SQL查询选择返回结果不在一个表中

Question

Say I have a table people with a name field. In that table are 3 rows with names: Steve, Mark, Ricky

Then I have a list of names: [Shane, Ricky, Mark]

I want to execute one query that will return the values that aren't already in the table.

I know I can do this with the following:

names = [Shane, Ricky, Mark]

for name in names:
    sql = '''SELECT * FROM people WHERE name=%s'''
    db.execute(sql, args=name)
    results = db.fetchall()
    
    # If no results are returned for a name 
    # this means the name isn't already in the table.
    if not results:
        print(name) # This would return 'Shane' as it's not already in the table.

Is there a way I could do this more efficiently with one SQL statement, rather than looping over each name in the list?

Answer 1

您可以使用： SELECT * FROM people WHERE name NOT IN(...)

Answer 2

You can use the SQL IN operator, selecting the values already in the table;

SELECT * FROM people WHERE name IN (...your values...)

and then finding the difference between the two lists:

not_in_db = list(set(names).difference(results))

Answer 3

UNTESTED:

Generate the list as a table (t1) and then return from that generated table, those values not in your people table.

The issue is a database can't return values that don't exist. So we first create a temporary table with your values so they exist and can be returned. Then we rub it up against the people table to find those that don't exist; thus allowing us to return those names not in people table.

SELECT *
FROM (values row('Shane'),
             row('Ricky'),
             row('Mark')) t1 (Name)
WHERE NOT EXISTS (SELECT 1 FROM people P where P.Name = T1.Name)

or perhaps: (I don't think you need to specify a table in mySQL if I recall.)

SELECT *
FROM (SELECT 'Shane' as Name UNION ALL
      SELECT 'Ricky' UNION ALL
      SELECT 'Mark') t1
WHERE NOT EXISTS (SELECT 1 FROM people P where P.Name = T1.Name)

So the above is the SQL for your sql = statement

Though in both cases we should specify data type for Name be it varchar, nvarchar and length...