有没有办法让NHibernate将字符串键视为不区分大小写？

Question

I'm using NHibernate's second level cache on an entity whos primary key is a string column. The database is case insensitive however when I retrieve the entity using the same Id but with different casing then NHibernate is treating the entity as a different entity in the cache. For example, I have a Setting table with a Name column as the primary key. If I try to retrieve the entity by using session.Get<Setting>("TestMode") , Get<Setting>("testmode") or Get<Setting>("TESTMODE") that will go to the database 3 times and put 3 separate entities in the 2nd level cache. If I then change something on my Setting entity through NHibernate, it will update the cached entity that actually matches the casing on the entity as it is stored in the database (in this case the one named "TestMode"). The problem is then that the other entities will still be in the cache but will be stale.

I understand why NHibernate would treat Ids as case-sensitive by default but there doesn't appear to be a way to set the case insensitivity at an entity level or configuration level as far as I can find. Can this be done, and if so how?

If it helps, I am using Fluent NHibernate. My Id column is specified as such:

Id(x => x.Name).GeneratedBy.Assigned().Column("Name");

UPDATE Included below are the implementations of Equals() and GetHashCode() from my Setting entity.

public override bool Equals(object obj)
{
    if (obj == null)
        return false;
    if (obj is Setting)
    {
        if (object.ReferenceEquals(this, obj))
            return true;
        if (this.Name.Equals(((Setting)obj).Name, StringComparison.OrdinalIgnoreCase)) // Ignore case because our databases are case insensitive!
            return true;
    }
    return false;
}

public override int GetHashCode()
{
    return Name.ToUpperInvariant().GetHashCode();
}

UPDATE 2 I've been profiling with NHibernate Profiler, and doing the retrievals for "TestMode", "testmode", "TestMode", "Testmode", "TESTMODE" (in that order) shows only a cache hit for the second "TestMode". It shows "2nd level cache load Setting (TestMode /* id */)", all the rest show SELECT ... FROM Setting setting0_ WHERE setting0_.Name = "testmode" (in whatever casing the call was asking for). When clicking the link "See the 1 row(s) resulting from this statement" it show me the row with the Name column having the correct casing of "TestMode", so it's clearly retrieving the entity with the right casing for the column itself, but it seems it's definately associating the entity with the id that the Get() function was called with. This doesn't seem like it should be the normal behaviour but I haven't been able to find any information about the specifics of how the 2nd level cache works, only the basics of how to use it from the NHibernate Cookbook.

Although I doubt it should make a difference to, I am using the HashtableCacheProvider for all these tests. (for production we'll be using a paid-for caching provider)

Answer 1

If you think of the cache as a dictionary with a string key, it's obvious that the keys "TEST" and "test" are different because C# is case sensitive.

But something doesn't seem right here. How do you know that NH is putting three entities in 2nd level cache? Assuming you're using a case-insensitive database, all three queries generated by the Get method will return the same row and the ID property on the entity will be set using the value returned by the query, not by the value supplied to the Get method. The entities will be cached using the entity's ID also. Do all three Gets return instances with the ID in the same case ("TestMode")?

---

Update based on the edits to your question, if I understand you correctly:

• Get("TestMode") caches the object as expected and retrieves it from cache on subsequent requests for "TestMode"
• Get("testmode") or any other case goes back to the database

That's the behavior I would expect. The value supplied for id is used to attempt to locate the entity in the cache. If a match is found then the entity is retrieved from cache; if not the database is queried. After the entity is retrieved, NHibnernate attempts to add it to the cache using its id as a key, but it already exists so the cached value is updated, replaced or ignored.

The key point is that the id supplied to the query is used as a key to locate the entity in the cache, but the entity's id is used as the key when it is added to the cache.

Answer 2

The cache depends on the Equals() implementation provided.

So, modify your Equals() implementation for this class to something that will take into account case issues. Either compare with insensitive settings or just .ToUpper()/.ToLower() the ID's.

UPDATE: there seems to be a duplicate question at NHibernate issue with assigned string ID and different case chars . Try to use the accepted answer solution

Answer 3

I think you can get around this by defining a custom string type, say CaseInsensitiveStringType , and then

Id(x => x.Name)
    .GeneratedBy
    .Assigned()
    .Column("Name")
    .CustomType<CaseInsensitiveStringType>();