Re: [xsl] Transforming XML Blockquotes - Mixed Content - XSLT 1.0 Solution

[JBryant@xxxxxxxxx]

Hi, Edward,

In a nutshell, p[not(preceding-sibling::*[1]/self::p)] says, "the p whose 
first preceding sibling is not a p". I used it to find the start of each 
group of paragraphs that should be grouped together.

You can get the num value into the resulting p blocks by changing this 
section:

    <p>
      <xsl:for-each select="../p[@group=$group]">
        <xsl:apply-templates/>
      </xsl:for-each>
    </p>

to this:

    <p num="{@num}">
      <xsl:for-each select="../p[@group=$group]">
        <xsl:apply-templates/>
      </xsl:for-each>
    </p>

I apologize for the oversight. In tangling with the grouping problem I 
forgot that you needed the num attribute to come through.

In a larger stylesheet meant to solve real problems (rather than the tiny 
snippet we have here), you could insert the method of handling blockquotes 
into a choose statement:

<xsl:template match="p">
  <xsl:choose>
    <xsl:when test="blockquote">
      <!-- Process the paragraph to handle the blockquote -->
    </xsl:when>
    <xsl:otherwise>
      <!-- Process the paragraph the usual, more simple, way -->
    </xsl:otherwise>
  </xsl:choose>
</xsl:template>

However, You'll still be checking each paragraph for the presence of a 
blockquote. However, any tool must do the same, yes? If any paragraph can 
contain a blockquote, doesn't the processing tool have to check every 
paragraph for the presence of a blockquote? So, some problems require this 
kind of processing. Of course, you could change the XML source to not have 
the root problem, but I assumed that was out of bounds.

Jay Bryant
Bryant Communication Services
(presently consulting at Synergistic Solution Technologies)





"Edward Bryant" <bryant_edward@xxxxxxxxxxx> 
04/27/2005 02:56 PM
Please respond to
xsl-list@xxxxxxxxxxxxxxxxxxxxxx


To
xsl-list@xxxxxxxxxxxxxxxxxxxxxx
cc

Subject
Re: [xsl] Transforming XML Blockquotes - Mixed Content - XSLT 1.0 Solution






Hello again,

Reply to David Carlisle and Wendell:

To the comments that I might try to use divs instead of the standard XHTML 

blockquote tag (because XHTML allows nested divs), I tried it and it 
doesn't 
work. If I remember correctly nested divs technically work but in IE6 a 
paragraph's initial indent is re-applied to the text following a 
blockquote 
(so, the continued part of the paragraph falsely appears to be a new 
paragraph). Hence, while some style features, such as line-height inherit 
correctly to the post blockquote text, it also incorrectly inherits the 
paragraph's initial indent.

Reply to David Carlisle and Michael Kay:

Thanks for David's explanation of why not to use DOE, as well as the 
tree-based reasoning behind XSLT (I understand why this kind of thing is a 

problem better). On the responses to my comment that Jay's solution was 
"overly complex", I did not mean that the offered solution was complex in 
that there were too many lines of code. What I meant was that the solution 

ends up handling or touching on a lot of data that isn't even involved in 
the blockquote problem. It is complex in that it results in a lot of extra 

processing of data, as in the example of a 20 page document with only one 
blockquote in one paragraph or the processing of documents without any 
blockquotes, etc.

Reply to Jay Bryant:

Thanks for your patience and your posting of a 1.0 compliant version of 
your 
solution. I did, however, have a few questions:

1. I tried to decipher this XPATH expression but I can't seem to 
understand 
it, can someone help explain what this is doing

         p[not(preceding-sibling::*[1]/self::p)]

2. In the below stylesheet (the second of the two you offered), how do you 

differentiate between the p representing the initial paragraph and the 
continued paragraph p tags. I need to figure out a way to output the 
original paragraph's "num" attribute next to the start of each real 
paragraph (but not repeat it for the new p tags surrounding the continued 
paragraph text).

<?xml version="1.0" encoding="ISO-8859-1"?>
<xsl:stylesheet version="1.0" 
xmlns:xsl="http://www.w3.org/1999/XSL/Transform";>

  <xsl:template match="x">
  <html>
  <head>
  <title>Paragraph Chunking Test</title>
  <link rel="stylesheet" type="text/css" href="style.css" />
  </head>
  <body>
    <xsl:for-each select="p[not(preceding-sibling::*[1]/self::p)]">
    <xsl:variable name="group" select="@group"/>
    <p>
      <xsl:for-each select="../p[@group=$group]">
        <xsl:apply-templates/>
      </xsl:for-each>
    </p>
    <xsl:apply-templates select="following-sibling::blockquote"/>
    </xsl:for-each>
  </body>
  </html>
  </xsl:template>

  <xsl:template match="p"/>

  <xsl:template match="blockquote">
    <xsl:copy-of select="."/>
  </xsl:template>

  <xsl:template match="span">
    <xsl:copy-of select="."/>
  </xsl:template>

</xsl:stylesheet>