Add schema name option to database introspection methods

[barnettben]

Adds an optional schemaName parameter to table introspection methods as discussed in #1463.

I have added the extra parameter to foreignKeyViolations() since that does support checking in other schemas - it then follows to also update checkForeignKeys(). I'm not totally happy with the doc comments on checkForeignKeys(), but I think they makes sense.

Commits in this PR are mostly grouped by which function they're updating, so are hopefully easy to review.

I have not updated the sample code in the README since the code in there still works and doesn't really need to show this optional addition.

One problem is that make smokeTest persistently fails on DatabaseCursorTests.testFastDatabaseValueCursorStep(). This is the first profiling test to run and takes about 3x longer than others in the same file. It appears unrelated to any of the code that I've touched with this PR. Other tests all pass.

Any and all critiques welcome!

---

Pull Request Checklist

• CONTRIBUTING: You have read https://github.com/groue/GRDB.swift/blob/master/CONTRIBUTING.md
• BRANCH: This pull request is submitted against the development branch.
• DOCUMENTATION: Inline documentation has been updated.
• DOCUMENTATION: README.md or another dedicated guide has been updated.
• TESTS: Changes are tested.
• TESTS: The make smokeTest terminal command runs without failure.

[groue]

Thank you @barnettben, it looks like a perfect PR :-)

I have not updated the sample code in the README since the code in there still works and doesn't really need to show this optional addition.

👍

One problem is that make smokeTest persistently fails on DatabaseCursorTests.testFastDatabaseValueCursorStep(). This is the first profiling test to run and takes about 3x longer than others in the same file. It appears unrelated to any of the code that I've touched with this PR. Other tests all pass.

Yes, it is unrelated, so it's ok to ignore it here. Do you know how it fails?

---

I'll help fixing the CI builds. Please stay tuned.

[barnettben]

Thank you @groue.
Looks like I mis-read which test was failing. It is actually testNullableDatabaseValueCursorForEach().

The log shows:

Thrown Error at SerializedDatabase.swift:125:failed: caught error: "SQLite error 10: disk I/O error - while executing `PRAGMA query_only = 0`" (0.00s)

I have not looked in to it further yet. Hopefully it's not the symptom of a failing SSD!

[groue]

I'm a little busy these days, but I'm in vacations at the end of the week. I should be able to look at the PR, then.

[barnettben]

Thank you, I'm not in a hurry. Don't spoil your holiday!

[groue]

Hi @barnettben,

I have skipped tests that involve attached databases when they run with SQLCipher + encryption. Another option would be to provide the password to the ATTACH statement. But this would just make the test more complicated without much testing benefit. SQLCipher is still tested (when not encrypted).

Another CI problem was this error message:

Run kishikawakatsumi/xcresulttool@v1
Error: Resource not accessible by integration

I have tried the techniques described in https://github.com/orgs/community/discussions/60820. Now I wait to see if CI is happier 😅

EDIT: nope, I could not solve the "Resource not accessible by integration" error. I just disabled kishikawakatsumi/xcresulttool, which wasn't very useful at reporting CI test failures anyway.

[groue]

All CI tests pass 🎉 Merging

[groue]

Shipped in v6.24.0 👍

Thank you very much @barnettben. Would you accept an invitation that grants you a push access to GRDB? I do intend to keep the lead of the project, but it is good for everyone that people of various levels of expertise can improve the repository. Of course, your free time would remain just as free as it was before!

[barnettben]

Thank you @groue, I'd be happy to accept. And thanks for merging this in too.

[groue]

You're very welcome, @barnettben 🙂🤝