refactor(omdb)!: remove meta:migrate goal + Java migration-conformance scenarios

Removes the consumers of the diff-and-converge schema-migration engine so the
engine itself can be deleted atomically in a follow-up commit. Schema migrations
(and live-DB schema-drift verification) move to the TypeScript toolchain
(@metaobjectsdev/cli migrate); the Java port retains runtime persistence only.

Maven plugin:
- Delete MetaDataMigrateMojo (the meta:migrate goal) and its two tests.
- Delete MetaDataVerifyMojo. The meta:verify goal was a thin subclass of the
  migrate mojo providing a live-DB schema-drift gate (SchemaMigrationEngine.verify),
  so it is part of the migration subsystem, not an independent codegen/metadata
  check. Prompt/template drift is still checked via the metaobjects-render Verify
  API; meta:gen and meta:editor are unaffected.
- Update server/java/README to list meta:gen / meta:editor only.

integration-tests:
- Delete MigrationScenarioRunner + MigrationScenarioTests; drop the migration
  record types from Scenarios and the migration parsing from ScenarioLoader.
  Java permanently stops running migration-conformance scenarios.
- Re-point QueryScenarioRunner's schema bootstrap off SchemaMigrationEngine onto
  the retained runtime auto-create path (MetaClassDBValidatorService + the
  drivers' legacy createTable/createIndex/createForeignKey/createSequence DDL).
- Defer the projection-aggregate query scenario: its aggregate-view body was
  synthesized by the engine's ViewBodyBuilder (removed); runtime auto-create only
  materializes views from an explicit dbViewSQL attr.

omdb:
- Quote column/PK/index/FK identifiers in PostgresDriver's legacy createTable/
  createIndex/createForeignKey so the runtime auto-create path produces a
  mixed-case-correct Postgres schema, matching the already-quoted CRUD/SELECT
  path. Previously masked because only the engine built the Postgres conformance
  schema; the legacy DDL was only exercised against lowercase Derby fixtures.

The migrate/ package still compiles (now unused) and is deleted in the next commit.

Verified: om/omdb/maven-plugin/core-spring/omdb-ktx green; integration-tests green
under Docker (query 9 pass + 1 deferred, api-contract 20, float 4).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: dmealing

server/java/README.md

@@ -33,7 +33,7 @@ All published to Maven Central under `com.metaobjects:*` at `7.1.0`:
 | `metaobjects-omdb` | Relational implementation of ObjectManager over JDBC + Spring-tx |
 | `metaobjects-omdb-ktx` | Kotlin facade over OMDB |
 | `metaobjects-core-spring` | Spring auto-configuration + `MetaDataService` |
-| `metaobjects-maven-plugin` | `mvn meta:gen` / `meta:migrate` / `meta:verify` |
+| `metaobjects-maven-plugin` | `mvn meta:gen` / `meta:editor` |
 
 The `archetype` and `examples` directories were removed in 7.1.0 (they had been out of the reactor since 7.0.0 and were not deployed to Central).
 
@@ -62,7 +62,7 @@ Spring REST + JPA stack:
 </dependency>
 ```
 
-Maven plugin for `meta:gen` / `meta:migrate` / `meta:verify`:
+Maven plugin for `meta:gen` / `meta:editor` (schema migrations and live-DB schema-drift verification are managed by the TypeScript toolchain — `@metaobjectsdev/cli migrate`; prompt/template drift is checked via the `metaobjects-render` `Verify` API):
 
 ```xml
 <plugin>

server/java/integration-tests/src/test/java/com/metaobjects/integration/MigrationScenarioRunner.java

@@ -6,9 +6,7 @@
 import com.metaobjects.manager.ObjectConnection;
 import com.metaobjects.manager.db.ObjectManagerDB;
 import com.metaobjects.manager.db.driver.PostgresDriver;
-import com.metaobjects.manager.db.migrate.AllowOptions;
-import com.metaobjects.manager.db.migrate.EmitResult;
-import com.metaobjects.manager.db.migrate.SchemaMigrationEngine;
+import com.metaobjects.manager.db.validator.MetaClassDBValidatorService;
 import com.metaobjects.object.MetaObject;
 import com.metaobjects.registry.MetaDataLoaderRegistry;
 import com.metaobjects.registry.ServiceRegistryFactory;
@@ -28,9 +26,12 @@
 /**
  * Mirrors C# QueryScenarioRunner / TS query-scenario.ts. End-to-end:
  *
- *   1. Apply the canonical schema (engine full-CREATE via
- *      {@link SchemaMigrationEngine#emit(Connection, AllowOptions)} against
- *      an empty actual snapshot).
+ *   1. Apply the canonical schema via the runtime auto-create path
+ *      ({@link MetaClassDBValidatorService} with {@code autoCreate=true},
+ *      which issues the drivers' legacy {@code createTable}/{@code createIndex}/
+ *      {@code createForeignKey}/{@code createSequence} DDL). Schema migrations
+ *      (diff-and-converge) are owned by the TS toolchain; the Java port only
+ *      retains runtime auto-create for test/dev schema bootstrap.
  *   2. Execute the scenario's seed-data SQL.
  *   3. For each {@link QuerySpec}: translate via {@link ObjectManagerDbAdapter}
  *      → {@link ObjectManagerDB#getObjects} / {@code getObjectsCount},
@@ -48,13 +49,14 @@ public static void run(QueryScenario scenario, PostgresContainer pg, Path canoni
         registry.registerLoader(loader);
 
         ObjectManagerDB omdb = newOmdb(pg);
-        SchemaMigrationEngine engine = new SchemaMigrationEngine(omdb, registry);
 
-        // 1. Apply canonical schema.
-        try (Connection c = openConnection(pg)) {
-            EmitResult emit = engine.emit(c, new AllowOptions());
-            executeSql(c, emit.up());
-        }
+        // 1. Apply canonical schema via the runtime auto-create path (the
+        //    drivers' legacy createTable/createIndex/createForeignKey DDL).
+        MetaClassDBValidatorService validator = new MetaClassDBValidatorService();
+        validator.setObjectManager(omdb);
+        validator.setAutoCreate(true);
+        validator.setMetaDataLoaderRegistry(registry);
+        validator.init();
 
         // 2. Seed data.
         if (scenario.seedData() != null && !scenario.seedData().isBlank()) {