SqlGenerator | 少量的数据存储 。 这个时候其实没有必要去使用第三方的orm库

The question can be considered as closed as the related feature request is registered in the issue tracker.

@Embedded means that while the referenced object is a separate entity in the java side of things it is part of the table mapped to the parent entity.

You therefore need to include the columns of the ContractStatuses entity in the suppliers_contracts table.

This of course would result in a name clash, so you need to specify a prefix in the @Embedded annotation and use that prefix with the column names coming from ContractStatuses.

Note: Since this is a follow up question from https://stackoverflow.com/a/64025347/66686 The advice there was to use a database view, so instead of the suppliers_contracts table you would map SupplierContractReadOnly to a database view which then contains all the columns discussed above.

Another note: @MappedCollection only applies to collection like fields (List, Set and Map). So it serves no purpose in this case and should be removed.

Could you share you application.properties or application.yaml in your resources folder?

I believe that your datasource isn't configured, could be the general one or the one in flyway.

I can imagine that you rely on env vars in that config but you didn't provided those on container startup.

The reason for your error is that you are returning an object from the method not a string. All you have to do is to override the ToString method of your class to return the string you want.

You can not connect to DB container using localhost, here localhost mean the Nodejs container not the DB container.

I didn't test, but according to this

The Column annotation and XML element defines insertable and updatable options. These allow for this column, or foreign key field to be omitted from the SQL INSERT or UPDATE statement. These can be used if constraints on the table prevent insert or update operations. They can also be used if multiple attributes map to the same database column, such as with a foreign key field through a ManyToOne and Id or Basic mapping. Setting both insertable and updatable to false, effectively mark the attribute as read-only.

You wrote it the wrong way round? The ReadOnly flag is true and you are hoping to avoid looping and set it as false.

Were you to provide a simple select, the query builder would be able to create matching update and delete queries, map columns to parameters and a writable dataset would be generated. It can't do that for complexities inherent in a union query because it would take some serious amount of coding logic to work out whether the columns are updatable, which table they're from, whether the same columns are presented in the same way etc.. it's not designed to be that sophisticated web mostlynits for consuming simple select queries

Just set the columns to be writable and provide an update query manually to the builder/adapter - trying to fudge things so that the process that creates the data table from the data can guess not to make these columns read only is going to be a considerable waste of time and probably quite fragile..

The other option you have is to not do your union, do two fills on two different queries from the adapter instead, it will have a better chance of generating an updatable dataset and command collection if the queries are free of joins, union, groups etc

Or use the dataset designer and create a new dataset that is designed and strongly typed to represent what you need

Of all the options, I'd pick the last you're trying to do this "properly" - data containers/classes should be designed and strongly typed; using a generic dataset isn't much "better" than storing everything in an object[][][]

It looks like you are trying to rewrite a catch-all query as LINQ. You don't need to write a catch-all query at all with LINQ, you can simply ignore the condition if you don't want it, eg:

This might be a error with liquibase drivers for SQLite.

https://liquibase.jira.com/browse/CORE-2468

If that's the case, you might want to switch to another DB.

Alternatively, you can try to use raw SQL in the changeset instead (based on your github code):