SchemaBuilder

A table defines what data a database stores. Schema creates or modifies that structure; Blueprint describes its columns within a closure. There are adapters for SQLite, MySQL, PostgreSQL and SQL Server, with dialect differences. Schema applies PREFIX/DB_PREFIX automatically.

Index · Before: models · After: migrations

The example requires connection and an owners table that supports the composite key. It is a schematic fragment, not a stand-alone program.

Schema::create("products", func(Blueprint $table) {
    $table->id()
    $table->string("sku", 50)->unique()
    $table->decimal("price", 10, 2)->default(0)
    $table->unsignedBigInteger("tenant_id")
    $table->unsignedBigInteger("owner_id")
    $table->unique(["tenant_id", "sku"])
    $table->foreign(["tenant_id", "owner_id"])
        ->references(["tenant_id", "id"])
        ->on("owners")
        ->onDelete("cascade")
    $table->timestamps()
})

Schema

• create($table, func(Blueprint $blueprint))
• table($table, func(Blueprint $blueprint))
• rename($from, $to)
• drop($table) and dropIfExists($table)
• hasTable($table) and hasColumn($table, $column)

##Blueprint

Types: id, increments, integer, tinyInteger, smallInteger, mediumInteger, bigInteger, unsignedInteger, unsignedBigInteger, float, double, decimal, char, string, text, mediumText, longText, date, dateTime, time, timestamp, timestamps, softDeletes, boolean, json and enum.

Last column modifiers: nullable, unsigned, unique(), default and comment. unsigned and the inline SQL comment are MySQL properties; the other engines preserve the portable type without inventing an equivalent semantics.

Table commands:

• dropColumn($column) or dropColumn([$a, $b])
• renameColumn($from, $to)
• index($columns, $name=nil)
• unique($columns, $name=nil) or uniqueIndex(...)
• dropIndex($name)
• foreign($columns, $name=nil)->references($columns)->on($table)->onDelete($action)->onUpdate($action)

SQLite rebuilds the table transactionally when a foreign key is added using Schema::table(), preserving data, indexes, and explicit triggers. PostgreSQL uses SERIAL/BIGSERIAL, JSONB and equivalent types. SQL Server uses IDENTITY(1,1) PRIMARY KEY, NVARCHAR(MAX), DATETIME2, and [...] delimiters.