Store API

This is a NodeJS web-store api appliaction that provides a variety of database Search, Sort and Filter options for web-store users.

The setup used here allows the Frontend to only send HTTP calls to request validated data from the backend.

Postman is used to test the endpoints, no frontend is available for this project.

Try it now online

To try the backend application, please visit the following URL to see all products in the database:

• https://webstore-api.azurewebsites.net

It is recommended to install a JSON Web Formatter such as json-Formatter to see the json data correctly.

NOTE:

To fully test the built features and utilise this backend-app, it is recommneded to use Postman. Point to the following REST API endpoint:

• https://webstore-api.azurewebsites.net/api/v1/products

See 'Valid Query Params' section below.

Kindly allow for 30 sec for database Rehydration and loading.

Running Locally: Prerequisites

Please make sure you have Node.js installed on your machine and Git version control system to run project locally.

Cloning the Repository

To clone this repository to your local machine, cd to your desired directory (e.g cd Desktop).

Then run the following git clone command. This will create a folder called Store-api on your specified directory.

git clone https://github.com/RhonnieAl/Store-api

Project Setup

Run npm install to install node dependencies

Run npm start or node app.jsto kickstart server.

In order to run the project, create a .env file, set your "MONGO_URI" with a correct "connection string" to your MongoDB.

Populate the Database with the provided initial data by running :

node populate.js

In order to avoid port collisions, port 3000 is used, feel free to change it.

Great :) Happy testing!

REST API Endpoints

• /api/v1/products - Run all query params on this endpoint

Valid Query Params

name - Search products by a specific name (search part of the name)

company - Search products by company brand (Ikea, JYSK, Isku, Masku)

featured - Search products on featured ('true') or not featured ('false)

sort - Determin which property the returned products should be sorted by (price, name, rating, date, company, featured)

fields - Select only the product properties you wish to view (price, name, rating, date, company, featured)

limit - Determin how many products you want to retrieve

page - Select the page number you wish to view

numericFilters - Search and Filter catalogue by "price" and "rating"

Error Handling

Errors are caught globally using middleware error-handler.js with the help of library express-async-errors for catching Async Errors. The library throws any async error without having to pass it to the next middleware as native Express does.

Instead of writing an Async-wrapper function with try-catch and next(), we throw the error with the help of express-async-errors to our custom error handler in error-handler.js and there do anything, for example show it on the console or return a message on the browser.

DB

MongoDB was auto populated with initial data using a Script in the file populate.js.

const start = async () => {
  try {
    await connectDB(process.env.MONGO_URI);
    // Delete any data that is currently in DB (optional)
    await Product.deleteMany();
    // Create and populate DB with initial data from jsonProducts.js
    await Product.create(jsonProducts); // Where the magic happens
    console.log("Sucess!!");
    // Once done, Close connection.
    process.exit(0); // Passing 0 means success
  } catch (error) {
    console.log(error);
    process.exit(1); // Passing 1 means error
  }
};
start();

DB Search Functionality

In Mongoose V5 an empty array is returned for queries that do not match the accepted Schema. To account for this, destructured needed queries

controllers >> products.js

const { featured, company, name } = req.query;

and passed an empty object into .find() so as not to return epmty array to the user.

Mongoose V6 takes care of this nateively by ignoreing erroneous queries.

To allow users to serach part of name string instead of typing full product name, MongoDB Regex query operators were used.

// regex: case incensitive and allows partial name input
{ $regex: name, $options: "i" }
{ $regex: company, $options: "i" }

DB Sort Functionality

This functionality enables the user to sort data in by any order according name, date, price, featured, company and rating.

By default MongoDB documents in a collection are sorted by date created.

To apply the .sort() functionality, the method is cahined as shown on Mongoose Docs.

As a rule .sort() method is applied on an Object query and not the resolved result of a query i.e

• Using a method on the Model with await returns the resolved result of a query.

• Using a method on the Model without await returns an object (a Query object), not the result of a query.

const { item, sort } = req.query;
result = Product.find(item);
if (sort) {
  sortList = sort.split(",").join(" ");
  result = result.sort(sortList);
} else {
  result = result.sort("createdAt");
}
const products = await result;
res.status(200).json({ products });

DB Select Functionality

This functionality enables a user to select certain fields such as name and price. Returning only data with name and price.

According to Mongoose Docs, the method is .select()

Similarly to sort functionality:

// Select
if (fields) {
  fieldList = fields.split(",").join(" ");
  result = result.select(fieldList);
}

const products = await result;

DB Pagination Functionality

This functionality allows the user to request a specified certian page if the number of items returned are numerous.

The user also has the ability to determin how many items a single page should have.

To achieve this the following methods were used: .limit() : determins how many results are returned

.select() : determins how many returned items are cut off. if you select(5) it will return from 6 and over.

.skip() : specifies the number of documents to skip in the filtered collection according to Mongoose Skip docs

// Extract user-given page number or default to 1
const page = Number(req.query.page) || 1;
// Extract user-given num of items requested or default to 10
const limit = Number(req.query.limit) || 10;
// set number of documents in collection to be skipped
const skip = (page - 1) * limit;

// If skip is zero page num begins from page 1 and returns items by limit given
result = result.skip(skip).limit(limit);

const products = await result;

DB Search by Price and Rating Functionality

This functionality allows the user to search and filter catalogue based on price or rating or both.

According to MongoDB docs, regex such as gt and gte are the only accepted comparison operators, instead of basic comparison operators (eg: <, >, =<, >=).

if (numericFilters) {
  // Object to map all user input operators to regex ones
  const operatorMap = {
    ">": "$gt",
    ">=": "$gte",
    "=": "$eq",
    "<": "$lt",
    "<=": "$lte",
  };
  // Find all occurences of user input operators
  const regEx = /\b(<|>|>=|=|<|<=)\b/g;

  // Replace all occurences in "numericFilters" with regex ones
  // "filters" is string with swapped operators
  let filters = numericFilters.replace(
    regEx,
    (match) => `-${operatorMap[match]}-`
  );

  const options = ["price", "rating"];
  // Split up the numeric queries passed by the user
  filters = filters.split(",").forEach((queryItem) => {
    // Destructure and name
    const [field, operator, value] = queryItem.split("-");
    // Restricing DB fields this fuction applies to
    if (options.includes(field)) {
      queryObject[field] = { [operator]: Number(value) };
    }
  });
}

Hobby Project maintined by Rhonnie Allan