As important as it is to know the control structures of a programming language, it is just as important to be able to name their contexts and surroundings. Using the right vocabulary is particularly effective in code reviews as it supports us to put our thoughts into words.
- 1. Algorithms & Functions
- 1.1. Algorithm
- 1.2. Declarative Programming
- 1.3. Imperative Programming
- 1.4. Function Declaration
- 1.5. Function Expression
- 1.6. Referentially Opaque Expressions
- 1.7. Referentially Transparent Expressions
- 1.8. Block Scope
- 1.9. Function Scope
- 1.10. Deterministic Functions
- 1.11. Identity Function
- 1.12. Pure Functions
- 2. TypeScript
- 2.1. Ambient Context
- 2.2. Array Destructuring
- 2.3. Collective Types
- 2.4. Compiler
- 2.5. Discriminated Unions
- 2.6. Literal Types
- 2.7. Module
- 2.8. Non-primitive types
- 2.9. Primitive Types
- 2.10. String Literal Type
- 2.11. Tagged template
- 2.12. Template Literal Type
- 2.13. Template Literal
- 2.14. Transpiler
- 2.15. Tuple Type
- 2.16. Type Annotation
- 2.17. Type Argument Inference
- 2.18. Type Argument
- 2.19. Type Inference
- 2.20. Type Variable
- 2.21. Union Types
Algorithms & Functions
An algorithm is a set of instructions to solve specific problems or to perform a computation. In TypeScript algorithms can be implemented with functions.
Algorithms have defining characteristics such as:
- Generality: The algorithm must apply to a set of defined inputs.
- Definiteness / Uniqueness: At any point in time, there is at most one possibility of continuation. Each instruction step is well-defined.
- Deterministic: Given a particular input, the algorithm will always produce the same output.
- Effectiveness: The algorithm terminates in a finite amount of time / steps. Every execution of the algorithm delivers a result.
- Finiteness: The algorithm is described in a finite amount of steps.
- Feasibility: It should be feasible to execute the algorithm with the available resources.
Declarative programming is a programming paradigm that defines the desired state of an application without explicitly listing statements that must be executed. In a nutshell, declarative programming defines an application’s execution from finish to start.
- Change UI behaviour by defining how it should look like based on its props
Imperative programming is a programming paradigm that uses statements to change an application’s state. In a nutshell, imperative programming defines a set of instructions from start to finish.
- Change UI behaviour by defining how to turn something off before turning it on
A function declaration gets hoisted and is written in the following way:
A function expression is part of an assignment and does not get hoisted:
Referentially Opaque Expressions
An expression is referentially transparent when it cannot be replaced with its return value.
At the time of writing the execution of
'2021-09-22T12:45:25.657Z'. This result will change over time, so we cannot replace
const isoDate = today() with
const isoDate = '2021-09-22T12:45:25.657Z' which makes this expression referentially opaque.
Referentially Transparent Expressions
An expression is referentially transparent when it can be replaced with its return value.
sayHello function always returns the same text, so we can safely replace our expression with
const message = Hello! which makes it referentially transparent.
let keyword can be used. It makes variables inaccessible from the outside of their blocks:
A deterministic function will always produce the same output given the same input. This makes the output of a deterministic function predictable as it does not rely on a dynamic state.
An identity function returns the identical value that was given to it:
Pure functions are a subset of deterministic functions. A pure function always produces the same result given a particular input. In addition, it does not cause side effects by avoiding I/O operations like printing to the console or writing to the disk.
A pure function does not mutate its passed parameters and is referentially transparent.
By default, the TypeScript compiler does not know in which runtime environment (for instance Node.js v16, Electron v16, Chrome v94) our code will be executed later. That’s why we can help the compiler knowing that by defining an ambience / ambient context.
Example: If you run your code in an environment where there is a “world” object that TypeScript does not know about, you can define that context using
A collective type unites all literals of its kind. Collective types can be used to widen a type.
Example: The type
number is a collective type of all integers as it includes different integer values:
Compilers transform high-level programming language code to low-level machine code or some low-level intermediate representation (e.g. Bytecode in Java).
Discriminated Unions (or tagged union types) are a form of type guards where the type is narrowed based on a shared property.
In the following example, the type of
Person have a shared property called
type. Depending on the value of this property, TypeScript can narrow down the type from
Dog | Person to either
Because this allows the type to be discriminated, this technique is referred to as discriminating unions. This concept also exists in F# (Discriminated Unions in F#).
A literal describes a specification of a collective type. A literal can be an exact number or an exact string instead of being any number or being any string:
In TypeScript a module is a file with at least one top level
All non-primitive data types (also composite data types) are passed by reference:
- object (and array, which is also an object)
All primitive data types are passed by value:
TypeScript’s type system additional has the following primitives:
String Literal Type
A string literal is a specific string (e.g. “click”) and a string literal type is the type describing a specific string:
Template Literal Type
A template literal type is a combination of Template literal and string literal type. It is capable of resolving string interpolations within a template string to use it for strong typing:
A tuple can contain values of different data types and works like an array with a fixed number of elements:
Note: A tuple (e.g.
[number]) is different to an array (e.g.
A type annotation is when you explicitly define the type that a variable can take on:
Type Argument Inference
TypeScript can infer types but also type arguments. This usually happens when the arguments of your generic functions are connected to the generic type of your function. In this case TypeScript can infer the type of your type variable by inspecting the input values of your function call:
In the following code,
<string> is the type argument (sometimes called type parameter):
When there is no explicit type annotation then TypeScript will infer the type for you:
A type variable is the placeholder for a generic type in your generic code:
Type variables are written by using the angle brackets and defining a name for the variable (e.g.
<T>). This construct is often referred to as the diamond operator because the angle brackets look like a diamond (
A union type describes a collection of types (i.e.
It is called a union because it unites the amount of possible types. The term union comes from set theory where it is used when two (or more) sets are combined.