Implementing personal data collection

To comply with the requirements of personal data regulations, such as the GDPR, you need to provide a way for administrators (data protection officers) to collect personal data stored within the system. This is necessary to resolve personal data queries from data subjects, and requests to transfer personal data to another system or application.

Kentico does not provide any personal data collection functionality by default. This requires exact knowledge of how your website gathers, processes and stores personal data. You need to implement the data collection based on the specifics of your website and the nature of the legal requirements that you wish to fulfill.

Use the following process to develop the personal data collection:

1. Open your Kentico solution in Visual Studio.

2. Create custom classes that implement collector interfaces:

   • Identity collectors – map real-world identifiers, such as email addresses or names, to corresponding Kentico objects that represent data subjects.

   • Data collectors – process the objects added by identity collectors, retrieve related personal data, and format the results into a string containing either human-readable text or machine-readable data (such as XML).

     Class location

     We recommend adding custom classes as part of a new assembly (Class library project) in your Kentico solution. You need to add the appropriate references to both the assembly and the main Kentico web project.

3. Add a custom module class and register your collector implementations within the module’s OnInit method:

   • You can register any number of collectors.
   • To register identity collectors, call the IdentityCollectorRegister.Instance.Add method. The registration order is significant – the identities added by a collector can be accessed by identity collectors that are registered after.
   • To register data collectors, call the PersonalDataCollectorRegister.Instance.Add method.

The registered collectors allow users to search for personal data in the Data protection application. To give users the option to delete the collected data from the system (or specific parts of it), you also need to implement erasure functionality – see Implementing personal data erasure.

Identity collectors

When searching for personal data in the Data protection application, users submit real-world identifiers of data subjects (email addresses, names, etc.). To collect personal data, you first need to create Identity collectors that convert these identifiers into Kentico objects representing matching data subjects, such as users (UserInfo), contacts (ContactInfo) or customers (CustomerInfo).

Identity collectors are classes that implement the IIdentityCollector interface. Every implementation must contain the Collect method, which processes the following parameters:

• IDictionary<string, object> – a dictionary holding the submitted identifiers and other filtering parameters. The default identifier input only provides an email address value, which is available under the “email” key.
• List<BaseInfo> – a list of Kentico objects representing data subjects. Contains all objects added by previously registered IIdentityCollector implementations.

To create your own IIdentityCollector implementations:

1. Use the ObjectQuery API to load Kentico objects (instances of Info classes) that match the submitted identifier values (available in the Collect method’s first parameter).
2. Add the retrieved objects to the list of identity objects (the Collect method’s second parameter).

Customizing the identifier inputs

By default, the Data protection application allows users to search for personal data based on an email address value. If you wish to set up personal data collection according to other types of identifiers or add filtering options, you need to create and register a custom control:

1. Open your Kentico solution in Visual Studio.

2. Create a new Web User Control (.ascx file) in the Kentico web project (CMSApp or CMS).

3. Add components to the control’s markup that allow users to input the required identifier values and/or set filtering parameters.

4. Switch to the code behind and make the control class inherit from DataSubjectIdentifiersFilterControl (available in the CMS.UIControls namespace).

5. Override the following methods:

   • GetFilter – return an IDictionary<string, object> collection containing all identifier values and filtering parameters that users input through the control.

     Example

     Copy

     
     
     
       public override IDictionary<string, object> GetFilter(IDictionary<string, object> filter)
       {
           filter.Add("email", txtEmail.Text);
     
           return filter;
       }
     
     
       

   • IsValid – return a bool value that indicates whether the control’s input is valid. Call the AddError method to display messages to users in cases where the input is not valid.

     Example

     Copy

     
     
     
       public override bool IsValid()
       {
           if (!CMS.Helpers.ValidationHelper.IsEmail(txtEmail.Text))
           {
               AddError("Please enter a valid email address.");
     
               return false;
           }
     
           return true;
       }
     
     
       

6. Edit the module class where you register your collector implementations, and register your identifier input control within the module’s OnInit method:

   • Call the DataProtectionControlsRegister.Instance.RegisterDataSubjectIdentifiersFilterControl method (available in the CMS.UIControls namespace).
   • Specify the path of the user control file in the method’s parameter, for example: ~/CMSModules/CustomDataProtection/DataSubjectIdentifiers.ascx

The system now uses your custom control to display the identifier inputs in the Data protection application. The keys and values added via the control’s GetFilter method are available as the second parameter of the Collect method in your IIdentityCollector implementations.

Data collectors

After you implement Identity collectors, you need to create Data collectors. These collectors retrieve personal data related to the identity objects provided by the registered identity collectors, and process the results into a suitable format. We recommend creating separate data collectors for different object types, depending on the types of personal data that you process on your website.

Data collectors are classes that implement the IPersonalDataCollector interface. Every implementation must contain the Collect method.

• The method’s IEnumerable<BaseInfo> parameter provides the identity objects added by your IIdentityCollector implementations. You can convert the BaseInfo objects to specific types, such as UserInfo, ContactInfo, etc.

• The method’s second parameter is a string that specifies the requested output format. By default, the following fixed values are used for the output format:

  • machine – when searching for data on the Data portability tab of the Data protection application. The value is available in the API under the PersonalDataFormat.MACHINE_READABLE constant. By default, the system assumes that machine-readable data is in XML format and adds a <PersonalData> root tag around the final output. To return data in another format, such as JSON, you need to Customize the data output.
  • human – when searching for data on the Right to access and Right to be forgotten tabs in the Data protection application. The value is available in the API under the PersonalDataFormat.HUMAN_READABLE constant.

To create your own IPersonalDataCollector implementations:

1. Collect all required personal data related to the provided identity objects.

2. Format the data into a string.

   We recommend creating a dedicated writer class for every required output format. You can reuse the API of general writer classes to build the personal data string for different types of objects.

   You may also need to define transformation functions that convert internally stored values into more understandable equivalents. For example, the system uses integer values for some fields that have a fixed set of possible options (e.g. the user gender field).

3. Return a PersonalDataCollectorResult object in the Collect method, with the resulting personal data string assigned into the Text property.

The text provided by registered data collectors is displayed when searching for personal data in the Data protection application.

Customizing the data output

By default, the overall data returned by the collection process is composed according to the registration order of your data collectors (starting from the first to the last). Additionally, the system assumes that machine-readable data is in XML format and adds a <PersonalData> root tag around the final output of all data collectors.

If you wish to return machine-readable data in another format, such as JSON, or otherwise adjust the data composition process, use the following customization approach:

1. Open your Kentico solution in Visual Studio.

2. Create a custom class that inherits from PersonalDataHelper (available in the CMS.DataProtection namespace).

3. Override the JoinPersonalDataInternal method of the helper class:

   • The method’s IEnumerable<string> parameter provides a collection of personal data strings added by your registered IPersonalDataCollector implementations.
   • The method’s second parameter is a string that specifies the requested output format.
   • Compose the final data output according to your custom requirements and return it as a string.

4. Register your helper class implementation using the RegisterCustomHelper assembly attribute.

using System;
using System.Linq;
using System.Text;
using System.Collections.Generic;

using CMS;
using CMS.DataProtection;

// Registers the CustomPersonalDataHelper class to replace the default PersonalDataHelper
[assembly: RegisterCustomHelper(typeof(CustomPersonalDataHelper))]

public class CustomPersonalDataHelper : PersonalDataHelper
{
    // Customizes the output of personal data in machine-readable format
    // Replaces the default <PersonalData> XML root tag with curly brackets for JSON data
    protected override string JoinPersonalDataInternal(IEnumerable<string> personalData, string outputFormat)
    {
        // Performs custom handling for personal data in machine-readable format
        if (outputFormat.Equals(PersonalDataFormat.MACHINE_READABLE, StringComparison.OrdinalIgnoreCase))
        {
            string indentation = "  ";
            string indentedNewLine = Environment.NewLine + indentation;

            var resultBuilder = new StringBuilder();

            // Adds the opening bracket to the result
            resultBuilder.AppendLine("{");
            resultBuilder.Append(indentation);

            // Updates all new line characters in the returned personal data to include the added indentation
            var modifiedPersonalData = personalData.Select(data => data.Replace(Environment.NewLine, indentedNewLine));

            // Adds the data provided by all registered personal data collectors
            resultBuilder.AppendLine(String.Join(indentedNewLine, modifiedPersonalData));

            // Adds the closing bracket to the result
            resultBuilder.Append("}");

            return resultBuilder.ToString();
        }

        // Runs the default method for human-readable or undefined output formats
        return base.JoinPersonalDataInternal(personalData, outputFormat);
    }
}

The system now uses your custom data joining implementation when displaying results in the Data protection application.

Example – Creating contact data collectors

The following example demonstrates how to implement personal data collection for basic contact data. The sample collectors work with the default email address identifier that users can submit in the Data protection application, and produce personal data output in plain text or XML format.

• Adding a custom class library
• Creating the identity collector
• Implementing writer classes
• Creating the data collector
• Registering the collectors

Adding a custom class library

Start by preparing a separate project for custom classes in your Kentico solution:

1. Open your Kentico solution in Visual Studio.

2. Create a new Class Library project in the Kentico solution (or reuse an existing custom project).

3. Add references to the required Kentico libraries (DLLs) for the new project:

   1. Right-click the project and select Add -> Reference.

   2. Select the Browse tab of the Reference manager dialog, click Browse and navigate to the Lib folder of your Kentico web project.

   3. Add references to the following libraries (and any others that you may need in your custom code):

      • CMS.Base.dll
      • CMS.ContactManagement.dll
      • CMS.Core.dll
      • CMS.DataEngine.dll
      • CMS.DataProtection.dll
      • CMS.Helpers.dll

4. Reference the custom project from the Kentico web project (CMSApp or CMS).

5. Edit the custom project’s AssemblyInfo.cs file (in the Properties folder).

6. Add the AssemblyDiscoverable assembly attribute:

   Copy

   
   
   
    using CMS;
   
    [assembly:AssemblyDiscoverable]
   
   
    

You can now add your collector implementations and other related classes under the custom project.

Creating the identity collector

To create an identity collector for contacts, add a new class implementing the IIdentityCollector interface under the custom project:

using System;
using System.Collections.Generic;
using System.Linq;

using CMS.DataEngine;
using CMS.DataProtection;
using CMS.ContactManagement;

public class ContactIdentityCollector : IIdentityCollector
{
    public void Collect(IDictionary<string, object> dataSubjectFilter, List<BaseInfo> identities)
    {
        // Does nothing if the identifier inputs do not contain the "email" key or if its value is empty
        if (!dataSubjectFilter.ContainsKey("email"))
        {
            return;
        }
        string email = dataSubjectFilter["email"] as string;
        if (String.IsNullOrWhiteSpace(email))
        {
            return;
        }

        // Finds contacts with a matching email address
        List<ContactInfo> contacts = ContactInfoProvider.GetContacts()
                                            .WhereEquals(nameof(ContactInfo.ContactEmail), email)
                                            .ToList();

        // Adds the matching contact objects to the list of collected identities
        identities.AddRange(contacts);
    }
}

The collector loads all contact objects (ContactInfo) that match the submitted email address identifier, and adds them to the list of collected identities.

Implementing writer classes

Continue by creating writer classes that convert Kentico objects into the required output formats (plain text and XML in this example).

For plain text, add the following writer class under the custom project:

using System;
using System.Collections.Generic;
using System.Text;

using CMS.DataEngine;

public class TextPersonalDataWriter
{
    private readonly StringBuilder stringBuilder;
    private int indentationLevel;

    public TextPersonalDataWriter()
    {
        stringBuilder = new StringBuilder();
        indentationLevel = 0;
    }

    // Writes horizontal tabs based on the current indentation level
    private void Indent()
    {
        stringBuilder.Append('\t', indentationLevel);
    }

    // Writes text representing a new section of data, and increases the indentation level
    public void WriteStartSection(string sectionName)
    {
        Indent();

        stringBuilder.AppendLine(sectionName + ": ");
        indentationLevel++;
    }

    // Writes the specified columns of a Kentico object (BaseInfo) and their values
    public void WriteObject(BaseInfo baseInfo, List<Tuple<string, string>> columns)
    {
        foreach (var column in columns)
        {
            // Gets the name of the current column
            string columnName = column.Item1;
            // Gets a user-friendly name for the current column
            string columnDisplayName = column.Item2;

            // Filters out identifier columns from the human-readable text data
            if (columnName.Equals(baseInfo.TypeInfo.IDColumn, StringComparison.Ordinal) || 
                columnName.Equals(baseInfo.TypeInfo.GUIDColumn, StringComparison.Ordinal))
            {
                continue;
            }

            // Gets the value of the current column for the given object
            object value = baseInfo.GetValue(columnName);

            if (value != null)
            {
                Indent();
                stringBuilder.AppendFormat("{0}: ", columnDisplayName);
                stringBuilder.Append(value);
                stringBuilder.AppendLine();
            }
        }
    }

    // "Closes" a text section by reducing the indentation level
    public void WriteEndSection()
    {
        indentationLevel--;
    }

    // Gets a string containing the writer's overall text
    public string GetResult()
    {
        return stringBuilder.ToString();
    }
}

For machine-readable data in XML format, add the following writer class under the custom project. The sample XML writer is an IDisposable implementation that uses the standard .NET XmlWriter class to create the required output.

using System;
using System.Collections.Generic;
using System.Text;
using System.Xml;

using CMS.DataEngine;
using CMS.Helpers;

public class XmlPersonalDataWriter : IDisposable
{
    private readonly StringBuilder stringBuilder;
    private readonly XmlWriter xmlWriter;

    public XmlPersonalDataWriter()
    {
        stringBuilder = new StringBuilder();
        xmlWriter = XmlWriter.Create(stringBuilder, new XmlWriterSettings { Indent = true, OmitXmlDeclaration = true });
    }

    // Writes an opening XML tag with a specified name
    public void WriteStartSection(string sectionName)
    {
        // Replaces period characters in object names with underscores
        sectionName = sectionName.Replace('.', '_');

        xmlWriter.WriteStartElement(sectionName);
    }

    // Writes XML tags representing the specified columns of a Kentico object (BaseInfo) and their values
    public void WriteObject(BaseInfo baseInfo, List<string> columns)
    {
        foreach (string column in columns)
        {
            object value = baseInfo.GetValue(column);

            if (value != null)
            {
                xmlWriter.WriteStartElement(column);
                xmlWriter.WriteValue(XmlHelper.ConvertToString(value));
                xmlWriter.WriteEndElement();
            }
        }
    }

    // Writes a closing XML tag for the most recent open tag
    public void WriteEndSection()
    {
        xmlWriter.WriteEndElement();
    }

    // Gets a string containing the writer's overall XML data
    public string GetResult()
    {
        xmlWriter.Flush();

        return stringBuilder.ToString();
    }

    // Releases all resources used by the current XmlPersonalDataWriter instance.    
    public void Dispose()
    {
        xmlWriter.Dispose();
    }
}

Creating the data collector

To create a data collector for contacts, add a class implementing the IPersonalDataCollector interface under the custom project:

using System;
using System.Collections.Generic;
using System.Linq;

using CMS.ContactManagement;
using CMS.DataEngine;
using CMS.DataProtection;

public class ContactDataCollector : IPersonalDataCollector
{
    // Prepares a list of contact columns to be included in the personal data
    // Every Tuple contains a column name, and user-friendly description of its content
    private readonly List<Tuple<string, string>> contactColumns = new List<Tuple<string, string>> {
        Tuple.Create("ContactFirstName", "First name"),
        Tuple.Create("ContactMiddleName", "Middle name"),
        Tuple.Create("ContactLastName", "Last name"),
        Tuple.Create("ContactJobTitle", "Job title"),
        Tuple.Create("ContactAddress1", "Address"),
        Tuple.Create("ContactCity", "City"),
        Tuple.Create("ContactZIP", "ZIP"),
        Tuple.Create("ContactMobilePhone", "Mobile phone"),
        Tuple.Create("ContactBusinessPhone", "Business phone"),
        Tuple.Create("ContactEmail", "Email"),
        Tuple.Create("ContactBirthday", "Birthday"),
        Tuple.Create("ContactGender", "Gender"),
        Tuple.Create("ContactNotes", "Notes"),
        Tuple.Create("ContactGUID", "GUID"),
        Tuple.Create("ContactLastModified", "Last modified"),
        Tuple.Create("ContactCreated", "Created"),
        Tuple.Create("ContactCampaign", "Campaign"),
        Tuple.Create("ContactCompanyName", "Company name")
    };    

    public PersonalDataCollectorResult Collect(IEnumerable<BaseInfo> identities, string outputFormat)
    {
        // Gets a list of all contact objects added by registered IIdentityCollector implementations
        List<ContactInfo> contacts = identities.OfType<ContactInfo>().ToList();

        // Uses a writer class to create the personal data, in either XML format or as human-readable text
        string contactData = null;
        if (contacts.Any())
        {
            switch (outputFormat.ToLowerInvariant())
            {
                case PersonalDataFormat.MACHINE_READABLE:
                    contactData = GetXmlContactData(contacts);
                    break;
                case PersonalDataFormat.HUMAN_READABLE:
                default:
                    contactData = GetTextContactData(contacts);
                    break;
            }
        }

        return new PersonalDataCollectorResult
        {
            Text = contactData
        };
    }

    private string GetXmlContactData(List<ContactInfo> contacts)
    {
        using (var writer = new XmlPersonalDataWriter())
        {
            // Wraps the contact data into a <OnlineMarketingData> tag
            writer.WriteStartSection("OnlineMarketingData");

            foreach (ContactInfo contact in contacts)
            {
                // Writes a tag representing a contact object
                writer.WriteStartSection(ContactInfo.OBJECT_TYPE);
                // Writes tags for the contact's personal data columns and their values
                writer.WriteObject(contact, contactColumns.Select(t => t.Item1).ToList());
                // Closes the contact object tag
                writer.WriteEndSection();
            }

            // Closes the <OnlineMarketingData> tag
            writer.WriteEndSection();

            return writer.GetResult();
        }
    }

    private string GetTextContactData(List<ContactInfo> contacts)
    {
        var writer = new TextPersonalDataWriter();

        writer.WriteStartSection("On-line marketing data");

        foreach (ContactInfo contact in contacts)
        {
            writer.WriteStartSection("Contact");
            // Writes user-friendly descriptions of the contact's personal data columns and their values
            writer.WriteObject(contact, contactColumns);
            writer.WriteEndSection();
        }

        return writer.GetResult();
    }
}

The sample data collector processes the contact objects provided by the identity collector, and then uses the writer classes to create personal data text in the requested output format.

Registering the collectors

To register the identity and data collectors, add a module class under the custom project, and run the required initialization code:

using CMS;
using CMS.DataEngine;
using CMS.DataProtection;

// Registers the custom module into the system
[assembly: RegisterModule(typeof(CustomDataProtectionModule))]

internal class CustomDataProtectionModule : Module
{
    // Module class constructor, the system registers the module under the name "CustomDataProtection"
    public CustomDataProtectionModule()
        : base("CustomDataProtection")
    {
    }

    // Contains initialization code that is executed when the application starts
    protected override void OnInit()
    {
        base.OnInit();

        // Adds the ContactIdentityCollector to the collection of registered identity collectors
        IdentityCollectorRegister.Instance.Add(new ContactIdentityCollector());

        // Adds the ContactDataCollector to the collection of registered personal data collectors
        PersonalDataCollectorRegister.Instance.Add(new ContactDataCollector());
    }
}

Save all changes and Build the custom project.

You can now search for the email addresses of contacts on the Data portability and Right to access tabs in the Data protection application. If matching contacts exist in the system, the registered collectors return their data in XML or plain text format.