Step 1. Launch Aspire and open the Content Source Management Page

Launch Aspire (if it's not already running). See:

Step 2. Add a new Content Source

  • For this step please follow the step from the Configuration Tutorial of the connector of you choice, please refer to Connector list

Step 3. Add a new Microsoft Search publisher to the Workflow

To add a Publish to Microsoft Search drag a Microsoft Search publisher rule from the Workflow Library and drop to the Workflow Tree where you want to add it. This will automatically open the Microsoft Search publisher window for the configuration of the publisher.


Step 3a. Specify Publisher Information

 In the publisher window, specify the connection information to publish to the Microsoft Search.

  • Name: Unique name for the publisher
  • Tenant Id:  The tenant Id provided by Microsoft
  • Client Id:  The client Id generated when the Application was registered in Prerequisites
  • Client Secret:  The client secret that was generated in Prerequisites
  • Index Name:  The name of the connection/index that will be created in MS Search
  • Groovy Transform:  The path to a Groovy transformation file that will process the document to make it match the expected structure from either the fixed ExternalFile or the custom ExternalItem data types in MS Search connection schemas. To see the default of the files, please refer to the Where are the default Groovy transformation and custom schema located?
  • Use custom schema:  Enables the usage of the limited custom schema in MS Search.  If disabled it will assume ExternalFile schema, otherwise it will be ExternalItem and will require the path to a JSON schema file
  • Start/end actions Clear:  In order to create the connection automatically on full crawls this must be enabled, otherwise it will be assumed the connection exists

Other configuration items are common to every publisher component.


ExternalFile vs ExternalItem

Microsoft Search allows the usage of one of two schemas:  fixed ExternalFile or custom ExternalItem.  ExternalItem allows limited freedom to define properties to be expected from crawled items.  Needless to say, the Groovy transformation file must yield an output that matches the expected schema.

ExternalFile schema

The fixed external file schema expects the following information:

  • acl:  The list of ACLs for the document
  • createdDateTime:  A standard UTC string that represents the creation date and time
  • modifiedDateTime:  A standard UTC string that represents the last modification date and time
  • createdBy:  The author's name
  • lastModifiedBy:  The name of the last person that modified the document
  • title:  The document's title
  • url:  The document's url
  • name:  The document's name
  • extension:  The document's file name extension
  • size:  The document size
  • content:  The document's content

This is a sample document in JSON format as expected by the REST API:

{
	"acl": [
		{
			"type": "user",
			"value": "d411eb08-42e2-4316-aab5-2df8e9d9c21b",
			"accessType": "grant",
			"identitySource": "Azure Active Directory"
		}
	],
	"createdDateTime": "2017-11-08T19:06:17Z",
	"modifiedDateTime": "2017-11-08T19:06:17Z",
	"createdBy": "empty",
	"lastModifiedBy": "empty",
	"title": "sample document",
	"url": "http://the.url.com",
	"name": "name.txt",
	"extension": "txt",
	"size": 10,
	"content": "the content/n"
}


ExternalItem schema

As mentioned before, the ExternalItem schema has limited customization capabilities.  It expects the following information:

  • acl:  The list of ACLs for the document
  • properties:  A name/value list of custom properties of predefined types
  • content:  The document's content

When configured to use custom schema, the publisher component expects a txt file with the following structure:


{
	"properties": [
		{
			"name": "propertyName",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		}
	]
}

Where:

  • name:  The property name in JSON friendly characters
  • type:  The data type of the property. Can be one of the following:
    • String
    • Int64
    • Double
    • DateTime (as UTC offset: YYYY-MM-DDTHH:mm:ss.nnZ)
    • Boolean
    • Collection(String)
    • Collection(Int64)
    • Collection(Double)
    • Collection(DateTime)
  • isSearchable:  (optional) If true, the property will be indexed as containing searchable content
  • isRetrievable:  (optional) If true, the property can be retrieved as part of search results
  • isQueryable:  (optional) If true, the property can be queried for specific values

This is a the default schema file provided with the component (schemaProperties.json):

{
	"properties": [
		{
			"name": "id",
			"type": "String"
		},
		{
			"name": "name",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		},
		{
			"name": "createdDateTime",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		},
		{
			"name": "modifiedDateTime",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		},
		{
			"name": "title",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true"
		},
		{
			"name": "url",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		},
		{
			"name": "description",
			"type": "String",
			"isSearchable": "true",
			"isRetrievable": "true",
			"isQueryable": "true"
		}
	]
}


Groovy Transformation file

The Groovy transformation file makes it easy to output data that is customized to the client's needs and also that can be safely conveyed to Microsoft Search through the REST API.  The output of the transformation must match the expected schema structure.

This is the default Groovy transformation file that is provided with the component (aspireToMicrosoftSearchBulk.groovy):


import java.math.BigInteger;
import java.security.MessageDigest;
import java.security.NoSuchAlgorithmException;
import java.nio.charset.Charset;

def connectorSpecificMap = [
	'isContainer':'is_container'
]

def getContent(String content) {
	try {
		if(content.getBytes().length > 16777216L) {
			return content.substring(0,10485760) + "...";
		} else {
			return content
		}
	} catch(Throwable t) {
		return "";
	}
}

def getMD5(String id) {
  
  MessageDigest digest = MessageDigest.getInstance("MD5")  
  String md5name = new BigInteger(1, digest.digest(id.getBytes())).toString(16)
  return md5name;
}

// Function that process the children of a connector specific field
def getChildren(name, parent) {

  builder."$name"() {
    parent.getChildren().each() { val ->

      def attr = val.getName();

      //if it has other children
      if(val.getChildren().size() > 0) {
        getChildren(attr, val);
      } else {
        builder."$attr"() {

          //All the attributes
          val.getAttributeNames().each() { attrName ->
            "@$attrName" val.getAttribute(attrName);
          }

          //Main content
          if (val?.getText() != null) {
            '_$' val?.getText();
          }
        }
      }
    }
  }
}

//***************************************************
//
// Main routine
//

// Action of the job
String action = doc.action.getText();

if ((action == "add") || (action == "update")) {
  /*****************
   * Add or Update *
   *****************/
  
  builder.$object() {
    '@search.action' "upload"

	String newId = "";
    // Get ID
    if (doc.id != null) {
      newId = getMD5(doc.id.getText())
    } else if (doc.fetchUrl != null) {
      newId = getMD5(doc.fetchUrl.getText())
    } else if (doc.url != null) {
      newId = getMD5(doc.url.getText())
    } else if (doc.displayUrl != null) {
      newId = getMD5(doc.displayUrl.getText())
    } else {
      newId = "ID-NOT-PROVIDED"
    }
	
	'id' newId


  // Document Source
    String[] nameDoc = doc.id.getText().split('/');
	String finalName = nameDoc[nameDoc.size()-1];

    name finalName
    title finalName
    description finalName	  

    if(doc.lastModified != null) {
      modifiedDateTime doc.lastModified
	  createdDateTime doc.lastModified
    }
	
    if (doc.fetchUrl != null) {
      url doc.fetchUrl
    } else if (doc.url != null) {
      url doc.url
    } else if (doc.displayUrl != null) {
      url doc.displayUrl
    } else {
      url "URL-NOT-PROVIDED"
    }

    // ACLs	
    if (doc.acls != null) {
       builder.acls() {
         $list { 
           doc.acls.getChildren().each() { val ->
             $object() {
               name val.getAttribute("name")
               access val.getAttribute("access")
               entity val.getAttribute("entity")
             }
           }
         }
       }
	
	
	//END
    }

    if(doc.content != null) {
      content doc.content?.getText()
    }

  }
} else {
  /**********
   * Delete *
   **********/

  builder.$object() {
    '@search.action' "delete"

	String delId = "";
    // Get ID
    if (doc.id != null) {
      delId = getMD5(doc.id.getText())
    } else if (doc.fetchUrl != null) {
      delId = getMD5(doc.fetchUrl.getText())
    } else if (doc.url != null) {
      delId = getMD5(doc.url.getText())
    } else if (doc.displayUrl != null) {
      delId = getMD5(doc.displayUrl.getText())
    } else {
      delId = "ID-NOT-PROVIDED"
    }
	
	'id' delId
  }
}




Once you've clicked on the Add button, it will take a moment for Aspire to download all of the necessary components (the Jar files) from the Maven repository and load them into Aspire. Once that's done, the publisher will appear in the Workflow Tree.

For details on using the Workflow section, please refer to Workflow introduction.