Refactoring before supporting CSS

This commit is contained in:
Eric van der Vlist 2012-04-13 11:16:40 +02:00
parent 11027c068a
commit 16cc943d48
4 changed files with 144 additions and 103 deletions

View File

@ -24,8 +24,7 @@
<!-- Store the archive in the database --> <!-- Store the archive in the database -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data"> <p:input name="data" transform="oxf:xslt" href="#data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath> <relpath>
@ -49,109 +48,33 @@
<p:choose href="#archive"> <p:choose href="#archive">
<!-- HTML document : need to update the links... --> <!-- HTML document : need to update the links... -->
<p:when test="/archive/response/document/@content-type='text/html'"> <p:when test="/archive/response/document/@content-type=('text/html')">
<!-- Store the document --> <!-- Call the corresponding pipeline to extract the links and rewrite them -->
<p:processor name="oxf:file-serializer">
<p:input name="config">
<config>
<scope>session</scope>
</config>
</p:input>
<p:input name="data" href="#archive#xpointer(/archive/response/document)"/>
<p:output name="data" id="url-written"/>
</p:processor>
<!-- 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="#archive">
<config xsl:version="2.0"> <config xsl:version="2.0">
<url> <url>
<xsl:value-of select="/*"/> <xsl:text>oxf:/actions/mediatypes/</xsl:text>
<xsl:value-of select="substring-after(/archive/response/document/@content-type, '/')"/>
<xsl:text>.xpl</xsl:text>
</url> </url>
<mode>html</mode>
</config> </config>
</p:input> </p:input>
<p:output name="data" id="html" debug="html"/> <p:output name="data" id="pipeline"/>
</p:processor> </p:processor>
<!-- Get a list of links to update --> <p:processor name="oxf:pipeline">
<!-- TODO: support links in inline CSS --> <p:input name="config" href="#pipeline"/>
<p:processor name="oxf:unsafe-xslt"> <p:input name="archive" href="#archive"/>
<p:input name="data" href="#html"/> <p:output name="rewritten" id="rewritten"/>
<p:input name="request" href="#archive#xpointer(/archive/request)"/> <p:output name="links" id="links"/>
<p:input name="config">
<xsl:stylesheet version="2.0">
<xsl:variable name="base" select="doc('input:request')/request/location"/>
<xsl:template match="/">
<links>
<xsl:variable name="links" as="node()*">
<xsl:apply-templates/>
</xsl:variable>
<xsl:for-each-group select="$links" group-by="@href">
<xsl:variable name="abs-href" select="resolve-uri(@href, $base)"/>
<xsl:variable name="tokens" select="tokenize($abs-href, '/')"/>
<xsl:variable name="last-token" select="$tokens[last()]"/>
<xsl:variable name="tokens2" select="tokenize($last-token, '\.')"/>
<xsl:variable name="extension" select="$tokens2[last()]"/>
<link abs-href="{$abs-href}" new-href="{saxon:string-to-hexBinary(substring($abs-href, 1, string-length($abs-href) - string-length($extension) - 1), 'utf-8')}.{$extension}"
filename="{saxon:string-to-hexBinary($abs-href, 'utf-8')}.xml">
<xsl:copy-of select="@*"/>
</link>
</xsl:for-each-group>
</links>
</xsl:template>
<xsl:template match="text()"/>
<xsl:template match="link[@rel='stylesheet']">
<link>
<xsl:copy-of select="@*"/>
</link>
</xsl:template>
<xsl:template match="img">
<link href="{@src}" type="image/*"/>
</xsl:template>
<xsl:template match="script[@src]">
<link href="{@src}" type="{@type}"/>
</xsl:template>
</xsl:stylesheet>
</p:input>
<p:output name="data" id="links" debug="links"/>
</p:processor> </p:processor>
<!-- Update the links -->
<p:processor name="oxf:unsafe-xslt">
<p:input name="data" href="#html"/>
<p:input name="request" href="#archive#xpointer(/archive/request)"/>
<p:input name="links" href="#links"/>
<p:input name="config">
<xsl:stylesheet version="2.0">
<xsl:variable name="links" select="doc('input:links')/links"/>
<xsl:variable name="base" select="doc('input:request')/request/location"/>
<xsl:key name="link" match="link" use="@href"/>
<xsl:template match="@*|node()">
<xsl:copy>
<xsl:apply-templates select="@*|node()"/>
</xsl:copy>
</xsl:template>
<xsl:template match="link[@rel='stylesheet']/@href|img/@src|script/@src">
<xsl:attribute name="{name(.)}">
<xsl:value-of select="$links/key('link', current())/@new-href"/>
</xsl:attribute>
</xsl:template>
<xsl:template match="link[@rel!='stylesheet']/@href|a/@href">
<xsl:attribute name="{name(.)}">
<xsl:value-of select="resolve-uri(., $base)"/>
</xsl:attribute>
</xsl:template>
</xsl:stylesheet>
</p:input>
<p:output name="data" id="rewritten" debug="rewritten"/>
</p:processor>
<!-- Store the rewritten document in the database --> <!-- Store the rewritten document in the database -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data"> <p:input name="data" transform="oxf:xslt" href="#data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath> <relpath>
@ -174,8 +97,7 @@
<!-- Update the archive index --> <!-- Update the archive index -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data"> <p:input name="data" transform="oxf:xslt" href="#data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath> <relpath>
@ -213,8 +135,7 @@ for $as in /archive-set
<!-- Update the queue --> <!-- Update the queue -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="aggregate('root', #data, #links)"> <p:input name="data" transform="oxf:xslt" href="aggregate('root', #data, #links)">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath>queue.xml</relpath> <relpath>queue.xml</relpath>
@ -266,8 +187,7 @@ for $a in /queue/action where $a/@uuid = $(uuid) return
<p:otherwise> <p:otherwise>
<!-- Update the archive index --> <!-- Update the archive index -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data"> <p:input name="data" transform="oxf:xslt" href="#data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath> <relpath>
@ -301,8 +221,7 @@ for $as in /archive-set
<!-- Update the queue --> <!-- Update the queue -->
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data"> <p:input name="data" transform="oxf:xslt" href="#data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath>queue.xml</relpath> <relpath>queue.xml</relpath>

View File

@ -40,8 +40,7 @@
</p:processor> </p:processor>
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" href="#data-access-data"/> <p:input name="data" href="#data-access-data"/>
<p:input name="param"> <p:input name="param">
<xquery><![CDATA[ <xquery><![CDATA[
@ -66,8 +65,7 @@ for $a in /queue/action where $a/@uuid = $(uuid) return
</p:processor> </p:processor>
<p:processor name="oxf:pipeline"> <p:processor name="oxf:pipeline">
<p:input name="config" href="/data-access.xpl <p:input name="config" href="/data-access.xpl"/>
"/>
<p:input name="data" transform="oxf:xslt" href="#data-access-data"> <p:input name="data" transform="oxf:xslt" href="#data-access-data">
<config xsl:version="2.0"> <config xsl:version="2.0">
<relpath> <relpath>

View File

@ -0,0 +1,13 @@
Pipelines in this directory are called by the archive-resource pipeline.
Their name is the name of the media subtype.
Inputs:
* archive: the archive
Outputs:
* rewritten: the rewritten version of the document
* links: the list of rewritten links

View File

@ -0,0 +1,111 @@
<p:config xmlns:p="http://www.orbeon.com/oxf/pipeline" xmlns:oxf="http://www.orbeon.com/oxf/processors" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" xmlns:saxon="http://saxon.sf.net/">
<p:param name="archive" type="input"/>
<p:param name="rewritten" type="output"/>
<p:param name="links" type="output"/>
<!-- Store the document -->
<p:processor name="oxf:file-serializer">
<p:input name="config">
<config>
<scope>session</scope>
</config>
</p:input>
<p:input name="data" href="#archive#xpointer(/archive/response/document)"/>
<p:output name="data" id="url-written"/>
</p:processor>
<!-- And read it as HTML -->
<p:processor name="oxf:url-generator">
<p:input name="config" transform="oxf:xslt" href="#url-written">
<config xsl:version="2.0">
<url>
<xsl:value-of select="/*"/>
</url>
<mode>html</mode>
</config>
</p:input>
<p:output name="data" id="html" debug="html"/>
</p:processor>
<!-- Get a list of links to update -->
<!-- TODO: support links in inline CSS -->
<!-- TODO: support iframes and objects -->
<p:processor name="oxf:unsafe-xslt">
<p:input name="data" href="#html"/>
<p:input name="request" href="#archive#xpointer(/archive/request)"/>
<p:input name="config">
<xsl:stylesheet version="2.0">
<xsl:variable name="base" select="doc('input:request')/request/location"/>
<xsl:template match="/">
<links>
<xsl:variable name="links" as="node()*">
<xsl:apply-templates/>
</xsl:variable>
<xsl:for-each-group select="$links" group-by="@href">
<xsl:variable name="abs-href" select="resolve-uri(@href, $base)"/>
<xsl:variable name="tokens" select="tokenize($abs-href, '/')"/>
<xsl:variable name="last-token" select="$tokens[last()]"/>
<xsl:variable name="tokens2" select="tokenize($last-token, '\.')"/>
<xsl:variable name="extension" select="$tokens2[last()]"/>
<link abs-href="{$abs-href}" new-href="{saxon:string-to-hexBinary(substring($abs-href, 1, string-length($abs-href) - string-length($extension) - 1), 'utf-8')}.{$extension}"
filename="{saxon:string-to-hexBinary($abs-href, 'utf-8')}.xml">
<xsl:copy-of select="@*"/>
</link>
</xsl:for-each-group>
</links>
</xsl:template>
<xsl:template match="text()"/>
<xsl:template match="link[@rel='stylesheet']">
<link>
<xsl:copy-of select="@*"/>
</link>
</xsl:template>
<xsl:template match="img">
<link href="{@src}" type="image/*"/>
</xsl:template>
<xsl:template match="script[@src]">
<link href="{@src}" type="{@type}"/>
</xsl:template>
</xsl:stylesheet>
</p:input>
<p:output name="data" id="links-local" debug="links"/>
</p:processor>
<p:processor name="oxf:identity">
<p:input name="data" href="#links-local"/>
<p:output name="data" ref="links"/>
</p:processor>
<!-- Update the links -->
<p:processor name="oxf:unsafe-xslt">
<p:input name="data" href="#html"/>
<p:input name="request" href="#archive#xpointer(/archive/request)"/>
<p:input name="links" href="#links-local"/>
<p:input name="config">
<xsl:stylesheet version="2.0">
<xsl:variable name="links" select="doc('input:links')/links"/>
<xsl:variable name="base" select="doc('input:request')/request/location"/>
<xsl:key name="link" match="link" use="@href"/>
<xsl:template match="@*|node()">
<xsl:copy>
<xsl:apply-templates select="@*|node()"/>
</xsl:copy>
</xsl:template>
<xsl:template match="link[@rel='stylesheet']/@href|img/@src|script/@src">
<xsl:attribute name="{name(.)}">
<xsl:value-of select="$links/key('link', current())/@new-href"/>
</xsl:attribute>
</xsl:template>
<xsl:template match="link[@rel!='stylesheet']/@href|a/@href">
<xsl:attribute name="{name(.)}">
<xsl:value-of select="resolve-uri(., $base)"/>
</xsl:attribute>
</xsl:template>
</xsl:stylesheet>
</p:input>
<p:output name="data" ref="rewritten" debug="rewritten"/>
</p:processor>
</p:config>