HHH-10668 - Add SQL dialect for SQLite 3

[vladmihalcea]

https://hibernate.atlassian.net/browse/HHH-10668

[vladmihalcea]

@luca-vercelli I fixed the tests in the documentation folder. However, there are several issues that prevent us for integrating it:

1. Foreign Keys don't work at all. Traditionally, we added FKs using an ALTER TABLE command, but that's not working with SQLite.
2. There is no way we can spawn two EntityManager instances and operate with 2 JDBC Connections concomitantly. This affects all TABLE generator tests and also Lock-related tests.

Until a fix is proposed at least for the FK issue, we should postpone this integration.

[luca-vercelli]

Yes, I know that ALTER TABLE ADD FK is not supported. Foreign keys must be declared during table definition, see https://sqlite.org/foreignkeys.html.
I can work on this. Can you tell me which parts of code I should look at? Where can I find tests for you issues 1. and 2.?
Thank you.

A possible solution for 1. is not to declare at all foreign keys.

[vladmihalcea]

@luca-vercelli Take a look on org.hibernate.tool.schema.internal.StandardTableExporter and StandardForeignKeyExporter which are now used to create the tables and the FK during hbm2ddl.

The Dialect already provides a way to customize them via the:

public Exporter<Table> getTableExporter() {
	return tableExporter;
}

public Exporter<ForeignKey> getForeignKeyExporter() {
	return foreignKeyExporter;
}

methods. We need to a way to export the FK in the Table exporter and just have a no-op getForeignKeyExporter.

For 2. We just need to exclude the tests that are causing failures, like LockModeTest. Check out this PR to know how you can do it. Look for @SkipForDialect annotation. You can work on top of this PR, and send a new PR when you are done.

[luca-vercelli]

I have big problems here... I cannot even compile hibernate anymore. The JVM hangs after a random number of tests.

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00007f91ba2acd73, pid=15684, tid=0x00007f91f0efe700
#
# JRE version: OpenJDK Runtime Environment (8.0_121-b13) (build 1.8.0_121-8u121-b13-0ubuntu1.16.04.2-b13)
# Java VM: OpenJDK 64-Bit Server VM (25.121-b13 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C  [sqlite-3.16.1-770159f8-1443-427b-bca0-e3aeccc14285-libsqlitejdbc.so+0x79d73]
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# An error report file with more information is saved as:
# /home/luca/Documenti/src/hibernate-orm/hibernate-core/hs_err_pid15684.log
22:55:54,230  WARN CachingRegionFactory:52 - CachingRegionFactory should be only used for testing.
#
# If you would like to submit a bug report, please visit:
#   http://bugreport.java.com/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

1670 tests completed, 39 failed, 162 skipped
:hibernate-core:test FAILED

I don't know what is changed, when I sent my previous PR I had not so many tests. there were 369 tests and they worked.

[vladmihalcea]

Maybe it's an OpenJDK bug. I'm using Oracle JDK and it's working just fine.

[luca-vercelli]

Same problem with Oracle JDK.
I'm building on Linux Mint (i.e. Ubuntu), on AMD64 architecture.
It seems to be a bug in sqlite JDBC (it creates a library in /tmp/sqlite-xxx-libsqlitejdbc.so ).

[cristianrgreco]

Would love to see official SQLite support in Hibernate

[cybermerlin]

hey, guys, how much time need to allow this merge request?

[beikov]

We are discussing the way forward for new Dialects in #3783

When we reach a conclusion, we might pick up this Dialect again.