fisharebest/laravel-floats

Floating point support for Laravel migrations

1.5.0 2020-09-24 12:32 UTC

This package is auto-updated.

Last update: 2024-04-15 15:04:30 UTC


README

License: MIT Latest Stable Version Build Status Coverage Status Scrutinizer Code Quality

laravel-floats

This package allows database migrations in Laravel 5, Laravel 6 and Laravel 7 to create FLOAT columns in MySQL.

Huh? You mean that Laravel does not support floating point columns?

Sadly no. It only supports double-precision floating point columns. If you want single-precision floating point, you need to use DB::raw().

This is despite the Laravel documentation at https://laravel.com/docs/5.8/migrations which states:

$table->float('amount', 8, 2); FLOAT equivalent column with a precision (total digits) and scale (decimal digits).

$table->double('amount', 8, 2); DOUBLE equivalent column with a precision (total digits) and scale (decimal digits).

You can read all about it at laravel/framework#3151 and many other issues.

In your migration Laravel 5.0 - 7.0 With this package
$table->float('col'); DOUBLE(8,2) FLOAT
$table->float('col', 0); DOUBLE FLOAT
$table->float('col', 5); DOUBLE(5,2) FLOAT
$table->float('col', 5, 3); DOUBLE(5,3) FLOAT(5,3)
$table->double('col'); DOUBLE DOUBLE
$table->double('col', 0); DOUBLE DOUBLE
$table->double('col', 5); DOUBLE DOUBLE
$table->double('col', 5, 3); DOUBLE(5,3) DOUBLE(5,3)

Installation

composer require fisharebest/laravel-floats

Package discovery takes care of everything on Laravel 5.5 and later. If you're using Laravel 5.4 or earlier, you'll need to replace an alias in config/app.php.

'aliases' => [
    ...
    'Schema' => \Fisharebest\LaravelFloats\Schema::class,
    ....
]

NOTE: this assumes you are using Laravel to autoload your facades using the aliases.

If you explicitly import Laravel's schema builder using use Illuminate\Support\Facades\Schema; then you will need to change this to use Fisharebest\LaravelFloats\Schema;.

How does this package work?

We extend the MySQL Grammar, modify the blueprint for float(), and then bind the updated grammar back into the IoC container.