When the ImportDocuments workflow has to create a new object (i.e. when there is no sourceName/sourceId matching as detailed above) it uses the XLST Config\Global\GN4_Tark4_Common\xsl_ImportDocumentsNameMap.xml to compute its name and destination folder.
Currently this XSLT use a new GUID for the name, and places the objects in a sub-folder of the root archiving folder named as the first publication referenced in the XML being imported - so for example if the importDocumentsRootFolderPath parameter is set to /system/Archive, importing the following:
<root>
<document id="doc0" name="pallino.indd">
<data mime="application/x-indesign">pallino.indd</data>
<issue publication="Demo" date="2011-01-30" name="Main" number="Default"/>
<pages>
<page name="1">
<preview mime="image/jpeg">Pages/1.jpg</preview>
</page>
</pages>
<stories>
<ref idRef="story257"/>
</stories>
<images>
<ref idRef="img493"/>
</images>
</document>
<story id="story257">
<data mime="text/plain">Stories/257.txt</data>
</story>
<image id="img493" name="iptc.jpg" >
<data mime="image/jpeg">Links/iptc.jpg</data>
</image>
</root>
will create three objects (a document, a story and an image) in the '/system/Archive/Demo' folder - named with three new GUID.
If there is no publication the workflow will use a sub-folder named '_'.
Note that these folders must already exist.
The naming and folder placement logic can be changed modifying xsl_ImportsDocumentNameMap.xml - provided that the name/folder pair for each object is always unique.
Here is an extract from XSL which generates the folderPath:
<xsl:attribute name="folderPath">
<xsl:choose>
<xsl:when test="string($publication)">
<xsl:value-of select="concat(@rootFolderPath,'/',$publication,'/',$year)"/>
</xsl:when>
<xsl:otherwise>
<xsl:value-of select="concat(@rootFolderPath,'/_')"/>
</xsl:otherwise>
</xsl:choose>
</xsl:attribute>
Note
The usage of many folder's levels is deprecated since the folderPath is only an internal and logical split and it is not visible anywhere in the archive. For example, avoid to create structure such as Archive/Publication/Year/Month/Day.