Consider the single-level class hierarchy shown in Figure 1.1 above, which is an incomplete disjoint segmentation of the class Book, as the design for the model classes of an MVC app. In such a case, whenever we have a model class hierarchy with only one level (or only a few levels) of subtyping and each subtype has only a few additional properties, it's preferable to use STI, so we model a single table containing columns for all attributes such that the columns representing additional attributes of segment subclasses ("segment attributes") are optional, as shown in the SQL table model in Figure 1.9 below.

Figure 1.9. An SQL table model with a single table representing the Book class hierarchy

It is a common approach to add a special discriminator column for representing the category of each row corresponding to the subclass instantiated by the represented object. Such a column would normally be string-valued, but constrained to one of the names of the subclasses. If the DBMS supports enumerations, it could also be enumeration-valued. We use the name category for the discriminator column, which, in the case of our Book class hierarchy example, has a frozen value constraint since the textbook-biography segmentation is rigid.

Based on the category of a book, we have to enforce that if and only if it is "TextBook", its attribute subjectArea has a value, and if and only if it is "Biography", its attribute about has a value. This implied constraint is expressed in the invariant box attached to the Book table class in the class diagram above, where the logical operator keyword "IFF" represents the logical equivalence operator "if and only if". It needs to be implemented in the database, e.g., with an SQL table CHECK clause or with SQL triggers.

When the given segmentation is disjoint, a single-valued enumeration attribute category is used for representing the information to which subclass an instance belongs. Otherwise, if it is non-disjoint, a multi-valued enumeration attribute categories is used for representing the information to which subclasses an instance belongs. Such an attribute can be implemented in SQL by defining a string-valued column for representing a set of enumeration codes or labels as corresponding string concatenations.

Consider the class hierarchy shown in Figure 1.8 above. With only three additional attributes defined in the subclasses Employee, Manager and Author, this class hierarchy can again be mapped with the STI approach, as shown in the SQL table model Figure 1.10 below.

Figure 1.10. An STI table model representing the Person roles hierarchy

Notice that now the discriminator column categories is multi-valued, since the segmentation of Person is not disjoint, but overlapping, implying that a Person object may belong to several categories. Notice also that, since a role segmentation (like Employee, Manager, Author) is not rigid, the discriminator column categories does not have a frozen value constraint.

An example of an admissible population for this model is the following:

Notice that the Person table contains four different types of people:

Pros of the STI approach: It leads to a faithful representation of the subtype relationships expressed in the original class hierarchy; in particular, any row representing a subclass instance (an employee, manager or author) also represents a superclass instance (a person).

Cons: (1) In the case of a multi-level class hierarchy where the subclasses have little in common, the STI approach does not lead to a good representation. (2) The structure of the given class hierarchy in terms of its elements (classes) is only implicitly preserved.

In a more realistic model, the subclasses of Person shown in Figure 1.8 above would have many more attributes, so the STI approach would be no longer feasible. In the TCI approach we get the SQL table model shown in Figure 1.11 below. A TCI model represents each concrete class of the class hierarchy as a table, such that each segment subclass is represented by a table that also contains columns for inherited properties, thus repeating the columns of the table that represents the superclass.

Figure 1.11. A TCI table model representing the Person roles hierarchy

A TCI table model can be derived from the information design model by performing the following steps:

Each table would only be populated with rows corresponding to the direct instances of the represented class. An example of an admissible population for this model is the following:

Pros of the TCI approach: (1) The structure of the given class hierarchy in terms of its elements (classes) is explicitly preserved. (2) When the segmentations of the given class hierarchy are disjoint, TCI leads to memory-efficient non-redundant storage.

Cons: (1) The TCI approach does not yield a faithful representation of the subtype relationships expressed in the original class hierarchy. In particular, for any row representing a subclass instance (an employee, manager or author) there is no information that it represents a superclass instance (a person). Thus, the TCI database schema does not inform about the represented subtype relationships; rather, this meta-information, which is kept in the app's class model, is de-coupled from the database. (2) The TCI approach requires repeating column definitions, which is a form of schema redundancy. (3) The TCI approach may imply data redundancy whenever the segment subclasses overlap. In our example, authors can also be employees, so for any person in the overlap, we would need to duplicate the data storage for all columns representing properties of the superclass (in our example, this only concerns the property name).

For avoiding the data redundancy problem of TCI in the case of overlapping segmentations, we could take the JTI approach as exemplified in the SQL table model shown in Figure 1.12 below. This model connects tables representing subclasses (subtables) to tables representing their superclasses (supertables) by defining their primary key column(s) to be at the same time a foreign key referencing their supertable's primary key. Notice that foreign keys are visualized in the form of UML dependency arrows stereotyped with «fkey» and annotated at their source table side with the name of the foreign key column.

Figure 1.12. A JTI table model representing the Person roles hierarchy

An example of an admissible population for this model is the following:

Pros of the JTI approach: (1) Subtyping relationships and the structure of class hierarchies are explicitly preserved. (2) Data redundancy in the case of overlapping segmentations is avoided.

Cons: (1) The main disadvantage of the JTI approach is that for querying a subclass, join queries (for joining the segregated entity data) are required, which may create performance issues.

These steps are discussed in more detail in the following sections.

The enumeration type BookCategoryEL is coded with the help of our library meta-class Enumeration at the beginning of the Book.js model class file in the following way:

BookCategoryEL = new Enumeration([ "Textbook", "Biography"]);

We code the model class Book in the form of an ES6 class definition where the category attribute as well as the segment attributes subjectArea and about are optional, with getters, setters and static check functions for all properties:

class Book {
  constructor ({isbn, title, year, category, subjectArea, about}) {
    this.isbn = isbn;
    this.title = title;
    this.year = year;
    // optional properties
    if (category) this.category = category;
    if (subjectArea) this.subjectArea = subjectArea;
    if (about) this.about = about;
  }
  get isbn() {...}
  static checkIsbn( isbn) {...}
  static checkIsbnAsId( isbn) {...}
  set isbn( isbn) {...}
  get title() {...}
  static checkTitle( t) {...}
  set title( t) {...}
  get year() {...}
  static checkYear( y) {...}
  set year( y) {...}
  get category() {...}
  static checkCategory( c) {...}
  set category( c) {...}
  get subjectArea() {...}
  static checkSubjectArea( sA, cat) {...}
  set subjectArea( s) {...}
  get about() {...}
  static checkAbout( a, cat) {...}
  set about( a) {...}
}

Notice that the constructor function is defined with a single record parameter using the ES6 syntax for function parameter destructuring.

We code the checkCategory and setCategory methods for the category attribute in the following way:

static checkCategory( c) {
  if (c === undefined || c === "") {
    return new NoConstraintViolation();  // category is optional
  } else if (!isIntegerOrIntegerString(c) || parseInt(c) < 1 ||
      parseInt(c) > BookCategoryEL.MAX) {
    return new RangeConstraintViolation(
        "Invalid value for category: "+ c);
  } else {
    return new NoConstraintViolation();
  }
};

set category( c) {
  var validationResult = null;
  if (this.category) {  // already set/assigned
    validationResult = new FrozenValueConstraintViolation(
        "The category cannot be changed!");
  } else {
    validationResult = Book.checkCategory( c);
  }
  if (validationResult instanceof NoConstraintViolation) {
    this._category = parseInt( c);
  } else {
    throw validationResult;
  }
}

While the getters for segment properties (in this example: subjectArea and about) follow the standard pattern, their checks and setters have to make sure that the property applies to the category of the instance being checked. This is achieved by checking a combination of a property value and a category, as in the following example:

static checkSubjectArea( sA, c) {
  if (c === BookCategoryEL.TEXTBOOK && !sA) {
    return new MandatoryValueConstraintViolation(
        "A subject area must be provided for a textbook!");
  } else if (c !== BookCategoryEL.TEXTBOOK && sA) {
    return new ConstraintViolation("A subject area must not " +
        "be provided if the book is not a textbook!");
  } else if (sA && (typeof(sA) !== "string" || sA.trim() === "")) {
    return new RangeConstraintViolation(
        "The subject area must be a non-empty string!");
  } else {
    return new NoConstraintViolation();
  }
} 

In the serialization function toString, we serialize the category attribute and the segment properties in a switch statement:

toString() {
  var bookStr = `Book{ ISBN: ${this.isbn}, title: ${this.title}, year: ${this.year}`;
  switch (this.category) {
    case BookCategoryEL.TEXTBOOK:
      bookStr += `, textbook subject area: ${this.subjectArea}`;
      break;
    case BookCategoryEL.BIOGRAPHY:
      bookStr += `, biography about: ${this.about}`;
      break;
  }
  return bookStr + "}";
};

In the update method of a model class, we only set a property if it is to be updated, that is, if there is a corresponding argument slot with a value that is different from the old property value. In the special case of a category attribute with a Frozen Value Constraint, we need to make sure that it can only be updated, along with an accompanying set of segment properties, if it has not yet been assigned. Thus, in the Book.update method, we perform the special test if book.category === undefined for handling the special case of an initial assignment, while we handle updates of the segment properties subjectArea and about in a more standard way:

Book.update = function ({isbn, title, year, 
    category, subjectArea, about}) {
  const book = Book.instances[isbn],
        objectBeforeUpdate = cloneObject( book);
  var noConstraintViolated=true, updatedProperties=[];
  try {
    ...
    if (category && book.category !== category) {
      book.category = category;
      updatedProperties.push("category");
    } else if (category === "" && "category" in book) {
      throw FrozenValueConstraintViolation(
          "The book category cannot be unset!");
    }
    if (subjectArea && book.subjectArea !== subjectArea) {
      book.subjectArea = subjectArea;
      updatedProperties.push("subjectArea");
    }
    if (about && book.about !== about) {
      book.about = about;
      updatedProperties.push("about");
    }
  } catch (e) {
    ...
  }
  ...
};

We have to take care of handling the category attribute and the segment properties subjectArea and about both in the "Retrieve and list all books" use case as well as in the "Create book" and "Update book" use cases by

We add a "Category" column to the view table of the "Retrieve/list all books" use case in books.html:

<table id="books">
  <thead><tr>
   <th>ISBN</th><th>Title</th><th>Year</th><th>Category</th>
  </tr></thead>
  <tbody></tbody>
</table>

A book without a special category will have an empty table cell in this column, while for all other books their category will be shown in this column, along with other category-specific information. This requires a corresponding switch statement in the v/books.js file:

if (book.category) {
  switch (book.category) {
  case BookCategoryEL.TEXTBOOK:
    row.insertCell().textContent = book.subjectArea + " textbook";
    break;
  case BookCategoryEL.BIOGRAPHY: 
    row.insertCell().textContent = "Biography about "+ book.about;
    break;
  }
}

In both use cases, we need to allow selecting a special category of book ('textbook' or 'biography') with the help of a selection field, as shown in the following HTML fragment:

<div class="field">
 <label>Category: <select name="category"></select></label>
</div>
<div class="field Textbook"><!-- conditional field -->
 <label>Subject area: <input type="text" name="subjectArea" /></label>
</div>
<div class="field Biography"><!-- conditional field -->
 <label>About: <input type="text" name="about" /></label>
</div>

Notice that we have added "Textbook" and "Biography" as additional values of the class attribute of the segment field container elements. This supports the rendering and un-rendering of "Textbook" and "Biography" form fields, depending on the value of the category attribute.

In the handleCategorySelectChangeEvent handler, segment property form fields are only displayed, with displaySegmentFields, when a corresponding book category has been selected:

handleCategorySelectChangeEvent = function (e) {
  const formEl = e.currentTarget.form,
        // the array index of BookCategoryEL.labels
        categoryIndexStr = formEl.category.value;
  if (categoryIndexStr) {
    displaySegmentFields( formEl, BookCategoryEL.labels,
        parseInt( categoryIndexStr) + 1);
  } else {
    undisplayAllSegmentFields( formEl, BookCategoryEL.labels);
  }
};

Recall that the category selection list contains a no-selection option "---" with the empty string as its return value, and a list of options formed by the enumeration labels of BookCategoryEL.labels such that their value is the corresponding array index (starting with 0) as a string. Consequently, the variable categoryIndexStr has either the value "" (empty string) or one of "0", "1", "2", etc.

In the case of a superclass like Person, we define a class-level property subtypes for having a mechanism to loop over all subtypes of a superclass.

class Person {...}
Person.subtypes = [];

The property subtypes holds a list of all subtypes of the given class. This list is initially empty.

The subtype relationships between the classes Employee and Person, as well as between Author and Person, are defined with the help of the ES6 keywords extends and super. For instance, in m/Author.js we define:

class Author extends Person {
  constructor ({personId, name, biography}) {
    super({personId, name});  // invoke Person constructor
    // assign additional properties
    this.biography = biography;
  }
  get biography() {return this._biography;}
  set biography( b) {this._biography = b;}  /***SIMPLIFIED CODE: no validation ***/
  toString() {...}
}
// add Author to the list of Person subtypes
Person.subtypes.push( Author);

When retrieving the instances of a class hierarchy's root class (in our example, Person) from a persistent data store organized according to the TCI approach, we have to retrieve not only its direct instances from the table representing the root class (people), but also all indirect instances from all tables representing its subclasses (employees and authors), as shown in the following code:

Person.retrieveAll = function () {
  var people = {};
  if (!localStorage["people"]) localStorage["people"] = "{}";
  try {
    people = JSON.parse( localStorage["people"]);
  } catch (e) {
    console.log("Error when reading from Local Storage\n" + e);
  }
  for (const key of Object.keys( people)) {
    try {  // convert record to (typed) object
      Person.instances[key] = new Person( people[key]);
    } catch (e) {
      console.log(`${e.constructor.name} ...`);
    }
  }
  // add all instances of all subtypes to Person.instances
  for (const Subtype of Person.subtypes) {
    Subtype.retrieveAll();
    for (const key of Object.keys( Subtype.instances)) {
      Person.instances[key] = Subtype.instances[key];
    }
  }
  console.log(`${Object.keys( Person.instances).length} records loaded`);
}

For any subtype (here, Author and Employee), each record is retrieved and a corresponding entry is created in the map Subtype.instances and copied to Person.instances.

Since the app's data is kept in main memory as long as the app is running (which is as long as the app's webpage is kept open in the browser), the data has to be saved to persistent storage when the app is terminated (e.g., by closing its browser tab). When saving the instances of Person (as records of the people table) to persistent storage in v/people.js, we also save the direct instances of ist subtypes Employee and Author (as records of the JS entity tables employees and authors in v/employees.js and v/authors.js). This is necessary because changes to Person instances may imply changes of Employee or Author instances.

We do this in v/people.js:

// save data when leaving the page
window.addEventListener("beforeunload", function () {
  Person.saveAll();
  // save all subtypes for persisting changes of supertype attributes
  for (const Subtype of Person.subtypes) {
    Subtype.saveAll();
  }
});