SELECT knowledge FROM sql_resources WHERE category='sql' AND slug='from'

FROM

The FROM clause is one of the fundamental building blocks of a SQL SELECT statement. In this article we'll explain how it works.

The FROM clause is formed of the reserved 'from' keyword followed by either:

Loading code...
  • A table name (most common)
  • A subquery
  • A JOIN clause (read more about the JOIN clause)

The sources defined in the FROM clause determine the set of rows that the SELECT statement can operate on, and so is the first part of the query to be executed.

In this article, we'll look at some examples of different FROM clauses.

Loading code...

In this article the code snippets are written in the Google BigQuery Standard SQL syntax.

FROM a table

The simplest FROM clause just lists a single table name - then the entire SELECT statement is operating on a single table.

stringsnumbers
a1
b2
c3

FROM a subquery

Instead of referring to a table by name, we can also refer to it as the result of a SELECT statement (since SELECT statements return tables). For example, we can rewrite the query above as

Loading code...
stringsnumbers
a1
b2
c3

or indeed as

stringsnumbers
a1
b2
c3

These subqueries can be as complex as you'd like, though for the sake of readability you should try to avoid putting too much logic into a subquery. In a Count notebook we encourage splitting complex queries into separate cells for this reason.

Including a JOIN clause

When you want to select rows from multiple tables, you'll need to tell the database how to merge the tables together (read more here). These instructions take the form of one or more JOIN clauses after the initial FROM clause. For example:

Loading code...
numbersnumbers_2
14
25
36

In the query above, our SELECT statement says that:

Loading code...
  • We want the numbers and numbers_2 columns (the database is smart enough to know which tables those columns come from)
  • FROMtable_1 and table_2
  • The rows of table_2 should be associated with those from table_1 by comparing the columns strings and strings_2