Advanced ColdFusion Administration
Indexing XML Documents

Style Files

The following style files are required to enable indexing of XML files. Default style files are installed into in the cfusion\verity\common\style directory (Windows) and opt/coldfusion/verity/common/style directory (Linux and UNIX).

Style File	Description
style.uni	Invokes the XML filter for indexing XML documents.
style.xml	Modifies the default behavior of the XML filter. (optional)
style.ufl	Defines custom fields in XML documents. The fields must also be defined in the style.xml file.
style.dft	Invokes the Verity universal filter by default so all document types can be indexed into one collection. You can modify the style.dft file to invoke the XML filter instead of the universal filter, as described below.

Configuring style files

This section discusses style file configuration used to support XML document filtering.

style.uni file

To index XML documents, the style.uni must include the following lines:

type: "text/xml"

  /format-filter = "flt_xml"

  /charset= guess

  /def-charset = 8859

Configuring the style.xml file

By default, the XML filter indexes regions of the document delimited by XML tags as zones, with the zones given the same name as the XML tag. META tags are automatically indexed as fields unless they are in a suppressed region.

To modify the default behavior, you create a style file named style.xml. You can specify field and zone indexing for regions of the document delimited by XML tags and skip regions of the document delimited by XML tags.

<?xml version="1.0" encoding="ISO-8859-1"?>

<?note: this is a sample comment line?>

<style.xml version="2.6.0">

  <?note:

   ? this following line dictates all xmltags be ignored

   ? <ignore xmltag="*" />

   ?>

  

  <?note:

   ? "ignore" will skip indexing xmltag, yet index contents

   ? between the beginning and end of this pair of xmltags

   ?>

  <?next 2 sample lines commented out: 

  <ignore xmltag="section_1" />

  <ignore xmltag="section_2" />

   ?>



  <?note:

   ? "preserve" indexes xmltag as zone with the presence of 

   ? <ignore xmltag="*" />

   ?>

  <?next 1 sample line commented out: 

  <preserve xmltag="section_3" />

   ?>



  <?note:

   ? "suppress" will suppress every xmltag embedded within

   ?>

  <?next 2 sample lines commented out: 

  <suppress xmltag="region_1" />

  <suppress xmltag="region_3" />

   ?>



  <?note:

   ? "field" will further index content between the beginning 

   ? and end of this pair of xmltags as field values

   ?>

  <?next 1 sample line commented out: 

  <field xmltag="column_1" />

   ?>



  <?note:

   ? if attribute "fieldname" is present, above content will

   ? be indexed into VDK field under the value of fieldname

   ? instead of the field under the name of xmltag

   ?>

  <?next 1 sample line commented out: 

  <field xmltag="column_2" fieldname="vdk_field_2" />

   ?>



  <?note:

   ? if attribute "index" is set to "override", above content

   ? will be indexed into VDK field overriding values read in

   ? from bulk insert file, if any

   ?>

  <?next 1 sample line commented out: 

  <field xmltag="column_3" index="override" />

   ?>



  <?note: 

   ? fieldname & index attributes could both exist

   ?>

</style.xml>

style.xml command syntax

<command attribute="value"/>

Use these commands in the style.xml file to manage how Verity handles individual XML elements. Refer to the style.xml file listing for examples of these commands.

Command	Description
field	Indexes the content between the pair of specified XML tags as field values. By default, the field name is the same as the xmltag value, unless otherwise specified by the fieldname attribute. Attributes: xmltag fieldname index
ignore	Skips indexing of xmltag but indexes the content between the pair of specified XML tags. Attribute: xmltag
preserve	Indexes specified xmltag as a zone if preceded by ignore xmltag = "*". Attribute: xmltag
suppress	Suppresses every xmltag embedded within the specified xmltag. Attribute: xmltag

style.xml command examples

The following command ignores all XML tags in the document, indexing only the content:

<ignore xmltag = "*"/>

The following command skips indexing the specified xmltag but indexes the content between the start and end tags of the specified xmltag:

<ignore xmltag = "section_1"/>

The following command indexes xmltag as a zone if there is also an ignore xmltag = "*" command:

<preserve xmltag = "section_1"/>

The following command suppresses the entire element identified by xmltag. The tag, attribute, and content are not indexed:

<suppress xmltag = "section_1"/>

The following command indexes the content between the start and end tags of the specified xmltag as a field, which is given the same name as xmltag:

<field xmltag = "column_1"/>

The following command indexes the content between the start and end tags of the specified xmltag as a field, which is given the name specified in the fieldname attribute:

<field xmltag = "column_2" fieldname = "vdk_field_2"/>

The following command indexes the content between the start and end tags of the specified xmltag as a field, overriding any existing value of the field:

<field xmltag = "column_2" index = "override"/>

style.ufl file

If administrators have defined custom fields to be populated in the style.xml file, the fields must also be defined in the style.ufl file or style.sfl file, using standard syntax.

style.dft file

To create a collection that contains only XML documents, administrators can modify the style.dft file to invoke the XML filter directly. In this case, the XML documents do not need a .xml extension.

The style.dft must include the following lines:

$control: 1

dft:

{

  field: DOC

  filter="flt_xml"

}

Copyright © 2001, Macromedia Inc. All rights reserved.

---