test: harden clear() namespace isolation, tag invalidation and deleteItem() return value

[GromNaN]

Three small additions to the contract, found while implementing a PSR-6 adapter for MongoDB (symfony/symfony#65468). In each case the current tests pass on an implementation that is actually wrong.

clear() must not touch another namespace on the same storage

When several pools share a single storage, a table or a collection for instance, clearing one pool must only remove its own keys. Nothing checks that today, because the tests only ever use one pool.

CachePoolTest gains an optional createCachePoolWithOtherNamespace(), which returns a second pool on the same storage under another namespace. It returns null by default, and the new testClearDoesNotAffectOtherNamespaces is then skipped, so existing implementations are unaffected. When a pool is given, the test writes the same key in both, clears the first one, and checks that the second still holds its own value.

testInvalidateTags now uses disjoint tags

key was tagged with tag1 and tag2 while key2 was tagged with tag1, so invalidating tag1 alone was enough to pass. Each item now carries a single distinct tag, and a third item, tagged with none of the invalidated tags, checks that the invalidation does not reach too far.

deleteItem() return value on a tagged item

testRemoveTagWhenItemIsRemoved ignored what deleteItem() returned. It is now asserted to be true.

[cryptiklemur]

Nice find on all three. Few things before I merge.

createCachePoolWithOtherNamespace() returns a second FilesystemAdapter, and FilesystemCommonTrait::init() does $directory .= \DIRECTORY_SEPARATOR.$namespace. Two separate directories, so the bug class this is aimed at, prefix collisions on one substrate, structurally can't happen there. Two PdoAdapters over one sqlite connection would bite: doClear() is DELETE ... WHERE id LIKE 'ns%', so the namespace is a real prefix in a shared table. That means moving createCachePool() too though. If you'd rather not, just say in the docblock that it's a smoke test.

Small one: $other->clear() on the last line only runs if every assertion above it passed. try/finally.

On testInvalidateTags, can you keep key tagged ['tag1', 'tag2'] alongside the new key2 and key3? It was the only spot in the suite where a single invalidateTags() call matched one item through two of the passed tags.