model.xml

<project name='project' pubsub='auto' threads='1' use-tagged-token='true'> 
<description> 
	This example is a basic demonstrtation of using the tokenization algorithm to
	perform text tokenization on streaming input. The project contains a single 
	continuous query consisting of the following: 
	1.	A source window that receives the text data to be analyzed 
	2.	A calculate window that tokenizes the text in incoming data events
		and publishes the results
</description>
<contqueries>
  <contquery name='contquery' include-singletons='true' trace='w_source w_calculate'>
   <windows>
    <window-source name='w_source' insert-only='true' index='pi_EMPTY'> 
    <description> 
	This source window receives the included example input data file via a file
	socket connector. The input stream is placed into two fields for each observation:
	a document ID that acts as the data stream's key, named docId; and a string of 
	incoming text, named doc.
    </description>
     <schema>
      <fields>
       <field name='docId'  type='int64' key='true'/>
       <field name='doc' type='string'/>
      </fields>
     </schema>
     <connectors>
      <connector class='fs' name='publisher'>
       <properties>
        <property name='type'>pub</property>
        <property name='fstype'>csv</property>
        <property name='fsname'>input.csv</property>
        <property name='transactional'>true</property>
        <property name='blocksize'>1</property>
       </properties>
      </connector>
     </connectors>
    </window-source>
    <window-calculate name="w_calculate" algorithm="Tokenization"> 
    <description> 
	This calculate window receives data events from the source window, and publishes 
	word tokens created by the Tokenization algorithm.In this example, the following
	input-map and output-map properties govern the calculate window: 

	docId:		Specifies the input variable for the unique docID. 
	doc:		Specifies the input variable for the input doc from the source window. 

	docIdOut:	Specifies the output variable for the unique doc ID. 
	tokenIdOut:	Specifies the output variable for the unique ID of the token. 
	wordOut:	Specifies the output variable for the word content in the token. 
	startPosOut:	Specifies the output variable for the starting position of the token word. 
	endPosOut:	Specifies the output variable for the ending position of the word. 

	The resulting output is written to the file result.out via a file socket connector.
    </description>
      <schema>
        <fields>
          <field name='docId' type='int64' key='true'/>
          <field name='tokenId' type='int64' key='true'/>
          <field name='word' type='string'/>
          <field name='startPos' type='int32'/>
          <field name='endPos' type='int32'/>
        </fields>
      </schema>
        <input-map>
          <properties>
           <property name="docId">docId</property>
           <property name="doc">doc</property>
          </properties>
        </input-map>
        <output-map>
          <properties>
           <property name="docIdOut">docId</property>
           <property name="tokenIdOut">tokenId</property>
           <property name="wordOut">word</property>
           <property name="startPosOut">startPos</property>
           <property name="endPosOut">endPos</property>
          </properties>
        </output-map>
     <connectors>
      <connector class='fs' name='sub'>
       <properties>
        <property name='type'>sub</property>
        <property name='fstype'>csv</property>
        <property name='fsname'>result.out</property>
        <property name='snapshot'>true</property> 
        <property name='header'>true</property>
       </properties>
      </connector>
     </connectors>
    </window-calculate>
   </windows>
   <edges>
    <edge source='w_source' target='w_calculate' role='data'/>
   </edges>
  </contquery>
</contqueries>
</project>