Lamda 表达式里的Join和GroupJoin的区别, 如何实现SQL的Left Join效果

例如，可以将产品表与产品类别表相联接，得到产品名称和与其相对应的类别名称

db.Products  
.Join  
 (  
      db.Categories,  
      p => p.CategoryID,  
      c => c.CategoryID,  
      (p, c) => new {p,c.CategoryName}  
)  
.Where(p => p.CategoryID == 1);  

这个Join对应SQL里的inner Join, 这个很容易理解. 上面代码里红色的c的类型是Category

如果我要达到SQL里Left Join的效果, 该怎么写呢? 想当然的以为,把Join改成GroupJoin,

db.Products  
.GroupJoin  
 (  
      db.Categories,  
      p => p.CategoryID,  
      c => c.CategoryID,  
      (p, cs) => new {p,cs.FirstOrDefault().CategoryName} 
)  
.Where(p => p.CategoryID == 1);  

上面代码里红色的cs的类型是IEnumerable<Category>, 你用FirstOrDefault()只是拿到一个目录,万一Product Left join是有多个目录呢?

所以要用.SelectMany把IEnumerable<Category> 把里面的每一个元素做投影.

要注意加上DefaultIfEmtyp() 才是Left join, 假如没有加这个,还是Inner Join

db.Products
.GroupJoin(db.Categories,
          (Product p) => p.CategoryId,
          (Category c) => c.CategoryId,
          (prod, cs) => new { prod, cs }) // cs is IEnumerable<Category>
          .SelectMany(prodCats => prodCats.cs.DefaultIfEmpty(), (prodCats, c) =>
                                new
                                {
                                    prodCats.prod
                                    categoryName = c.CategoryName
                                });

用LinQ query表达式 对应的写法是

            var qry =
            from p in db.Products
            join c in db.Categories on  p.CategoryId equals c.CategoryId
            into g from pc in g.DefaultIfEmpty()
            select new { 
                    prodId = p.ProdId,
                    prodCode = p.ProdCode,
                    prodName = p.ProdName,
                    categoryName = g.FirstOrDefault().CategoryName
            };

如果我只用DefaultIfEmpty(),不用GroupJoin,这样写能达到这个效果吗?

            var qry = db.Products.SelectMany
            (
                p => db.Categories.Where(c => c.CategoryId == p.CategoryId).DefaultIfEmpty(),
                (p, c) => new
                {
                    p.ProdId,
                    p.ProdCode,
                    p.ProdName,
                    c.CategoryName
                }
            );

这样写,如果数据量大,性能会很差. 因为它会把Products和Categories都Load到内存里,再Match, 而不是用SQL Left Join

参考文章: 

https://stackoverflow.com/questions/584820/how-do-you-perform-a-left-outer-join-using-linq-extension-methods

http://linqsamples.com/linq-to-objects/join/GroupJoin-linq

https://docs.microsoft.com/en-us/dotnet/csharp/linq/perform-left-outer-joins