Syd: bug fixes

2008-08-08T15:34:33Z

bug fixes

Syd: fix some comments

2008-08-03T12:35:11Z

fix some comments

Syd at 14:40, 28 July 2008

2008-07-28T14:40:08Z

Syd at 01:44, 27 July 2008

2008-07-27T01:44:35Z

Syd: New page: A stylesheet to read in TEI (or other XML files) and remove line-end soft hyphens. See header comment for details.

@@ Line 10: / Line 10: @@
 <!-- (or after it, if it has no internal whitespace). -->
 <!-- Written 2008-07-24/26 by Syd Bauman -->
 <!-- Copyleft 2008 Syd Bauman and the Brown University Women Writers Project -->
 <!--  -->
@@ Line 30: / Line 36: @@
 <!-- in this stylesheet test for shy only at end-of-line (except for whitespace). -->
 <xsl:stylesheet version="1.0"
-   xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+   xmlns:xsl="http://www.w3.org/1999/XSL/Transform" >
    <!-- housekeeping -->
@@ Line 49: / Line 54: @@
        <xsl:apply-templates select="@*|node()"/>
      </xsl:copy>
    </xsl:template>
    <!-- Match (or handle when called explicitly) any text node that contains -->
    <!-- a soft hyphen. -->
-   <xsl:template match="text()[contains(.,'&#x00AD;')]" name="hasShy">
+   <xsl:template match="text()[contains(.,'&amp;#xAD;')]" name="hasShy">
      <!-- Should we be stripping off the first word of this text node? -->
      <xsl:param name="strip1st">false</xsl:param> <!-- not unless asked -->
@@ Line 75: / Line 108: @@
        </xsl:choose>
      </xsl:variable>
-     <!-- get the string content of the next text node (put it in variable $nextText) -->
+     <!-- Get the string content of the next text node (other than text nodes that -->
      <xsl:variable name="nextText">
-       <xsl:value-of select="normalize-space(following-sibling::text()[1])"/>
+       <xsl:call-template name="strip-leading-space">
      </xsl:variable>
      <!-- some rudimentary error-checking -->
      <xsl:choose>
-       <xsl:when test="contains(substring-after($me,'&#xAD;'),'&#xAD;')">
+       <xsl:when test="contains(substring-after($me,'&amp;#xAD;'),'&amp;#xAD;')">
-         <xsl:message>Oh dear. More than 1 &amp;shy; in this text sequence, I'm not going to get this one right.</xsl:message>
+         <xsl:message>Oh dear. More than 1 &amp;amp;shy; in this text sequence, I'm not going to get this one right.</xsl:message>
        </xsl:when>
-       <xsl:when test="contains(substring-after($me,'&#xAD;'),'&#xAD;')">
+       <xsl:when test="string-length( substring-after( $me,'&amp;#xAD;') ) > 0">
-         <xsl:message>Oh dear. This &amp;shy; is not the last (non-whitespace) character of the text node. I'm probably going to mess this up.</xsl:message>
+         <xsl:message>Oh dear. This &amp;amp;shy; is not the last (non-whitespace) character of the text node. I'm probably going to mess this up.</xsl:message>
        </xsl:when>
      </xsl:choose>
      <!-- output myself up to, but not including, shy -->
-     <xsl:value-of select="substring-before($me,'&#xAD;')"/>
+     <xsl:value-of select="substring-before($me,'&amp;#xAD;')"/>
      <!-- first token of next text node -->
      <xsl:choose>
@@ Line 103: / Line 140: @@
      </xsl:choose>
      <!-- a newline to separate this word from the rest -->
-     <xsl:text>&#x0A;</xsl:text>
+     <xsl:text>&amp;#x0A;</xsl:text>
    </xsl:template>
-   <!-- Now match text nodes for which the preceding text node had a shy -->
+   <!-- Now match text nodes for which the preceding text node had a shy. -->
-   <!-- we want this template to be higher priority than 'hasShy', so that -->
+   <!-- We want this template to be higher priority than 'hasShy', so that -->
    <!-- when we hit a line that meets both criteria (has a shy, previous -->
    <!-- text node had a shy), we come here first -->
-   <xsl:template match="text()[preceding-sibling::text()[1][contains(.,'&#xAD;')]]" priority="1">
+   <xsl:template match="text()[preceding-sibling::text()[not(normalize-space(.)='')][1][contains(.,'&amp;#xAD;')]]" priority="1">
      <!-- preceding text node ended in shy, so it was caught by the hasShy template, -->
      <!-- which grabbed and printed out the first token of this text node. So our -->
@@ Line 116: / Line 153: @@
      <!-- to process it accordingly, too. -->
      <xsl:choose>
-       <xsl:when test="contains(.,'&#x00AD;')">
+       <xsl:when test="contains(.,'&amp;#xAD;')">
          <xsl:call-template name="hasShy">
            <xsl:with-param name="strip1st">true</xsl:with-param>
@@ Line 122: / Line 159: @@
        </xsl:when>
        <xsl:otherwise>
-         <xsl:value-of select="substring-after(normalize-space(.),' ')"/>
+        <!-- create a copy of myself (remember, I'm a text node) without leading -->
        </xsl:otherwise>
      </xsl:choose>

@@ Line 12: / Line 12: @@
 <!-- Copyleft 2008 Syd Bauman and the Brown University Women Writers Project -->
 <!--  -->
-<!-- Known issues (some would consider them bugs, other features): -->
+<!-- Known issues (some would consider them bugs, others would say features): -->
-<!-- * If there are more than one shy in a given text node, we end up -->
+<!-- * If there is more than one shy in a given text node, we end up -->
 <!--   doing the wrong thing: deleting all the text after the *first* -->
 <!--   shy, not the last. However, in this case we issue a warning -->
@@ Line 26: / Line 26: @@
 <!--   removed. It should probably be removed, but IMHO a warning would be a -->
 <!--   good idea. -->
-<!-- In truth, the right thing to do might be to validate that shys are palced -->
+<!-- In truth, the right thing to do might be to validate that shys are plqced -->
 <!-- appropiately with some other software (like a Schematron schema), and then -->
 <!-- in this stylesheet test for shy only at end-of-line (except for whitespace). -->
@@ Line 114: / Line 114: @@
      <!-- which grabbed and printed out the first token of this text node. So our -->
      <!-- goal is to print out the rest. But if this node also ends in a shy, we have -->
-     <!-- process it accordingly, too. -->
+     <!-- to process it accordingly, too. -->
      <xsl:choose>
        <xsl:when test="contains(.,'&#x00AD;')">
@@ Line 129: / Line 129: @@
 </xsl:stylesheet>
 </nowiki></pre>
 I make no claim that this is even a good, let alone the best, way to do this. It did work on my test files, though.
 [[Category:XSLT]][[Category:XSLT:1.0]]

← Older revision		Revision as of 14:40, 28 July 2008
Line 130:		Line 130:
	</nowiki></pre>		</nowiki></pre>
	I make no claim that this is even a good, let alone the best, way to do this. It did work on my test files, though.		I make no claim that this is even a good, let alone the best, way to do this. It did work on my test files, though.
		+
		+	[[Category:XSLT]][[Category:XSLT:1.0]]

@@ Line 3: / Line 3: @@
 <pre><nowiki>
 <?xml version="1.0" encoding="UTF-8"?>
-<!-- shy-remover.xslt -->
+<!-- unshy.xslt -->
 <!-- Reads in an XML file, writes out the same file with end-of-line soft hyphens -->
 <!-- removed. Any whitespace between the soft hyphen and the next text node is -->

Unshy.xslt - Revision history

Syd: bug fixes

Syd: fix some comments

Syd at 14:40, 28 July 2008

Syd at 01:44, 27 July 2008

Syd: New page: A stylesheet to read in TEI (or other XML files) and remove line-end soft hyphens. See header comment for details.