Re: unparsed-text and analyze-string

Thank you, Ken, for your quick response to my query. Your solution
works like a charm. Unfortunately, I will not be able to use it on
this particular project. My "unparsed" text will contain some element
structures such as <u>underlined</u>, <b>bold,</b> and <i>italicized</
i> words and phrases.

It appears that both the function replace() and the element
xsl:analyze-string strip such elements from the selected text. Is
there any way to get around this?

Note the <i> element in choice B of the second question in this this
import xml:

<document>
<part>
    <par class="stem_mc">&#9;19&#9;Choose the word that is a
homophone for <inline style="font-style: italic;">Ill.</inline> </par>
    <par class="choice-a">&#9;A&#9;ill</par>
    <par class="choice-b">&#9;B&#9;aisle</par>
    <par class="choice-c">&#9;C&#9;I will</par>
    <par class="choice-d">&#9;D&#9;idle</par>
    <par class="answer">B</par>
    <par class="g_code">G5U1S6</par>

    <par class="stem_mc">&#9;20&#9;Read the sentence. Then choose
the word that fits in the blank.</par>
    <par class="display">The man needed some ____ to get into the
building.</par>
    <par class="choice-a">&#9;A&#9;assist</par>
    <par class="choice-b">&#9;B&#9;<i>unessisted</i></par>
    <par class="choice-c">&#9;C&#9;assisting</par>
    <par class="choice-d">&#9;D&#9;assistance</par>
    <par class="answer">D</par>
    <par class="g_code">G5U1S2</par>
  </part>
</document>

This stylesheet strips the <i> element while it replaces the
&#9;A&#9; structures from the choices. (There is probably a more
elegant way to make these changes. All suggestions welcome!)

<?xml version="1.0" encoding="ISO-8859-1"?>
<xsl:stylesheet version="2.0"
	xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
	<!--<xsl:strip-space elements="*"/>-->

	<!--identity template-->
<xsl:template match="@*|node()">
	<xsl:copy>
		<xsl:apply-templates select="@*|node()"/>
	</xsl:copy>
</xsl:template>

<xsl:template match='par[@class="choice-a"]'>
<xsl:element name="choice-a">
          <xsl:copy-of select="replace(.,'\tA\t','')"/>
          </xsl:element>
</xsl:template>

<xsl:template match='par[@class="choice-b"]'>
<xsl:element name="choice-b">
          <xsl:copy-of select="replace(.,'\tB\t','')"/>
          </xsl:element>
</xsl:template>

<xsl:template match='par[@class="choice-c"]'>
<xsl:element name="choice-c">
          <xsl:copy-of select="replace(.,'\tC\t','')"/>
          </xsl:element>
</xsl:template>

<xsl:template match='par[@class="choice-d"]'>
<xsl:element name="choice-d">
          <xsl:copy-of select="replace(.,'\tD\t','')"/>
          </xsl:element>
</xsl:template>
</xsl:stylesheet>

 <question type="mc">        19      Choose the word that is a
homophone for  Ill.  </question>
    <choice-a>ill</choice-a>
    <choice-b>aisle</choice-b>
    <choice-c>I will</choice-c>
    <choice-d>idle</choice-d>
    <answer>B</answer>
    <par class="g_code">G5U1S6</par>

    <question type="mc">        20      Read the sentence. Then
choose the word that fits in the blank.</question>
    <par class="display">The man needed some ____ to get into the
building.</par>
    <choice-a>assist</choice-a>
    <choice-b>unessisted</choice-b>
    <choice-c>assisting</choice-c>
    <choice-d>assistance</choice-d>
    <answer>D</answer>
    <par class="g_code">G5U1S2</par>
  </part>
</document>

In case you are wondering, I am trying to develop a workflow that
starts with Microsoft Word documents. My current experiment is to
export styled word documents to rtf which are then converted using
UpCast. That gives me the current input document. As I noted before,
I am trying to avoid marking up this copy manually. I am not sure I
will succeed, as there are more hurdles yet to jump.