Batch job processing in upCast Enterprise is controlled by so-called Batch Configuration Files (BCFs) that describe the batch job operations to be executed on source documents. They are executed by the Batch Processor import filter.

In order to access the batch job functionality you have to select the Batch Processor import filter. Do this via the Import Filter popup menu.

After you have selected the Batch Processor import filter you can choose a batch configuration file either via the Browse… button next to the Source File input field or you can directly type in the name and access path of the batch configuration file. In doing so, to separate folder names, you must use the forward slash ('/') on Unix and Macintosh systems. On Windows machines, use the backslash ('\') character instead. The slashes will be converted to the appropriate, system specific separator character during file access. Batch job file names must end in .bcf.

In order to initiate a batch job, click Start Conversion. This initiates a detailed check of the chosen batch configuration file.

This check comprises the following:

At the end of the check you get a survey about the number of documents to be converted and their accumulated size.

If you want to proceed with the execution of the batch job, click Execute Batch, otherwise click Cancel Batch.

A batch job consists of several phases, as determined in the batch configuration file. During the processing of these phases a progress bar will be displayed that provides a Cancel button. You can abort the batch job at any time by clicking this button. All already successfully written files will remain untouched.

A Batch Configuration File (BCF) is actually an instance of an XML document written in the BCF language. It consists of one or more blocks describing the filter configurations and the source files to be converted using these filter configurations.

A BCF is considered an input file for the Batch Processor import filter. In other words, what normally in upCast is the RTF file to be converted by the RTF 1.6 import filter, is now the BCF file.

Internally, a BCF sets up the conversion core of upCast and calls either the (none) or the RTF 1.6 import filter for each file according to the rules specified in the BCF.

A batch run as described in a BCF consists of one or more conversion elements.

A conversion specifies the complete filter configuration and one or more source elements, which specify what documents (specified by file elements) have to be converted.

A configuration consists of the specification of the export filter elements to be used, where each can have its own param element set.

A param element has a a name, value attribute pair.

The source element specifies the sourcepath, i.e. the path where the file elements to be converted will be found, and the destpath, where the converted files will be written to. In conjunction with the (optional) use of so-called Post Actions, you may also specify successpath and failurepath attributes.

A file element can either have a specificname attribute and thus specify a single source document. Or it can be of a selective nature in order to collect several files based on certain criteria, like all files starting or ending with a certain character combination or even all RTF files in the sourcepath. If the file element is empty and has no selection attributes, it indicates that the filters set up should be executed exactly once.

<run> 
 <conversion>
   <configuration>
     <import type="rtf" />
     <filter type="XML">
       <param name="Extension" value=".xml" />
     </filter>
     <filter type="XHTML">
       <param name="Extension" value=".html" />
     </filter>
     <filter type="CSS">
       <param name="Extension" value=".css" />
     </filter>
   </configuration>
   <source sourcepath="file:///C:/upCast/rtf"
           destpath="file:///C:/upCast/rtf/converted">
     <file endswith=".rtf" skipexisting="yes" />
   </source>
 </conversion>
</run>

In the following we describe the actual DTD for BCFs in human language for each element. For those who better like the rather concise DTD notation, here it is (though a bit sloppy when it comes to attribute values):

<ELEMENT run (setvar*,conversion*)>
<ATTLIST run
         mode (interactive|silent) 'interactive'>
<ELEMENT setvar EMPTY>
<ATTLIST name CDATA #REQUIRED
         value CDATA #REQUIRED>
<ELEMENT conversion (configuration,source*)>
<ATTLIST conversion
         maxdocs CDATA #IMPLIED>
<ELEMENT configuration (import,filter+)>
<ELEMENT import (param)*>
<ATTLIST import
         type (none|rtf) #REQUIRED>
<ELEMENT filter (param)*>
<ATTLIST filter
         type CDATA #REQUIRED
<ELEMENT param EMPTY>
<ATTLIST param
         name CDATA #REQUIRED
         value CDATA #IMPLIED>
<ELEMENT source (file)*>
<ATTLIST source
         sourcepath CDATA #IMPLIED
         destpath CDATA #IMPLIED
         imagedestpath CDATA #IMPLIED
         postaction CDATA #IMPLIED
         successfolder CDATA #IMPLIED
         failurefolder CDATA #IMPLIED>
<ELEMENT file EMPTY>
<ATTLIST file
         endswith CDATA #IMPLIED
         startswith CDATA #IMPLIED
         skipexisting (yes|no) 'no'
         name CDATA #IMPLIED
         anyfile (yes|no) 'no'>

Deleting results of previous run. Suppose you wish to delete generated files of a previous run before performing a conversion job. This can be done as follows:

<run mode="interactive">
  <conversion >
    <configuration>
      <import type="none" />
      <filter type="Commandline">
        <param name="Commandline" value="rm &quot;%S&quot;" />
      </filter>
    </configuration>
    <source sourcepath="/Users/chris/conv/" postaction="none" >
      <file endswith=".css" />
      <file endswith=".xml" />
    </source>
  </conversion>
  <conversion>
    <configuration>
      <import type="rtf" />
      <filter type="XML">
        <param name="Extension" value=".xml" />
        <!-- further parameters omitted... -->
      </filter>
      <filter type="CSS">
        <param name="Extension" value=".css"/>
      </filter>
    </configuration>
    <source sourcepath="/Users/chris/orig/" 
            destpath="/Users/chris/conv/"
            postaction="none">
      <file endswith=".rtf" />
    </source>
  </conversion>
</run>

The above example deletes all *.css and *.xml files from the output directory of the second conversion by utilizing the Unix command rm in conjunction with the upCast variable %S. Note that the argument to rm must be quoted to allow file names containing spaces or special shell characters and that the quotes itself need to be quoted using "since they are within an attribute.

Validating a folder of XML files. Suppose you wish to use upCast for quickly validating all XML files in the folder /Users/chris/xml/, then this should do it:

<run mode="interactive">
  <conversion >
    <configuration>
      <import type="none" />
      <filter type="XMLValidator">
        <param name="InputFile" value="%S" />
      </filter>
    </configuration>
    <source sourcepath="/Users/chris/xml/" postaction="none" >
      <file endswith=".xml" />
    </source>
  </conversion>
</run>

The above example validates all *.xml files in the /Users/chris/xml/ directory.