Added images treatement

wikipedia_to_hdoc/ant/wiki_to_hdoc.ant

@@ -36,6 +36,19 @@
         <xslt classpath="lib/saxon9he.jar" in="input/${inputFile}" out="${tmpFolderPath}/${filename}_prepared.xml" style="${xsltFolderPath}/prepare_wiki_to_hdoc.xsl">
             <factory name="net.sf.saxon.TransformerFactoryImpl"/>
         </xslt>
+
+
+        <!-- Prepare the ant script that will get our ressources(images )-->
+
+        <xslt classpath="lib/saxon9he.jar" 
+            in="${tmpFolderPath}/${filename}_prepared.xml" 
+            out="${tmpFolderPath}/${filename}_ressources.xml" style="${xsltFolderPath}/get_ressources_urls.xsl">
+            <factory name="net.sf.saxon.TransformerFactoryImpl"/>
+
+        </xslt>
+
+        <ant antfile="${tmpFolderPath}/${filename}_ressources.xml"/>  
+
     </target>
     
     <!-- Create the hdoc structure of the new file -->
@@ -65,7 +78,11 @@
         <copy todir="${tmpHdocResultFolderPath}/listing" >  
             <fileset dir="${hdocResultPath}/listing" includes="**"/>  
         </copy>
-        
+
+        <copy todir="${tmpHdocResultFolderPath}/ressources" >  
+            <fileset dir="${hdocResultPath}/ressources" includes="**"/>  
+        </copy>  
+                
         <!-- Copy of the html file, hdoc needs "container.xml" file, not a html. The html file can be useful to the user -->
         <copy file="${hdocResultPath}/${filename}.html" tofile="${tmpHdocResultFolderPath}/content.xml"/>
     </target>

wikipedia_to_hdoc/ant/wiki_to_hdoc_fetcher.ant

@@ -42,6 +42,19 @@
         <xslt classpath="lib/saxon9he.jar" in="${inputFile}" out="${tmpFolderPath}/${filename}_prepared.xml" style="${xsltFolderPath}/prepare_wiki_to_hdoc.xsl">
             <factory name="net.sf.saxon.TransformerFactoryImpl"/>
         </xslt>
+
+
+        <!-- Prepare the ant script that will get our ressources(images )-->
+
+        <xslt classpath="lib/saxon9he.jar" 
+            in="${tmpFolderPath}/${filename}_prepared.xml" 
+            out="${tmpFolderPath}/${filename}_ressources.xml" style="${xsltFolderPath}/get_ressources_urls.xsl">
+            <factory name="net.sf.saxon.TransformerFactoryImpl"/>
+
+        </xslt>
+
+        <ant antfile="${tmpFolderPath}/${filename}_ressources.xml"/>    
+
     </target>
     
     <!-- Create the hdoc structure of the new file -->
@@ -77,6 +90,11 @@
             <fileset dir="${hdocResultPath}/tables" includes="**"/>  
         </copy>
         
+
+
+        <copy todir="${tmpHdocResultFolderPath}/ressources" >  
+            <fileset dir="${hdocResultPath}/ressources" includes="**"/>  
+        </copy>        
         <!-- Copy of the html file, hdoc needs "container.xml" file, not a html. The html file can be useful to the user -->
         <copy file="${hdocResultPath}/${filename}.html" tofile="${tmpHdocResultFolderPath}/content.xml"/>
     </target>

wikipedia_to_hdoc/xslt/get_ressources_urls.xsl

+<?xml version="1.0" encoding="UTF-8"?>
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+    xmlns:xs="http://www.w3.org/2001/XMLSchema"
+    exclude-result-prefixes="xs"
+    version="2.0"
+    xmlns:saxon="http://saxon.sf.net/" 
+    extension-element-prefixes="saxon"
+    >
+    
+      <!-- This XSL create an ANT script that will download our ressources (images) using the GET TASK-->
+    
+    <xsl:output method="xml" indent="yes" encoding="UTF-8"/>
+    
+    <xsl:template match="/">
+        <project name="getRessourcesFiles" basedir="." default="getRessourcesFiles">
+         
+            <target name="getRessourcesFiles">
+                <xsl:apply-templates select="//div[@class='thumb tleft' or @class='thumb tright']//img"/>
+            </target>
+        </project>
+    </xsl:template>
+    
+    
+    
+    <xsl:template match="img" priority="5">
+        <xsl:param name="dest-rep">${hdocResultPath}/ressources</xsl:param>
+        <xsl:param name="dest-file">${hdocResultPath}/ressources/<xsl:value-of select="tokenize(@src, '/')[last()]"/></xsl:param>
+        <mkdir dir="{$dest-rep}"/>
+        
+        <get src="{@src}" dest="{$dest-file}" />
+    </xsl:template>
+    
+    
+    
+</xsl:stylesheet>
\ No newline at end of file

wikipedia_to_hdoc/xslt/prepare_wiki_to_hdoc.xsl

@@ -10,6 +10,31 @@
         </xsl:copy>
     </xsl:template>
     
+    <xsl:template match ="/html/body/div/div/div/div[@class='thumb tleft' or @class='thumb tright']" priority="1">
+
+        <div class="{@class}"> 
+            <xsl:apply-templates select=".//img" />
+            <xsl:apply-templates select=".//div[@class='thumbcaption']" />
+            
+        </div>
+    </xsl:template>
+
+    <xsl:template match="img">
+            <img src="http:{@src}" alt="{@alt}"/>
+    </xsl:template>
+    <xsl:template match="div[@class='thumbcaption']">
+            <p class="thumbcaption"> 
+                <xsl:apply-templates select="./text() | .//a/text()" mode="thumbcaption"/>
+            </p>
+    </xsl:template>
+
+    <xsl:template match="text()" mode="thumbcaption">
+          <xsl:value-of select="."/>
+    </xsl:template>
+
+
+
+
     <!-- Scripts are not useful to us -->
     <xsl:template match="script"/>
 

wikipedia_to_hdoc/xslt/wiki_to_hdoc.xsl

@@ -77,6 +77,13 @@
             <!-- Storing current section to know when apply template has to be called in the next parts of the template -->
             <xsl:variable name="currentSectionTitle" select="." />
 
+
+            <!-- Apply this template for all elements between the actual title h{2,3,4,5} and the next one  -->
+            <xsl:variable name="nextTitle" select="following-sibling::*[starts-with(name(), 'h')][1]"/>
+
+            <xsl:apply-templates select="following-sibling::div[contains(@class,'thumb tleft') or contains(@class,'thumb tright')] intersect $nextTitle/preceding-sibling::*"/>
+
+
             <!-- If there is text right below the section name, copy it -->
             <xsl:if test="not(following-sibling::*[1] intersect following-sibling::h3) and not(following-sibling::*[1] intersect following-sibling::h4) and not(following-sibling::*[1] intersect following-sibling::h5) and not(following-sibling::*[1] intersect following-sibling::h6)">
                 <div>
@@ -320,6 +327,24 @@
         <sup><xsl:value-of select="."/></sup>
     </xsl:template>
     
+
+
+    <!-- Images -->
+    <xsl:template match="div[contains(@class,'thumb')]">
+        <div> 
+            <xsl:apply-templates select=".//img"/>
+            <xsl:apply-templates select=".//p[@class='thumbcaption']"/>
+        </div>
+    </xsl:template>  
+
+    <xsl:template match="img">
+     <img src="./ressources/{tokenize(@src, '/')[last()]}" alt="{@alt}"/>
+    </xsl:template>
+
+    <xsl:template match="p[@class='thumbcaption']">
+        <p> <xsl:value-of select="text()"/> </p>
+    </xsl:template>
+
     
     <!-- ===== Ignored content ===== -->