Blog

In this post, we will look at listing users with all their addresses. Both users and addresses tables are joined with a pivot table, user_addresses, which holds the foreign key of each of users and address table primary key.

In this post, we will look at getting a list of countries along with its addresses.

There are three ways of loading this kind of relationship:

1. Perform one sql query on both tables
2. Perform a left join from countries to addresses table,
3. Use array agg (for postgres) function or group concat for mysql/mariadb.

In this post, we will look at how deleting a single record in a table works for these various libraries and ORMs.

In general, this operation is very simple because we only care about deleting and checking for an error.

In short, we want to achieve

DELETE FROM users where id=$1;

In this post we will see a simple update of a resource by using its ID. We will try to perform an UPDATE <TABLE> SET <COLUMN1>=<VALUE1> etc, if not, we simply retrieve the item first, set to new values, and then save the record.

We only show how to do a full PUT updating instead of a PATCH, where a patch means we only update certain fields only, leaving other columns untouched.

As a good practice, we scan user’s request to a custom UserUpdateRequest struct. Then we hand over this struct to our database access layer.

In this post, we will look at how these libraries and ORM deal with fetching a single record given an ID. There should not be any drama as we only be doing a simple query as follows:

 SELECT * FROM users WHERE id = $1; 

In this post, we will look at comparing listing of user resource. We limit the record to the first thirty records.

Basically in all of them we want to make the following query

SELECT * FROM users ORDER BY id LIMIT 30 OFFSET 0;

In this post, we compare and contrast how these libraries and ORMs handle a record insertion.

As a standard approach to all example in these blog series, our controller accepts and parses client request to a custom ‘request’ struct, and if required, parses query parameter(s). We hash the password and together with CreateUserRequest we try to insert to database. Then the struct is passed down to our data access layer.

type CreateUserRequest struct {
	ID         uint   `json:"id"`
	FirstName  string `json:"first_name"`
	MiddleName string `json:"middle_name"`
	LastName   string `json:"last_name"`
	Email      string `json:"email"`
	Password   string `json:"password"`
}

All CRUD operations are done in crud.go files of each of the libraries and ORMs directories.

You should perform validation in production. In our case, we skip this part to keep this blog series focused.