Question

我需要按时间顺序返回警报列表。下面的方法1做到了这一点。

我还需要获取某种警报类型的警报，以便如果AlertTypeID = {1,2,3,6,8,9，x}，则通过CreateDate返回这些警报并按月进行排序。下面的方法2就是这样做的。

现在，我需要将方法2的结果插入到方法1的结果中。即使方法2返回的项目按月组织，我们也只是关心最近的。只有方法2中最新的子组项需要遵循从方法1返回的项的时间顺序。此子组 item将确定子组2中其余每月项目的位置。

最后一项要求是必须删除重复项。如果正在退货从子组警报中，它不能也存在于主组警报中。

下面我举例说明了我想要达到的效果：

december alert 1 (12/23/2012)
december alert 2 (12/21/2012)
december alert 3 (12/20/2012)
december subalert 1 (12/19/2012)
december subalert 2 (12/18/2012)
december subalert 3 (12/04/2012)
december subalert 4 (12/01/2012)
december alert 4 (12/18/2012)
december alert 5 (12/12/2012)
november alert 1 (11/22/2012)
november alert 2 (11/16/2012)
november subalert 1 (11/14/2012)
november subalert 2 (11/08/2012)
november alert 3 (11/12/2012)

代码：

所有提醒按日期时间

List<Alert> result = new List<Alert>();
using(NeuroLabLinqDataContext dc = conn.GetContext())
{
    IEnumerable<Alert> alerts = (from a in dc.Alerts
                                 where a.AccountID == AccountID
                                 orderby a.CreateDate descending
                                 select a).Take(40);
    result = alerts.ToList();
}
return result;

按月提醒的警报类型

List<Alert> result = new List<Alert>();
int[] alertTypes = {1,2,3,4,5,6,7,8,9};

using (NeuroLabLinqDataContext dc = conn.GetContext())
{
    IEnumerable<Alert> alerts = (from a in dc.Alerts
                                 where a.AccountID == AccountID &&
                                 alertTypes.Contains(a.AlertTypeID)
                                 orderby ((DateTime)a.CreateDate).Month ascending
                                 select a).ToList();
}

return result;

最终select语句的分组应如下所示：

 select new { Date = alerts.CreateDate, 
              Message = alerts.Message, 
              Type = alert.AlertTypeID, 
              RecentActivity = [if from method 2, then true] };

更新：已更新方法

 public List<Alert> GetAlertsByAccountID(Int32 AccountID, params int[] alertTypes)
        {

            List<Alert> result = new List<Alert>();

            using (NeuroLabLinqDataContext dc = conn.GetContext())
            {
                var all = (from a in dc.Alerts
                           where a.AccountID == AccountID
                           orderby a.CreateDate descending
                           select a);

                int abc = all.Count();

                var first = all
                    .Where(a => a.AccountID == AccountID) && !alertTypes.Contains(a.AlertTypeID))
                    .OrderByDescending(a => a.CreateDate)
                    .GroupBy(a => a.CreateDate.Date)
                    .ToDictionary(g => g.Key);

                var firstKeys = first.Keys.Cast<DateTime>()
                    .ToList().OrderBy(k => k);

                var second = all
                    .Where(a => a.AccountID == AccountID) && alertTypes.Contains(a.AlertTypeID))
                    .OrderBy(a => a.CreateDate.Month)
                    .GroupBy(a => a.CreateDate.Month)
                    .ToDictionary(g => firstKeys
                        .First(k => k > g.OrderByDescending(a => a.CreateDate)
                        .FirstOrDefault().CreateDate));

                var combined = first
                    .GroupJoin(
                        second,
                        fk => fk.Key,
                        sk => sk.Key,
                        (d, l) => d.Value
                            .Union(l.SelectMany(i => i.Value).ToArray()))
                    .SelectMany(i => i);

                    result = combined.ToList(); 
            }

            return result;

        }

感谢约翰，我更进一步。目前，我收到以下错误：

序列不包含匹配元素

在这一行（我很确定）：

.First(k => k > g.OrderByDescending(a => a.CreateDate)

对于它的价值，这里是我的警报表中的数据。

AlertID AccountID   CreateDate  Timestamp   AlertTypeID    Message
122 5   2008-03-11 20:48:07.983 0x00000000000128FB  9        sdfs
123 1   2008-03-11 20:48:39.957 0x00000000000128FE  8        sdfsd
124 5   2008-03-11 20:48:39.977 0x00000000000128FF  8        sdfs
125 5   2008-03-11 20:48:40.017 0x0000000000012901  8        asdfa
126 1   2008-03-12 22:57:42.160 0x00000000000130B3  4        sfsf
127 5   2008-03-12 22:57:42.337 0x00000000000130B4  4        sdfsd
128 5   2008-03-13 09:42:14.237 0x0000000000013889  4        sdfsd
129 5   2008-03-13 09:42:31.957 0x000000000001388B  4        sdfsd
130 5   2008-03-13 09:42:45.397 0x000000000001388D  5        asdfsdf
131 1   2008-03-16 14:52:17.197 0x0000000000014822  9        asdfsdf
132 1   2008-04-12 15:25:17.330 0x000000000001B582  3        sfasdf
133 5   2008-04-12 15:25:17.700 0x000000000001B583  3        dfsfds
134 6   2008-04-14 08:37:03.273 0x000000000001BD87  3        aasfsd
135 6   2008-04-14 08:37:15.270 0x000000000001BD89  3        fhfsdf
136 6   2008-04-14 08:38:45.120 0x000000000001BD8B  2        ghsdgd
137 6   2008-04-14 08:41:30.407 0x000000000001BD9A  4        fghsdfg
138 6   2008-04-14 08:42:30.800 0x000000000001BD9C  4        gfsdf
139 6   2008-04-14 08:42:43.763 0x000000000001BD9E  5        sdfsdf
140 6   2008-04-14 08:49:25.450 0x000000000001BDAA  9        sdfasdfa
141 6   2008-04-14 08:49:34.237 0x000000000001BDAC  9        sdfasdf
142 1   2008-04-14 08:50:23.380 0x000000000001BDAF  8        sdfhdfhsg
143 6   2008-04-14 08:50:23.567 0x000000000001BDB0  8        dgasdf
144 5   2008-04-14 08:50:23.690 0x000000000001BDB1  8        dgasdf
145 6   2008-04-14 08:50:23.747 0x000000000001BDB2  8         dgasdf
147 1   2008-06-24 14:22:41.183 0x00000000000222E6  14       dgasdf
148 5   2008-06-24 14:22:41.617 0x00000000000222E7  14       dgasdf
149 6   2008-06-24 14:22:41.623 0x00000000000222E8  14       dgasdf
150 1   2008-06-24 20:11:57.757 0x0000000000022AB3  13     dgasdf
151 5   2008-06-24 20:11:57.947 0x0000000000022AB4  13       dgasdf
152 6   2008-06-24 20:11:57.953 0x0000000000022AB5  13       dgasdf
153 1   2008-07-03 18:41:51.067 0x0000000000028888  14       dgasdf
154 5   2008-07-03 18:41:51.230 0x0000000000028889  14        dgasdf
155 6   2008-07-03 18:41:51.237 0x000000000002888A  14        dgasdf
156 1   2008-07-03 18:46:17.873 0x000000000002888D  14       dgasdf
157 5   2008-07-03 18:46:17.937 0x000000000002888E  14       dgasdf
158 6   2008-07-03 18:46:17.940 0x000000000002888F  14        dgasdf

Answer 1

关键是将两个组分成字典，使用第一个列表中的日期作为字典的键，并在第二个列表项日期之后选择最接近的键作为第二个字典的键。

如果您有两个词典，每个词典都使用类型和子类型的公共键值，您只需执行GroupJoin和SelectMany即可将结果输入到排序列表中。

（*请注意，答案是基于稍微不同的早期版本的问题，我不会花时间更新答案，因为我认为已经在这个答案中说明和解决了基本问题）

更新2 我意识到你在First（）调用中看到的问题是你的一些subalert项可能比任何其他警报项更新，这会导致你的异常。我通过使用DateTime::MaxValue向第一个字典添加'代理'键来解决这个问题，然后我不再从第一个列表中过滤出子图，我只是在最终结果上使用.Distinct()来删除重复项

使用linqpad我模拟了这个问题，并使用词典和GroupJoin

解决了这个问题

var all = new []{
    new {date = DateTime.Parse("2012-12-23"), type = "alert", value = 1, accountId = 333 },
    new {date = DateTime.Parse("2012-12-21"), type = "alert", value = 2, accountId = 333 },
    new {date = DateTime.Parse("2012-12-20"), type = "alert", value = 3, accountId = 333 },
    new {date = DateTime.Parse("2012-12-18"), type = "alert", value = 4, accountId = 333 },
    new {date = DateTime.Parse("2012-12-12"), type = "alert", value = 5, accountId = 333 },
    new {date = DateTime.Parse("2012-11-22"), type = "alert", value = 1, accountId = 333 },
    new {date = DateTime.Parse("2012-11-16"), type = "alert", value = 2, accountId = 333 },
    new {date = DateTime.Parse("2012-11-12"), type = "alert", value = 3, accountId = 333 },
    new {date = DateTime.Parse("2012-12-19"), type = "subalert", value = 1, accountId = 333 },
    new {date = DateTime.Parse("2012-12-18"), type = "subalert", value = 2, accountId = 333 },
    new {date = DateTime.Parse("2012-12-04"), type = "subalert", value = 3, accountId = 333 },
    new {date = DateTime.Parse("2012-12-01"), type = "subalert", value = 4, accountId = 333 },
    new {date = DateTime.Parse("2012-11-14"), type = "subalert", value = 1, accountId = 333 },
    new {date = DateTime.Parse("2012-11-08"), type = "subalert", value = 2, accountId = 333 },  
/*add*/ new {date = DateTime.Parse("2012-12-25"), type = "subalert", value = 9, accountId = 333 },  
};

var first = all
    .Where(a=>a.accountId == 333 /* removed && type != "alert" */)
    .OrderByDescending(a=>a.date)
    .GroupBy(a=>a.date.Date)
    .ToDictionary(g=>g.Key);

var firstKeys = first.Keys
    .Cast<DateTime>()
    .Union(new []{DateTime.MaxValue}) /* added this 'surrogate' key */
    .OrderBy(k=>k)
    .ToArray();

var second = all
    .Where(a=>a.accountId == 333 && a.type == "subalert")
    .OrderBy(a=>a.date.Month)
    .GroupBy(a=>a.date.Month)
    .ToDictionary(g=>firstKeys.First(k=>k > g.OrderByDescending(a=>a.date).FirstOrDefault().date));


var combined = first
    .GroupJoin(
        second,
        fk=>fk.Key,
        sk=>sk.Key,
        (d,l)=>d.Value
            .Union(l.SelectMany(i=>i.Value).ToArray()))
    .SelectMany(i=>i)
    .Distinct(); /* Added this to remove duplicates */

combined.Dump();

哪个收益率：

Updated Results

将一个List <customtype>复杂合并到另一个</customtype>中

1 个答案: