From d04838af2ca4b9e30ad1b8b43c5498f112b66263 Mon Sep 17 00:00:00 2001
From: skyanth <skyanth@gmail.com>
Date: Thu, 20 Feb 2020 17:16:08 +0100
Subject: [PATCH] better cleaning

---
 xml/xslt/gitlab_export_cleanup.xsl  | 43 ++++++++++++++---------------
 xml/xslt/gitlab_export_splitter.xsl | 27 ++++++++++++++++--
 2 files changed, 46 insertions(+), 24 deletions(-)

diff --git a/xml/xslt/gitlab_export_cleanup.xsl b/xml/xslt/gitlab_export_cleanup.xsl
index 5208d80..d26dcee 100644
--- a/xml/xslt/gitlab_export_cleanup.xsl
+++ b/xml/xslt/gitlab_export_cleanup.xsl
@@ -2,6 +2,7 @@
 <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">
 
+    <xsl:strip-space elements="*"/>
     <xsl:output indent="yes" method="xml"/>
 
     <xsl:template match="@* | node()">
@@ -10,13 +11,8 @@
         </xsl:copy>
     </xsl:template>
 
-    <!-- remove all spans; they are not used in pentext -->
-    <xsl:template match="span">
-        <xsl:apply-templates/>
-    </xsl:template>
-
-    <!-- remove all divs; they are not used in findings -->
-    <xsl:template match="div">
+    <!-- remove all of the following elements; they are not used in pentext -->
+    <xsl:template match="span | div | font">
         <xsl:apply-templates/>
     </xsl:template>
 
@@ -27,19 +23,19 @@
 
     <!-- remove selected attributes from selected elements -->
     <xsl:template match="pre/@class | a/@class | tr/@class | img/@alt"/>
-    
+
     <!-- change em to i -->
     <xsl:template match="em">
-            <i>
-                <xsl:apply-templates/>
-            </i>
+        <i>
+            <xsl:apply-templates/>
+        </i>
     </xsl:template>
-    
+
     <!-- change strong to b -->
     <xsl:template match="strong">
-            <b>
-                <xsl:apply-templates/>
-            </b>
+        <b>
+            <xsl:apply-templates/>
+        </b>
     </xsl:template>
 
     <!-- remove h*, make bold paragraph -->
@@ -53,9 +49,14 @@
 
     <!-- add .. to <img src="/uploads/[long code]/file.png"/> -->
     <xsl:template match="img/@src">
-        <xsl:choose><xsl:when test="starts-with(., '/uploads/')">
-            <xsl:attribute name="src" select="concat('..', .)"/>
-        </xsl:when><xsl:otherwise><xsl:copy></xsl:copy></xsl:otherwise></xsl:choose>
+        <xsl:choose>
+            <xsl:when test="starts-with(., '/uploads/')">
+                <xsl:attribute name="src" select="concat('..', .)"/>
+            </xsl:when>
+            <xsl:otherwise>
+                <xsl:copy/>
+            </xsl:otherwise>
+        </xsl:choose>
     </xsl:template>
 
     <!-- get rid of superfluous breaks before images or h3 tags -->
@@ -63,7 +64,7 @@
     <xsl:template
         match="br[following-sibling::img] | br[following-sibling::h3] | br[following-sibling::p]"> </xsl:template>
 
-    <!-- insert default img width to nudge pentesters :) -->
+    <!-- insert default img width to familiarize pentesters with the concept of image size :) -->
     <xsl:template match="img[not(@height) and not(@width)]">
         <xsl:copy>
             <xsl:attribute name="width">17</xsl:attribute>
@@ -77,8 +78,6 @@
         <xsl:apply-templates/>
     </xsl:template>
 
-<xsl:template match="p[child::img[not(preceding-sibling::*)][not(following-sibling::*)]]">
-        <xsl:apply-templates select="@* | node()"/>
-    </xsl:template>
+    
 
 </xsl:stylesheet>
diff --git a/xml/xslt/gitlab_export_splitter.xsl b/xml/xslt/gitlab_export_splitter.xsl
index 31c78b0..4ede2ca 100644
--- a/xml/xslt/gitlab_export_splitter.xsl
+++ b/xml/xslt/gitlab_export_splitter.xsl
@@ -2,7 +2,8 @@
 <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">
 
-    <xsl:output indent="yes" method="xml"/>
+<xsl:strip-space elements="*"/>
+    <xsl:output indent="yes" method="xml"></xsl:output>
 
 
 
@@ -13,8 +14,30 @@
     </xsl:template>
 
 
-    <xsl:template match="p[child::img[not(preceding-sibling::*)][not(following-sibling::*)]]">
+    <!--<xsl:template match="p[child::img[not(preceding-sibling::*)][not(following-sibling::*)]]">
         <xsl:copy-of select="img"/>
+    </xsl:template>-->
+    
+    <!-- find p with only img children & free the children -->
+    <xsl:template match="p[not(text()) and * and not(*[not(self::img)])]" priority="10">
+        <xsl:copy-of select="*"/>
+    </xsl:template>
+    
+    <!-- find p with intro text and single or double img and move img out of p -->
+    <xsl:template match="
+        p[node()[1][self::text()] and node()[2][self::img] and 
+        (not(node()[3]) 
+        or 
+        (node()[3][self::img] and not(node()[4])))
+        ]" priority="5">
+        <p><xsl:copy-of select="node()[1]"/></p>
+        <xsl:copy-of select="*"/>
     </xsl:template>
 
+<!-- find p where last element is img and move img out of p (picking up some stragglers this way) -->
+<xsl:template match="
+        p[node()[last()][self::img]]" priority="1">
+        <p><xsl:copy-of select="node()[position() &lt; last()]"/></p>
+        <xsl:copy-of select="node()[last()]"/>
+    </xsl:template>
 </xsl:stylesheet>