From c79bd8e49c8035ec657bd2b6f98507e0f9164ca7 Mon Sep 17 00:00:00 2001
From: Eric van der Vlist <vdv@dyomedea.com>
Date: Sat, 28 Apr 2012 09:42:21 +0200
Subject: [PATCH] Forcing HTML content type for XHTML documents

---
 archiver/pipelines/actions/mediatypes/warc-html.xpl | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/archiver/pipelines/actions/mediatypes/warc-html.xpl b/archiver/pipelines/actions/mediatypes/warc-html.xpl
index 7d30e57..f7411d1 100644
--- a/archiver/pipelines/actions/mediatypes/warc-html.xpl
+++ b/archiver/pipelines/actions/mediatypes/warc-html.xpl
@@ -11,6 +11,8 @@
     <p:input name="config">
       <config>
         <scope>session</scope>
+        <encoding>utf-8</encoding>
+        <force-encoding>true</force-encoding>
       </config>
     </p:input>
     <p:input name="data" href="#record#xpointer(/record/content/document)"/>
@@ -24,6 +26,10 @@
         <url>
           <xsl:value-of select="/*"/>
         </url>
+        <encoding>utf-8</encoding>
+        <force-encoding>true</force-encoding>
+        <content-type>text/html</content-type>
+        <force-content-type>true</force-content-type>
         <mode>html</mode>
       </config>
     </p:input>