copy table_name from 's3://bucket/path/' CREDENTIALS 
'aws_access_key_id=[access_key];aws_secret_access_key=[secret_key]' 
delimiter '\t' gzip maxerror 100;

When importing csv data from s3 to redshift, We can use delimiter ',' or CSV option. There two options are different. If we use CSV option, we can import such as this file

123,aaa,bbb,"223,54234",111
345,aaa,bbb,777,666

If we use delimiter ',' escape option, we can import such as this file

1,aaa
2,bbb\,bbb
3,ccc
4,ddd