Changeset ad35672

Timestamp:

15/04/2012 00:12:29 (20 months ago)

Author:

Eric van der Vlist <vdv@dyomedea.com>

Branches:

master

Children:

c4c4108

Parents:

ba51ddf

git-author:

Eric van der Vlist <vdv@dyomedea.com> (15/04/2012 00:12:29)

git-committer:

Eric van der Vlist <vdv@dyomedea.com> (15/04/2012 00:12:29)

Message:

Still work in progress, but the WARC archive now validates with warc-tools' warcvalid.py...

Location:

archiver/pipelines/actions

Files:

• package-archive.xpl (modified) (1 diff)
• warc-lib.xsl (modified) (4 diffs)

archiver/pipelines/actions/package-archive.xpl

@@ -212 +212 @@
             <document xsl:version="2.0" xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:type="xs:string" content-type="text/plain">
               <xsl:apply-templates select="$request" mode="warc"/>
-              <xsl:apply-templates select="$response" mode="warc"/>
+              <xsl:apply-templates select="$response" mode="warc">
+                <xsl:with-param name="document-length" as="xs:integer" select="string-length(translate(/archive/response/document, ' &#xa;&#xd;', '')) * 3 div 4" tunnel="yes"/>
+              </xsl:apply-templates>
             </document>
           </xsl:template>

archiver/pipelines/actions/warc-lib.xsl

@@ -31 +31 @@
 
     <xsl:template match="record" mode="warc">
+        <xsl:param name="document-length" as="xs:integer" select="0" tunnel="yes"/>
+        <xsl:call-template name="CRLF"/>
+        <xsl:call-template name="CRLF"/>
         <xsl:apply-templates select="header" mode="warc"/>
         <xsl:variable name="block">
@@ -39 +42 @@
                 <name>Content-Length</name>
                 <value>
-                    <xsl:value-of select="string-length($block)"/>
+                    <xsl:value-of select="string-length($block) + $document-length "/>
                 </value>
             </field>
@@ -46 +49 @@
         <xsl:call-template name="CRLF"/>
         <xsl:value-of select="$block"/>
-        <xsl:call-template name="CRLF"/>
-        <xsl:call-template name="CRLF"/>
     </xsl:template>
 
@@ -72 +73 @@
 
     <xsl:template match="response" mode="warc-http">
+        <!--<xsl:message>
+            <xsl:value-of select="string-length(document)"/>
+            <xsl:text> - </xsl:text>
+            <xsl:value-of select="string-length(translate(document, ' &#xa;&#xd;', ''))"/>
+        </xsl:message>-->
         <line>
             <!-- TODO: get the HTTP version and status-->