Javascript Connect to Link -- Pseudocode examples

Create an instance of the DataService class. Failure likely means Link is not installed.

dataService = new ActiveXObject(“Scantron.Scanning.DataService”);

Hopefully you stored the 14-char SDK license key in web.config or other secure place on your server. Call your app to fetch the key and assign it to Link. Send the status back to your application, perhaps for troubleshooting downstream version/license issues you might encounter.

getLicenseKey(function(licenseKey){

dataService.License = licenseKey

log (dataService.Version)

log (dataService.License)

})

To do web based data collection, follow these steps…

Create a zero footprint session.

clientContext = {}

clientContext.DataFileOption = L_DATA_FILE_OPTON_NOFILES

Populate with URLs to the application definition and profiles. Be aware of case sensitivity for your web server!

clientContext.Application = “../../Content/MyApp.StxAppDef.bin”

We want Link to perform data validation, so use an edit profile

clientContext.EditProfile = “../../Content/MyApp.StxEdtPrf.bin”

Also want scoring performed after successful correction (optional)

clientContext.ScoreProfile = “../../Content/MyApp.StxScrPrf.bin”

Prepare to open a connection to the data service. This entails downloading the scanning metadata and finding the connected scanner. OpenAsync does all of this and notifies us.

Create a client callback object and add the following handlers

clientCallback = {}

Handle the OpenCompleted callback

function OpenCompleted(dataServiceEventArgs){

Obtain a reference to the scanning engine bound to the data service.

scanningManager = dataServiceEventArgs.ScanningManager

}

As each document is successfully scanned, the scanning engine raises the DocComplete event. See if the scanned data has any data errors. If so, somebody will correct the data later, so save it to a temporary store on your server. If the data passed the edit examination phase, transform the data to XML. You should specify a SQL query that generates the XML schema expected by the business tier in your app. Notice that the data is saved in the native ScanTools format. This allows us to apply an edit profile when the same or another user is ready to correct the data, possibly in the same or another web session.

Handle the ScanningManager.DocComplete event.

function DocComplete(scanEventArgs){

// obtain a reference to the document just scanned

doc = scanEventArgs.Document

if (doc.EditState == EditState.esPass){

// document passes edit examination

// transform data to XML

myXMLQuery = “My FOR XML SQL select statement”

xmlData = doc.ExecuteQuery(myXmlQuery)

// send the data home along with the perfected data

saveStudent(doc.DataRecordInNCSFormat, xmlData)

}

else{

// send the data home. It will be queued.

saveStudent (doc.DataRecordInNCSFormat)

}

}

Open the data service, the metadata is downloaded by Link and is available till you call Close

dataService.OpenAsync(ClientContext, clientCallback)

When done, have Link clean up any local resources it used.

dataService.Close()

Add a help topic…

Perfecting Scanned Data for Downstream Use

You can make use of an edit profile to drive the data validation and correction process. You use ScanTools Plus at design-time tool to build edit profiles which are a declarative way to specify data validation rules. An edit profile can be applied at scan-time. It can also be used in a correction-only session. You can use one edit profile at scan-time and another post-scan. You can choose to implement your own edit correction rules and methodology or complement the features provided by Link. You can choose to use the edit correction user interface in Link or you can write your own interface. Link allows you to make the tradeoffs, supporting you where needed and staying out of your way at other times. The most important advantage of the built in user interface is time to market. If this is your first exposure to ScanTools Link programming, you saved yourself a bunch of time and effort using well-known functionality. Typical reasons for implementing a custom user interface would be to support additional features such as international cultures, regular expression based validation, spell-check, database driven validation or simply to match the look and feel of your solution.

To create a web based edit-only correction session, follow this procedure.

1. Specify the application definition and profiles via appropriate URLs that are meaningful to the site of origin of your solution. By default, web servers are unaware of ScanTools file extensions and may not allow these files to be downloaded. This example assumes that the web server allows the “.bin” extension, so the ScanTools files are stored with an additional web server friendly extension:

An application must be specified for us to obtain resolved data (as opposed to the OMR read levels).

clientContext.Application = “../../Content/MyApp.StxAppDef.bin”

2. Set an edit profile

clientContext.EditProfile = “../../Content/MyApp.StxEdtPrf.bin”

3. Have scoring performed after successful correction (optional)

clientContext.ScoreProfile = “../../Content/MyApp.StxScoPrf.bin”

4. Open the data service, the metadata is downloaded by Link

dataService.OpenAsync(ClientContext,clientCallback)

5. Obtain the first (or subsequent) data record to be corrected by calling your web service.

getStudent(function (data){

// got back data that was previous stored in NCS ASCII format

dataRecordInNCSFormat = data

// Hydrate an IDocument providing object

doc = dataService.LoadDocument (dataRecordInNCSFormat)

// perhaps enable an Edit button on the UI.

})

6. Present the record to the user, perhaps in response to an Edit button click event

succeded = dataService.ShowEditCorrectionDialog(doc)

If (succeeded){

// scanned data is perfected, transform it in a format suitable to your business tier

xmlData = doc.ExecuteQuery(myXmlQuery)

// send the data home

saveStudent (doc.DataRecordInNCSFormat, xmlData)

….

// fetch the next record for edit

getStudent()

}

else{

// data record is still failing one or more edits

// Data may be partially corrected, so send it back

// Maybe present it to another user at a later time

saveStudent (doc.DataRecordInNCSFormat)

}

7. When done, have Link clean up any local resources it used. This is typically done when the page unloads.

dataService.Close()

To correct data at scan-time, you have two options:

• do it inline as each document is scanned
• decouple correction from scanning

Which method you use depends upon the scanner, your application and the customer preference. We do not recommend that you do any inline correction on a high speed scanner like the iNSIGHT 70/150. If you have an iNSIGHT 4 and doing real-time printing of say a server issued ticket or a score, correction may have to be done inline.

To customize the data correction user interface, create your own UI and data binding. Use a SQL select statement to get the scanned data. Use a SQL Update statement to send the data back to Link. Reapply the edit profile using a select on the Edit Failures table.