EssOEss posted:

My experience with Entity Framework is that it tends to get in trouble very fast once you start using fancy words like group by and join. I immediately ran into client-side evaluation or EF just giving up with some internal query builder exception when I last tried to use it for nontrivial cases. I only have positive experience using it as a simple row serializer/deserializer and suggest you expect no more from it.

Thus I do not know if EF can actually turn this into nice SQL but my first try at this would be something like:

code:

record Row(int Id, string Message);

class Program
{
    static void Main(string[] args)
    {
        var data = new Row[]
        {
            new(1, "apple"),
            new(2, "apple"),
            new(3, "apple"),
            new(4, "apple"),
            new(5, "banana"),
            new(6, "banana"),
            new(7, "dusty"),
            new(8, "elephant"),
            new(9, "elephant"),
            new(10, "hello"),
        };

        var groupedByMessage = data.GroupBy(row => row.Message);

        var messageAndMaxId = groupedByMessage.Select(group => new
        {
            Message = group.Key,
            MaxId = group.Max(row => row.Id)
        });

        var fiveMostRecentMessages = messageAndMaxId
            .OrderByDescending(x => x.MaxId)
            .Select(x => x.Message)
            .Take(5);

        Console.WriteLine(string.Join(", ", fiveMostRecentMessages));
    }
}

Supersonic posted:

I'm looking to upload a file to Amazon S3 in multiple parts using C# and either WebClient or HttpClient. The upload process is a 4-step operation:

1. Initiate a multipart upload
2. Create pre-signed URLs for each part
3. Upload the object’s parts
4. Complete multipart upload

epswing posted:

I believe Distinct for both Linq to Entities and Linq to Objects will not maintain ordering.

New Yorp New Yorp posted:

Although I'm not a huge fan of "throw another nuget package at it!" and I'm not sure if it works with IQueryable vs IEnumerable, MoreLINQ has a DistinctBy extension method that probably fits the bill here.

Personally I'd just do the groupby.

mistermojo posted:

yeah groupby should be what you want

code:

var tupleList = new List<(int id, string message)>
  {
    (1, "apple"),
    (2, "apple"),
    (3, "apple"),
    (4, "apple"),
    (5, "banana"),
    (6, "banana"),
    (7, "dusty"),
    (8, "elephant"),
    (9, "elephant"),
    (10, "hello")
 };

var last5distinct = tupleList.OrderByDescending(x => x.id).GroupBy(p => p.message).Select(y => y.First().message).Take(5);

Console.WriteLine(string.Join(Environment.NewLine, last5distinct));

gets

hello
elephant
dusty
banana
apple


.NET 6 finally adds DistinctBy so we won't even need to use another package soon!

e: if you know its always ordered by id you could simplify it with something like

code:

var last5distinct = tupleList.GroupBy(p => p.message).Select(y => y.First().message).Reverse().Take(5);

raminasi posted:

Is there a reason you're not just using the utility class AWS provides for this?

pointsofdata posted:

Re the EF question, just write some SQL (assuming you want the logic to happen on the db) rather than fighting with EF. Most SQL variants are really good at that sort of logic and it's easy to develop with interactive queries.

NecroBob posted:

code:

  SELECT TOP 5 [Message]
    FROM [dbo].[idk]
GROUP BY [Message]
ORDER BY MAX([Id]) DESC

EssOEss posted:

What is all y'all's experience with nullable reference types? Any great success stories to tell? I tried them out and just found it to be a nice idea that in practice is just a nuisance that did not help me at all. Wondering if this is a common experience or if there is a scenario where they prove worth bothering about.

EssOEss posted:

What is all y'all's experience with nullable reference types? Any great success stories to tell? I tried them out and just found it to be a nice idea that in practice is just a nuisance that did not help me at all. Wondering if this is a common experience or if there is a scenario where they prove worth bothering about.

LOOK I AM A TURTLE posted:

The only annoying thing I've encountered is when libraries I'm using don't support them yet, forcing me to temporarily disable them in a a few cases. This is the the case that comes to mind:

C# code:

// Defined in 3rd party library:
public interface IFooGetter
{
    Foo GetFoo(int fooId);
}

// In my code:
public class FooGetter : IFooGetter
{
    private readonly IDictionary<int, Foo> _foos;

    public Foo GetFoo(int fooId) => _foos.TryGetValue(fooId, out var result) ? result : null;
}

The compiler doesn't allow this. GetFoo can't return null because Foo isn't nullable, but you can't make the return type Foo?, because then you're not implementing the interface as defined. The workaround is to either return null!, or temporarily disable NNRTs with the #nullable disable pragma.

LOOK I AM A TURTLE posted:

The workaround is to either return null!

Chrungka posted:

Are you sure about example above? When I try to implement interface from library that does not support NNRTs, I have no problem implementing the interface method with Foo? return type. On the other hand, if library does support NNRTs, then compiler rightfully complains and correct implementation would be to throw exception instead of returning null.

NihilCredo posted:

I understand why this is necessary (if nothing else for the exact scenario you described) but lol that this is even allowed.

LOOK I AM A TURTLE posted:

Hmm, you may be right. Perhaps the scenario I'm describing actually only applies to cases where the library has <Nullable>enable</Nullable>, but the particular interface hasn't been updated to correctly reflect the true nullability of the return value? I had this problem specifically with the IUserStore<T> interface in the Identity part of ASP.NET Core, where the FindById/Name/Email methods are defined as returning T rather than T?. I could implement the interface as IUserStore<User?>, but then it ends up being nullable in lots of other methods where it isn't desirable (UpdateUser etc.).

LOOK I AM A TURTLE posted:

The first time I tried it I fully expected it to give some sort of "this expression is hosed" error, like when you do if (0 == 1), but nope.

EssOEss posted:

What is all y'all's experience with nullable reference types? Any great success stories to tell? I tried them out and just found it to be a nice idea that in practice is just a nuisance that did not help me at all. Wondering if this is a common experience or if there is a scenario where they prove worth bothering about.

Chrungka posted:

I think you should go with IUserStore<User?>. Having to implement Task<IdentityResult> UpdateAsync(User? user, CancellationToken cancellationToken) with regular null check via ArgumentNullException might be inconvenient. But returning null when contract explicitly says it doesn't return null is IMHO a bug in code.

Hammerite posted:

snip

This has been working up to now under Oracle, using Oracle.ManagedDataAccess.Client. If I don't find a way to get it working I will experiment with explicitly constructing the database connection using the types from Npgsql, but I was hoping that at least in this aspect it'd prove to be a drop-in replacement...

Mr Shiny Pants posted:

Hammerite posted:

I've got it working now. The problem was something much dumber - I was putting that db provider stuff in the app.config of the library that communicates with the db, and it needed to be in the app.config of the C++ app that uses the library. Now I'm on to the next problem.

Hammerite posted:

I'm going to ask this here too because it doesn't look likely that Stack Overflow is going to be much help.

How do I get C# to print out the exact value represented by a floating-point number in memory?

I'm looking to create a function f such that this code:

code:

float myFloat = 0.1;
string s = f(myFloat);
Console.WriteLine(s);

prints 0.100000001490116119384765625

... which is, if I got my maths right, the actual value represented by the bit pattern of the float.

myFloat.ToString("F50") is no help. It just prints out 0.10000...000, which is

mystes posted:

Try G50 instead.

Hammerite posted:

drat, it does the right thing on net core 3.1 as well. Just Framework where it doesn't. Unfortunately the place where I wanted to do this is a framework app. Well, I will just have to do without.