-weight="normal" text-align="center">
<xsl:value-of select="string_1"/>
</fo:block>
</fo:table-cell>
</xsl:if>
</xsl:for-each>
Eg,
My query1 may give results like:
String_1
======
Started
In Progress
Complete
Started
Started
In Progress
In Progress
Complete
My PDF should be
=========
Started
In Progress
Complete
Started
In Progress
In Progress
Complete
Sorry if I didn't provide much information. I am new to XSL.

Answer 1

I don't know, how your input XML exactly looks like, but you should be able to adapt from this XML
<root>
<query1>
<row>
<string_1>Started</string_1>
</row>
<row>
<string_1>In Progress</string_1>
</row>
<row>
<string_1>In Progress</string_1>
</row>
<row>
<string_1>In Progress</string_1>
</row>
<row>
<string_1>Complete</string_1>
</row>
</query1>
</root>
and this XSLT:
<xsl:template match="/">
<xsl:for-each select="root/query1/row">
#> <xsl:if test="not(preceding-sibling::*[1]=string_1)">
<fo:table-row height="0.9cm">
<!-- <xsl:if test="event_name = 'Started'"> -->
<fo:table-cell border-style="solid" border-width="0.5pt" number-columns-spanned="5">
<fo:block font-family="Courier" color="Blue" font-size="10pt" font-weight="normal" text-align="center">
<xsl:value-of select="string_1" />
</fo:block>
</fo:table-cell>
<!-- </xsl:if> -->
</fo:table-row>
#> </xsl:if>
</xsl:for-each>
</xsl:template>
I have highlighted the important lines by #>. I have commented out one of the if-statements which seems to be there for doing different formatting on started nodes.
preceding-sibling::*[1] gives you the last processed node. It is compared to string_1 which is the current node. If it is different (not()), it should become part of the output.