New answers tagged ddl
Not real sure but you could try to use the Linux version of mysqldump in a Cygwin or similar interface. It has a cleaner output file than the windows version.
One way to accomplish this, using Java, would be to prepare the query statement, then use the PreparedStatement.getMetaData() method to obtain the ResultSetMetaData object that has the list of columns that would be returned, as well as their data types. Other database interfaces, such as ODBC and OCI, have similar functionality.
If you talk about continuous integration then I assume it's a development environment. In that case I would say the person doing structural changes has to test them to ensure not to break things for others, much the same way someone updating a common library: Test in your own sandbox before commiting such changes. In a production deployment process you ...
The best way to enforce uniqueness in this scenario then is to not split the artist "types" into separate tables. Normalizing your schema will show that there are fields in common, so at the least your artists table would contain id, artist_type ("band" or "singer"), and name. Any fields that are not shared should be split into more generalized tables. ...
Top 50 recent answers are included