Dyomedea
/
owark
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Trying to guess content types

This commit is contained in:
Eric van der Vlist 2012-04-28 23:12:20 +02:00
parent bc581fabf9
commit 16ef7979b0
1 changed files with 31 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

39
archiver/pipelines/actions/mediatypes/warc-html.xpl
View File

@ -5,28 +5,51 @@
<p:param name="index" type="input"/> <p:param name="index" type="input"/>
<p:param name="rewritten" type="output"/> <p:param name="rewritten" type="output"/>
<!-- Try to guess the encoding... -->
<p:processor name="oxf:xslt">
<p:input name="data" href="#record"/>
<p:input name="config">
<encoding xsl:version="2.0">
<xsl:choose>
<xsl:when test="contains(/record/content/headers/header[@name='Content-Type'], 'charset=')">
<xsl:value-of select="substring-before(concat(substring-after(/record/content/headers/header[@name='Content-Type'], 'charset='), ';'), ';')"/>
<xsl:message>
ENCODING :
<xsl:value-of select="substring-before(concat(substring-after(/record/content/headers/header[@name='Content-Type'], 'charset='), ';'), ';')"/>
</xsl:message>
</xsl:when>
<xsl:otherwise>utf-8</xsl:otherwise>
</xsl:choose>
</encoding>
</p:input>
<p:output name="data" id="encoding" debug="encoding"/>
</p:processor>
<!-- Store the document --> <!-- Store the document -->
<p:processor name="oxf:file-serializer"> <p:processor name="oxf:file-serializer">
<p:input name="config"> <p:input name="config" transform="oxf:xslt" href="#encoding">
<config> <config xsl:version="2.0">
<scope>session</scope> <scope>session</scope>
<encoding>utf-8</encoding> <encoding>
<xsl:value-of select="/encoding"/>
</encoding>
<force-encoding>true</force-encoding> <force-encoding>true</force-encoding>
</config> </config>
</p:input> </p:input>
<p:input name="data" href="#record#xpointer(/record/content/document)"/> <p:input name="data" href="#record#xpointer(/record/content/document)"/>
<p:output name="data" id="url-written"/> <p:output name="data" id="url-written" debug="url-written"/>
</p:processor> </p:processor>
<!-- And read it as HTML --> <!-- And read it as HTML -->
<p:processor name="oxf:url-generator"> <p:processor name="oxf:url-generator">
<p:input name="config" transform="oxf:xslt" href="#url-written"> <p:input name="config" transform="oxf:xslt" href="aggregate('root', #url-written, #encoding)">
<config xsl:version="2.0"> <config xsl:version="2.0">
<url> <url>
<xsl:value-of select="/*"/> <xsl:value-of select="/root/url"/>
</url> </url>
<encoding>utf-8</encoding> <encoding>
<xsl:value-of select="/root/encoding"/>
</encoding>
<force-encoding>true</force-encoding> <force-encoding>true</force-encoding>
<content-type>text/html</content-type> <content-type>text/html</content-type>
<force-content-type>true</force-content-type> <force-content-type>true</force-content-type>
@ -75,7 +98,7 @@
<p:processor name="oxf:html-converter"> <p:processor name="oxf:html-converter">
<p:input name="config"> <p:input name="config">
<config> <config>
<content-type>application/xml</content-type> <content-type>text/html</content-type>
<encoding>utf-8</encoding> <encoding>utf-8</encoding>
</config> </config>
</p:input> </p:input>