您的位置:时时app平台注册网站 > 编程知识 > 基于Dapper二次封装了一个易用的ORM工具类:SqlD

基于Dapper二次封装了一个易用的ORM工具类:SqlD

2019-10-13 18:43

依赖Dapper三回封装了二个易用的ORM工具类:SqlDapperUtil,把常备能用到的各类CRUD都开展了简化封装,让普通程序员只需关怀职业就可以,因为特轻松,故直接贴源代码,我们若需使用能够直接复制到项目中,该SqlDapperUtil已常见用于公司项目中。

2、针对Dapper封装了常用的CRUD方法,有单独的查询语法。

  1. 我们定义三个轻松UserInfoMo实体(包罗mobile等特性)

 别的说爱他美(Aptamil)下,为了防范和压缩因DB连接未立刻放出导致的连接池不足等原因,故默许实行全体的CRUD方法都以用完即自由,但有一种情状不会自由正是应用了作业,若选拔专业,则少不了配套使用:UseDbTransaction、Commit、或退步推行Rollback,不然大概引致未能及时放出对象,当然最后当SqlDapperUtil实例被回收后事务若未有交到或回滚,会强制实践回滚操作并释放专门的学问及连接对象,幸免大概的能源浪费意况。

自然那也是可取, 好声音的先生们有时说某一个人就是张白纸……

    比相当多ORM的实体映射通过反射来成功,那一点上Dapper再度显现其魅力,在Commond参数赋值,以至实体转变等关键模块,使用了Reflection.Emit功用,直接完结了MSIL编写翻译层面包车型客车赋值达成,之所以说直接,是因为其本身代码还亟需编写翻译器生成IL代码。在运转时依据项目属性动态创制赋值委托方法。

            //使用事务创建多张表,多条SQL语句写在一起
            try
            {
                dapper.UseDbTransaction();
                dapper.ExecuteCommand(@"create table T_Person(Name nvarchar(20) primary key,Age int,BirthDay datetime,HomeAddress nvarchar(200));
                                                    create table T_CarInfo(DriverName nvarchar(20) primary key,Brand nvarchar(50),ManufactureDate datetime)");
                dapper.Commit();
            }
            catch (Exception ex)
            {
                dapper.Rollback();
                //记日志
            }

            //使用事务批量插入多张表的多个记录,多条SQL分多次执行(参数支持批量集合对象传入,无需循环)
            try
            {
                dapper.UseDbTransaction();
                dapper.ExecuteCommand(@"insert into T_Person
select N'张三' as Name,22 as Age,'2018-1-1' as BirthDay,N'中国广东深圳' as HomeAddress union all
select N'李四' as Name,25 as Age,'2018-10-1' as BirthDay,N'中国广东深圳' as HomeAddress union all
select N'王五' as Name,35 as Age,'1982-10-1' as BirthDay,N'中国广东广州' as HomeAddress");


                var carInfos = dapper.GetModelList<CarInfo>(@"
select N'张三' as DriverName,N'大众' as Brand,'2018-8-8' as ManufactureDate union all
select N'李四' as DriverName,N'奔驰' as Brand,'2018-1-8' as ManufactureDate union all
select N'王五' as DriverName,N'奥迪' as Brand,'2017-8-8' as ManufactureDate");

                dapper.ExecuteCommand(@"insert into T_CarInfo(DriverName,Brand,ManufactureDate) Values(@DriverName,@Brand,@ManufactureDate)", carInfos);

                dapper.Commit();
            }
            catch (Exception ex)
            {
                dapper.Rollback();
                //记日志
            }

            //执行删除,有参数,参数可以是实体类、匿名对象、字典(如有需要,可以是集合,以支持批量操作)
            bool deleteResult = dapper.ExecuteCommand("delete from T_CarInfo where DriverName=@DriverName", new { DriverName = "李四" });

            //构建动态执行SQL语句(以下是更新,查询类似)
            StringBuilder updateSqlBuilder = new StringBuilder();
            var updateParams = new Dictionary<string, object>();

            if (1 == 1)
            {
                updateSqlBuilder.Append(",Age=@Age");
                updateParams["Age"] = 20;
            }

            if (2 == 2)
            {
                updateSqlBuilder.Append(",BirthDay=@BirthDay");
                updateParams["BirthDay"] = Convert.ToDateTime("2010-1-1");
            }

            if (3 == 3)
            {
                updateSqlBuilder.Append(",HomeAddress=@HomeAddress");
                updateParams["HomeAddress"] = "中国北京天安门";
            }

            string updateSql = string.Concat("update T_Person set ", updateSqlBuilder.ToString().TrimStart(','), "  where  Name=@Name");
            updateParams["Name"] = "张三";

            bool updateResult = dapper.ExecuteCommand(updateSql, updateParams);

            //查询返回动态自定义结果,之所以不直接返回Dynamic就好,是因为可读性差,故尽可能的在执行后就转成指定的类型
            Tuple<string, int> hasCarInfo = dapper.GetDynamicModel<Tuple<string, int>>((rs) =>
            {
                var result = rs.First();
                return Tuple.Create<string, int>(result.Name, result.CarCount);
            }, @"select a.Name,count(b.DriverName) as CarCount from T_Person a left join T_CarInfo b on a.Name=b.DriverName where a.Name=@Name group by a.Name", new { Name = "张三" });

图片 1图片 2

一. 数目存款和储蓄层供给

以上示例方法用到了七个类,如下:

ORM亦不是万能的,比方做大数量的批量插入什么的,仍旧要求SqlHelper,加上有些人就喜爱DataTable恐怕DataSet。

      2. 包裹功用轻松强盛,对原生SQL的支撑上很灵活

本来已经想计算一下这篇小说,但直接由于职业太忙没有的时候间,今天应用加班探讨.NET CORE的空隙时间实现,请我们支持,有好东西小编自然会享用的,纵然不自然高大上,但一定实用且项目中有实战过的。

图片 3图片 4

   上边的完成部分中大概要求您对.NET的 泛型,委托,扩充,表明式等有八个基础驾驭。正是因为那些语言特色,方便大家对操作共性的抽出统一。

2.底下是身体力行怎么样开展增、删、改以致动态查询的状态:

1、好几年没更新了

  首先,如图所示,我们贯彻了读写分离的四个扩充,其实说起底都会经过Excute方法,那么这里突显下方法的切实可行落到实处:

SqlDapperUtil类中凭仗了事先小编封装的类:如:MemoryCacheUtil(本地内存依赖缓存实用工具类)、ConfigUtil(配置文件管理工科具类)、EncryptUtil(加密工具类),假设项目中不想援引这个类,能够移除或改成任何格局即可。

using System;using System.Collections.Generic;using System.Data;using Dapper;using DapperExtensions;using HY.DataAccess;namespace HY.ORM{    /// <summary>    /// Repository基类    /// </summary>    public class RepositoryBase : RepositoryServiceBase, IDataRepository    {        public RepositoryBase()        {        }        public new void SetDBSession(IDBSession dbSession)        {            base.SetDBSession(dbSession);        }        public RepositoryBase(IDBSession dbSession)            : base(dbSession)        {        }        /// <summary>        /// 根据条件筛选出数据集合        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="sql"></param>        /// <param name="param"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public IEnumerable<T> Get<T>(string sql, dynamic param = null, bool buffered = true) where T : class        {            return DBSession.Connection.Query<T>(sql, param as object, DBSession.Transaction, buffered);        }        /// <summary>        /// 根据条件筛选数据集合        /// </summary>        /// <param name="sql"></param>        /// <param name="param"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public IEnumerable<dynamic> Get(string sql, dynamic param = null, bool buffered = true)        {            return DBSession.Connection.Query(sql, param as object, DBSession.Transaction, buffered);        }        /// <summary>        /// 分页查询        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="pageIndex"></param>        /// <param name="pageSize"></param>        /// <param name="allRowsCount"></param>        /// <param name="sql"></param>        /// <param name="param"></param>        /// <param name="allRowsCountSql"></param>        /// <param name="allRowsCountParam"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public IEnumerable<T> GetPage<T>(int pageIndex, int pageSize, out long allRowsCount, string sql, dynamic param = null, string allRowsCountSql = null, dynamic allRowsCountParam = null, bool buffered = true) where T : class        {            IEnumerable<T> entityList = DBSession.Connection.GetPage<T>(pageIndex, pageSize, out allRowsCount, sql, param as object, allRowsCountSql, null, null, buffered, databaseType: DBSession.DatabaseType);            return entityList;        }        /// <summary>        /// 根据表达式筛选        /// </summary>        /// <typeparam name="TFirst"></typeparam>        /// <typeparam name="TSecond"></typeparam>        /// <typeparam name="TReturn"></typeparam>        /// <param name="sql"></param>        /// <param name="map"></param>        /// <param name="param"></param>        /// <param name="transaction"></param>        /// <param name="buffered"></param>        /// <param name="splitOn"></param>        /// <param name="commandTimeout"></param>        /// <returns></returns>        public IEnumerable<TReturn> Get<TFirst, TSecond, TReturn>(string sql, Func<TFirst, TSecond, TReturn> map,            dynamic param = null, IDbTransaction transaction = null, bool buffered = true, string splitOn = "Id",            int? commandTimeout = null)        {            return DBSession.Connection.Query(sql, map, param as object, transaction, buffered, splitOn);        }        /// <summary>        /// 根据表达式筛选        /// </summary>        /// <typeparam name="TFirst"></typeparam>        /// <typeparam name="TSecond"></typeparam>        /// <typeparam name="TReturn"></typeparam>        /// <param name="sql"></param>        /// <param name="map"></param>        /// <param name="param"></param>        /// <param name="transaction"></param>        /// <param name="buffered"></param>        /// <param name="splitOn"></param>        /// <param name="commandTimeout"></param>        /// <returns></returns>        public IEnumerable<TReturn> Get<TFirst, TSecond, TThird, TReturn>(string sql, Func<TFirst, TSecond, TThird, TReturn> map,            dynamic param = null, IDbTransaction transaction = null, bool buffered = true, string splitOn = "Id",            int? commandTimeout = null)        {            return DBSession.Connection.Query(sql, map, param as object, transaction, buffered, splitOn);        }        /// <summary>        /// 获取多实体集合        /// </summary>        /// <param name="sql"></param>        /// <param name="param"></param>        /// <param name="transaction"></param>        /// <param name="commandTimeout"></param>        /// <param name="commandType"></param>        /// <returns></returns>        public SqlMapper.GridReader GetMultiple(string sql, dynamic param = null, IDbTransaction transaction = null,            int? commandTimeout = null, CommandType? commandType = null)        {            return DBSession.Connection.QueryMultiple(sql, param as object, transaction, commandTimeout, commandType);        }        /// <summary>        /// 执行sql操作        /// </summary>        /// <param name="sql"></param>        /// <param name="param"></param>        /// <returns></returns>        public int Execute(string sql, dynamic param = null, IDbTransaction transaction = null)        {            return DBSession.Connection.Execute(sql, param as object, transaction);        }    }}

  2. 积攒基类完毕(BaseRep)

 1.先来模拟各类查询数据(由于是直接写模拟SQL输出,故并未有条件,也便于大家COPY后一直能够测量检验结果)

图片 5

使用示例方法如下:

using (IDbConnection conn = GetConnection    {        const string query = "select * from XO order by id desc";        return conn.Query<XOEntity>(query,null);    }

  当然,如若为了轻易和品质,直接ADO.NET连接理论上来讲是相比异常的快速的做法,不过如此会促成大批量的再次操作逻辑代码,同时也会产生代码的混杂,扩大入保险护复杂度。作为技能人士,不独有需要减轻事情难题提升作用,同一时间也要增进和谐的频率,所以作者会选取一个ORM框架来成功部分基础事业。

3.还会有七个章程:BatchCopyData、BatchMoveData,那是异样封装的,不是遵照Dapper而是基于原生的Ado.net及BCP,指标是火速多量跨DB跨表COPY数据或改造数据,使用也不复杂,提出想通晓的网上朋友可以查阅自己过去的作品

HY.Web.Entity

在HY.Web.Entity的门类中新建Sys_UsersEntity.cs 定义实体类

图片 6View Code

  这一个地点根本分为四个部分,a. 表明式的剖析,乃至参数化的拍卖   b. 扩大Connection的Insert,Update...等Dapper未有扩充的主意:

using Dapper;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Reflection;
using System.Text;
using System.Threading.Tasks;

namespace KYExpress.Common
{
    public class ColumnAttributeTypeMapper<T> : FallbackTypeMapper
    {
        public ColumnAttributeTypeMapper()
            : base(new SqlMapper.ITypeMap[]
                {
                    new CustomPropertyTypeMap(
                       typeof(T),
                       (type, columnName) =>
                           type.GetProperties().FirstOrDefault(prop =>
                               prop.GetCustomAttributes(false)
                                   .OfType<ColumnAttribute>()
                                   .Any(attr => attr.Name == columnName)
                               )
                       ),
                    new DefaultTypeMap(typeof(T))
                })
        {
        }
    }

    [AttributeUsage(AttributeTargets.Property, AllowMultiple = true)]
    public class ColumnAttribute : Attribute
    {
        public string Name { get; set; }
    }

    public class FallbackTypeMapper : SqlMapper.ITypeMap
    {
        private readonly IEnumerable<SqlMapper.ITypeMap> _mappers;

        public FallbackTypeMapper(IEnumerable<SqlMapper.ITypeMap> mappers)
        {
            _mappers = mappers;
        }


        public ConstructorInfo FindConstructor(string[] names, Type[] types)
        {
            foreach (var mapper in _mappers)
            {
                try
                {
                    ConstructorInfo result = mapper.FindConstructor(names, types);
                    if (result != null)
                    {
                        return result;
                    }
                }
                catch (NotImplementedException)
                {
                }
            }
            return null;
        }

        public SqlMapper.IMemberMap GetConstructorParameter(ConstructorInfo constructor, string columnName)
        {
            foreach (var mapper in _mappers)
            {
                try
                {
                    var result = mapper.GetConstructorParameter(constructor, columnName);
                    if (result != null)
                    {
                        return result;
                    }
                }
                catch (NotImplementedException)
                {
                }
            }
            return null;
        }

        public SqlMapper.IMemberMap GetMember(string columnName)
        {
            foreach (var mapper in _mappers)
            {
                try
                {
                    var result = mapper.GetMember(columnName);
                    if (result != null)
                    {
                        return result;
                    }
                }
                catch (NotImplementedException)
                {
                }
            }
            return null;
        }


        public ConstructorInfo FindExplicitConstructor()
        {
            return _mappers
                .Select(mapper => mapper.FindExplicitConstructor())
                .FirstOrDefault(result => result != null);
        }
    }

}

聊到DapperExtensions修改的小地方还蛮多的,下图是一个代码比较的截图。所以一会把代码打包贴上来吧。

  其次,在重重种类中会出现客商和订单在不一致库中的那类情形,因为涉嫌到分库的图景,所以须要子类中能有修改连接串技术,那么这里小编经过构造函数的格局,提供了四个可空参数:

        class Person
        {
            public string Name { get; set; }


            public int Age { get; set; }

            public DateTime BirthDay { get; set; }

            [Column(Name = "HomeAddress")]
            public string HomeAddr { get; set; }

            public CarInfo Car { get; set; }

            public List<CarInfo> Cars { get; set; }
        }

        class CarInfo
        {
            public string Brand { get; set; }

            public DateTime ManufactureDate { get; set; }

            public string DriverName { get; set; }
        }

图片 7

当前项目还在搭建中,要是有意思味的同室也招待出席进去,代码详见Github下OSS.Core

ColumnAttributeTypeMapper协理类相关代码如下:(假使不思虑实体类的习性与表字段不平等的景观,如下映射类能够无需加上,同偶尔间SqlDapperUtil中移除相关信赖ColumnAttributeTypeMapper逻辑就能够)

SqlConnectionFactory.cs 那些类是选拔工厂情势创建DB连接的包裹,代码如下:

  a. 熟习Expression表明式的对象应该比较领悟,表明式本人是贰个树形接口,依据差别的类型,能够持续的深入分析其子表明式,直到不富有持续深入分析的大概。所以这些就极粗略正是递归的接连不断迭代,依据其分歧的NodeType能够创立不相同的sql成分,因为代码较长,能够参见github下的SqlExpressionVisitor.cs类,个中参数的赋值部分,未有运用反射,而是使用的反光发射,代码详见SqlParameterEmit.cs

            //实例化SqlDapperUtil对象,构造函数是config文件中的connectionStrings的Name名
            var dapper = new SqlDapperUtil("LmsConnectionString");

            //查询1个值
            DateTime nowTime = dapper.GetValue<DateTime>("select getdate() as nowtime");


            //查询1行值,并转换成字典(这对于临时查询多个字段而无需定义实体类有用)
            Dictionary<string, dynamic> rowValues = dapper.GetFirstValues("select 0 as col0,1 as col1,2 as col2");


            //查询1行并返回实体类
            Person person = dapper.GetModel<Person>("select '张三' as Name,22 as Age,'2018-1-1' as BirthDay,'中国广东深圳' as HomeAddr");


            //查询1行表字段与实体类属性不一致映射
            Person person2 = dapper.GetModel<Person>("select '张三' as Name,22 as Age,'2018-1-1' as BirthDay,'中国广东深圳' as HomeAddress");


            //查询多行返回实体集合
            var persons = dapper.GetModelList<Person>(@"select '张三' as Name,22 as Age,'2018-1-1' as BirthDay,'中国广东深圳' as HomeAddr union all
                                                                            select '李四' as Name,25 as Age,'2018-10-1' as BirthDay,'中国广东深圳' as HomeAddress union all
                                                                            select '王五' as Name,35 as Age,'1982-10-1' as BirthDay,'中国广东广州' as HomeAddress
                                                                        ");


            //查询多行返回1对1关联实体结果集
            var personWithCarResult = dapper.GetMultModelList<Person>(@"select t1.*,t2.* from
                                                                                                    (select '张三' as Name,22 as Age,'2018-1-1' as BirthDay,'中国广东深圳' as HomeAddr union all
                                                                                                    select '李四' as Name,25 as Age,'2018-10-1' as BirthDay,'中国广东深圳' as HomeAddress union all
                                                                                                    select '王五' as Name,35 as Age,'1982-10-1' as BirthDay,'中国广东广州' as HomeAddress)as t1 inner join
                                                                                                    (
                                                                                                    select '张三' as DriverName,'大众' as Brand,'2018-8-8' as ManufactureDate union all
                                                                                                    select '李四' as DriverName,'奔驰' as Brand,'2018-1-8' as ManufactureDate union all
                                                                                                    select '王五' as DriverName,'奥迪' as Brand,'2017-8-8' as ManufactureDate
                                                                                                    )as t2
                                                                                                    on t1.Name=t2.DriverName
                                                                        ", new[] { typeof(Person), typeof(CarInfo) }, (objs) =>
                                                                         {
                                                                             Person personItem = objs[0] as Person;
                                                                             CarInfo carItem = objs[1] as CarInfo;
                                                                             personItem.Car = carItem;
                                                                             return personItem;
                                                                         }, splitOn: "DriverName");




            //查询多行返回1对多关联实体结果=personWithManyCars
            List<Person> personWithManyCars = new List<Person>();
            dapper.GetMultModelList<Person>(@"select t1.*,t2.* from
                                                                                                    (select '张三' as Name,22 as Age,'2018-1-1' as BirthDay,'中国广东深圳' as HomeAddr union all
                                                                                                    select '李四' as Name,25 as Age,'2018-10-1' as BirthDay,'中国广东深圳' as HomeAddress union all
                                                                                                    select '王五' as Name,35 as Age,'1982-10-1' as BirthDay,'中国广东广州' as HomeAddress)as t1 inner join
                                                                                                    (
                                                                                                    select '张三' as DriverName,'大众' as Brand,'2018-8-8' as ManufactureDate union all
                                                                                                    select '张三' as DriverName,'奔驰' as Brand,'2018-1-8' as ManufactureDate union all
                                                                                                    select '张三' as DriverName,'奥迪' as Brand,'2017-8-8' as ManufactureDate
                                                                                                    )as t2
                                                                                                    on t1.Name=t2.DriverName
                                                                        ", new[] { typeof(Person), typeof(CarInfo) }, (objs) =>
                                                                        {
                                                                            Person personItem = objs[0] as Person;
                                                                            CarInfo carItem = objs[1] as CarInfo;

                                                                            Person personItemMain = personWithManyCars.FirstOrDefault(p => p.Name == personItem.Name);
                                                                            if (personItemMain == null)
                                                                            {
                                                                                personItem.Cars = new List<CarInfo>();
                                                                                personItemMain = personItem;
                                                                                personWithManyCars.Add(personItemMain);
                                                                            }

                                                                            personItemMain.Cars.Add(carItem);
                                                                            return personItemMain;

                                                                        }, splitOn: "DriverName");

上边说说ORM那块的兑现方式。见下截图

 

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using Dapper;
using System.Data;
using System.Data.Common;
using System.Reflection;
using System.IO;
using System.Collections.Concurrent;
using System.Data.SqlClient;

namespace Zuowj.Common
{
    /// <summary>
    /// 基于Dapper的数据操作类封装的工具类
    /// Author:左文俊
    /// Date:2017/12/11
    /// </summary>
    public class SqlDapperUtil
    {
        private static string dbConnectionStringConfigPath = null;
        private readonly static ConcurrentDictionary<string, bool> dbConnNamesCacheDic = new ConcurrentDictionary<string, bool>();

        private string dbConnectionName = null;
        private string dbConnectionString = null;
        private string dbProviderName = null;
        private IDbConnection dbConnection = null;
        private bool useDbTransaction = false;
        private IDbTransaction dbTransaction = null;


        #region 私有方法

        private IDbConnection GetDbConnection()
        {
            bool needCreateNew = false;
            if (dbConnection == null || string.IsNullOrWhiteSpace(dbConnection.ConnectionString))
            {
                needCreateNew = true;
            }
            else if (!MemoryCacheUtil.Contains(dbConnectionName))
            {
                needCreateNew = true;
            }

            if (needCreateNew)
            {
                dbConnectionString = GetDbConnectionString(dbConnectionName, out dbProviderName);
                var dbProviderFactory = DbProviderFactories.GetFactory(dbProviderName);
                dbConnection = dbProviderFactory.CreateConnection();
                dbConnection.ConnectionString = dbConnectionString;
            }

            if (dbConnection.State == ConnectionState.Closed)
            {
                dbConnection.Open();
            }

            return dbConnection;
        }

        private string GetDbConnectionString(string dbConnName, out string dbProviderName)
        {
            //如果指定的连接字符串配置文件路径,则创建缓存依赖,一旦配置文件更改就失效,再重新读取
            string[] connInfos = MemoryCacheUtil.GetOrAddCacheItem(dbConnName, () =>
            {
                var connStrSettings = ConfigUtil.GetConnectionStringForConfigPath(dbConnName, SqlDapperUtil.DbConnectionStringConfigPath);
                string dbProdName = connStrSettings.ProviderName;
                string dbConnStr = connStrSettings.ConnectionString;
                //LogUtil.Info(string.Format("SqlDapperUtil.GetDbConnectionString>读取连接字符串配置节点[{0}]:{1},ProviderName:{2}", dbConnName, dbConnStr, dbProdName), "SqlDapperUtil.GetDbConnectionString");
                return new[] { EncryptUtil.Decrypt(dbConnStr), dbProdName };
            }, SqlDapperUtil.DbConnectionStringConfigPath);

            dbProviderName = connInfos[1];
            return connInfos[0];
        }


        private T UseDbConnection<T>(Func<IDbConnection, T> queryOrExecSqlFunc)
        {
            IDbConnection dbConn = null;

            try
            {
                Type modelType = typeof(T);
                var typeMap = Dapper.SqlMapper.GetTypeMap(modelType);
                if (typeMap == null || !(typeMap is ColumnAttributeTypeMapper<T>))
                {
                    Dapper.SqlMapper.SetTypeMap(modelType, new ColumnAttributeTypeMapper<T>());
                }

                dbConn = GetDbConnection();
                if (useDbTransaction && dbTransaction == null)
                {
                    dbTransaction = GetDbTransaction();
                }

                return queryOrExecSqlFunc(dbConn);
            }
            catch
            {
                throw;
            }
            finally
            {
                if (dbTransaction == null && dbConn != null)
                {
                    CloseDbConnection(dbConn);
                }
            }
        }

        private void CloseDbConnection(IDbConnection dbConn, bool disposed = false)
        {
            if (dbConn != null)
            {
                if (disposed && dbTransaction != null)
                {
                    dbTransaction.Rollback();
                    dbTransaction.Dispose();
                    dbTransaction = null;
                }

                if (dbConn.State != ConnectionState.Closed)
                {
                    dbConn.Close();
                }
                dbConn.Dispose();
                dbConn = null;
            }
        }

        /// <summary>
        /// 获取一个事务对象(如果需要确保多条执行语句的一致性,必需使用事务)
        /// </summary>
        /// <param name="il"></param>
        /// <returns></returns>
        private IDbTransaction GetDbTransaction(IsolationLevel il = IsolationLevel.Unspecified)
        {
            return GetDbConnection().BeginTransaction(il);
        }

        private DynamicParameters ToDynamicParameters(Dictionary<string, object> paramDic)
        {
            return new DynamicParameters(paramDic);
        }

        #endregion

        public static string DbConnectionStringConfigPath
        {
            get
            {
                if (string.IsNullOrEmpty(dbConnectionStringConfigPath))//如果没有指定配置文件,则取默认的配置文件路径作为缓存依赖路径
                {
                    dbConnectionStringConfigPath = BaseUtil.GetConfigPath();
                }

                return dbConnectionStringConfigPath;
            }
            set
            {
                if (!string.IsNullOrWhiteSpace(value) && !File.Exists(value))
                {
                    throw new FileNotFoundException("指定的DB连接字符串配置文件不存在:"   value);
                }

                //如果配置文件改变,则可能导致连接字符串改变,故必需清除所有连接字符串的缓存以便后续重新加载字符串
                if (!string.Equals(dbConnectionStringConfigPath, value, StringComparison.OrdinalIgnoreCase))
                {
                    foreach (var item in dbConnNamesCacheDic)
                    {
                        MemoryCacheUtil.RemoveCacheItem(item.Key);
                    }
                }

                dbConnectionStringConfigPath = value;
            }
        }

        public SqlDapperUtil(string connName)
        {
            dbConnectionName = connName;
            if (!dbConnNamesCacheDic.ContainsKey(connName)) //如果静态缓存中没有,则加入到静态缓存中
            {
                dbConnNamesCacheDic[connName] = true;
            }

        }


        /// <summary>
        /// 使用事务
        /// </summary>
        public void UseDbTransaction()
        {
            useDbTransaction = true;
        }


        /// <summary>
        /// 获取一个值,param可以是SQL参数也可以是匿名对象
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public T GetValue<T>(string sql, object param = null, int? commandTimeout = null, CommandType? commandType = null)
        {
            return UseDbConnection((dbConn) =>
             {
                 return dbConn.ExecuteScalar<T>(sql, param, dbTransaction, commandTimeout, commandType);
             });
        }

        /// <summary>
        /// 获取第一行的所有值,param可以是SQL参数也可以是匿名对象
        /// </summary>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public Dictionary<string, dynamic> GetFirstValues(string sql, object param = null, int? commandTimeout = null, CommandType? commandType = null)
        {
            return UseDbConnection((dbConn) =>
            {
                Dictionary<string, dynamic> firstValues = new Dictionary<string, dynamic>();
                List<string> indexColNameMappings = new List<string>();
                int rowIndex = 0;
                using (var reader = dbConn.ExecuteReader(sql, param, dbTransaction, commandTimeout, commandType))
                {
                    while (reader.Read())
                    {
                        if ((  rowIndex) > 1) break;
                        if (indexColNameMappings.Count == 0)
                        {
                            for (int i = 0; i < reader.FieldCount; i  )
                            {
                                indexColNameMappings.Add(reader.GetName(i));
                            }
                        }

                        for (int i = 0; i < reader.FieldCount; i  )
                        {
                            firstValues[indexColNameMappings[i]] = reader.GetValue(i);
                        }
                    }
                    reader.Close();
                }

                return firstValues;

            });
        }

        /// <summary>
        /// 获取一个数据模型实体类,param可以是SQL参数也可以是匿名对象
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public T GetModel<T>(string sql, object param = null, int? commandTimeout = null, CommandType? commandType = null) where T : class
        {
            return UseDbConnection((dbConn) =>
            {
                return dbConn.QueryFirstOrDefault<T>(sql, param, dbTransaction, commandTimeout, commandType);
            });
        }

        /// <summary>
        /// 获取符合条件的所有数据模型实体类列表,param可以是SQL参数也可以是匿名对象
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="buffered"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public List<T> GetModelList<T>(string sql, object param = null, bool buffered = true, int? commandTimeout = null, CommandType? commandType = null) where T : class
        {
            return UseDbConnection((dbConn) =>
            {
                return dbConn.Query<T>(sql, param, dbTransaction, buffered, commandTimeout, commandType).ToList();
            });
        }

        /// <summary>
        /// 获取符合条件的所有数据并根据动态构建Model类委托来创建合适的返回结果(适用于临时性结果且无对应的模型实体类的情况)
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="buildModelFunc"></param>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="buffered"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public T GetDynamicModel<T>(Func<IEnumerable<dynamic>, T> buildModelFunc, string sql, object param = null, bool buffered = true, int? commandTimeout = null, CommandType? commandType = null)
        {
            var dynamicResult = UseDbConnection((dbConn) =>
           {
               return dbConn.Query(sql, param, dbTransaction, buffered, commandTimeout, commandType);
           });

            return buildModelFunc(dynamicResult);
        }

        /// <summary>
        /// 获取符合条件的所有指定返回结果对象的列表(复合对象【如:1对多,1对1】),param可以是SQL参数也可以是匿名对象
        /// </summary>
        /// <typeparam name="T"></typeparam>
        /// <param name="sql"></param>
        /// <param name="types"></param>
        /// <param name="map"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="buffered"></param>
        /// <param name="splitOn"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>

        public List<T> GetMultModelList<T>(string sql, Type[] types, Func<object[], T> map, object param = null, bool buffered = true, string splitOn = "Id", int? commandTimeout = null, CommandType? commandType = null)
        {
            return UseDbConnection((dbConn) =>
            {
                return dbConn.Query<T>(sql, types, map, param, dbTransaction, buffered, splitOn, commandTimeout, commandType).ToList();
            });
        }




        /// <summary>
        /// 执行SQL命令(CRUD),param可以是SQL参数也可以是要添加的实体类
        /// </summary>
        /// <param name="sql"></param>
        /// <param name="param"></param>
        /// <param name="transaction"></param>
        /// <param name="commandTimeout"></param>
        /// <param name="commandType"></param>
        /// <returns></returns>
        public bool ExecuteCommand(string sql, object param = null, int? commandTimeout = null, CommandType? commandType = null)
        {
            return UseDbConnection((dbConn) =>
            {
                int result = dbConn.Execute(sql, param, dbTransaction, commandTimeout, commandType);
                return (result > 0);
            });
        }

        /// <summary>
        /// 批量转移数据(利用SqlBulkCopy实现快速大批量插入到指定的目的表及SqlDataAdapter的批量删除)
        /// </summary>
        public bool BatchMoveData(string srcSelectSql, string srcTableName, List<SqlParameter> srcPrimarykeyParams, string destConnName, string destTableName)
        {

            using (SqlDataAdapter srcSqlDataAdapter = new SqlDataAdapter(srcSelectSql, GetDbConnectionString(dbConnectionName, out dbProviderName)))
            {
                DataTable srcTable = new DataTable();
                SqlCommand deleteCommand = null;
                try
                {
                    srcSqlDataAdapter.AcceptChangesDuringFill = true;
                    srcSqlDataAdapter.AcceptChangesDuringUpdate = false;
                    srcSqlDataAdapter.Fill(srcTable);

                    if (srcTable == null || srcTable.Rows.Count <= 0) return true;

                    string notExistsDestSqlWhere = null;
                    string deleteSrcSqlWhere = null;

                    for (int i = 0; i < srcPrimarykeyParams.Count; i  )
                    {
                        string keyColName = srcPrimarykeyParams[i].ParameterName.Replace("@", "");
                        notExistsDestSqlWhere  = string.Format(" AND told.{0}=tnew.{0}", keyColName);
                        deleteSrcSqlWhere  = string.Format(" AND {0}=@{0}", keyColName);
                    }

                    string dbProviderName2 = null;
                    using (var destConn = new SqlConnection(GetDbConnectionString(destConnName, out dbProviderName2)))
                    {
                        destConn.Open();

                        string tempDestTableName = "#temp_"   destTableName;
                        destConn.Execute(string.Format("select top 0 * into {0} from {1}", tempDestTableName, destTableName));
                        string destInsertCols = null;
                        using (var destSqlBulkCopy = new SqlBulkCopy(destConn))
                        {
                            try
                            {
                                destSqlBulkCopy.BulkCopyTimeout = 120;
                                destSqlBulkCopy.DestinationTableName = tempDestTableName;
                                foreach (DataColumn col in srcTable.Columns)
                                {
                                    destSqlBulkCopy.ColumnMappings.Add(col.ColumnName, col.ColumnName);
                                    destInsertCols  = ","   col.ColumnName;
                                }

                                destSqlBulkCopy.BatchSize = 1000;
                                destSqlBulkCopy.WriteToServer(srcTable);
                            }
                            catch (Exception ex)
                            {
                                //LogUtil.Error("SqlDapperUtil.BatchMoveData.SqlBulkCopy:"   ex.ToString(), "SqlDapperUtil.BatchMoveData");
                            }

                            destInsertCols = destInsertCols.Substring(1);

                            destConn.Execute(string.Format("insert into {1}({0}) select {0} from {2} tnew where not exists(select 1 from {1} told where {3})",
                                             destInsertCols, destTableName, tempDestTableName, notExistsDestSqlWhere.Trim().Substring(3)), null, null, 100);
                        }
                        destConn.Close();
                    }

                    deleteCommand = new SqlCommand(string.Format("DELETE FROM {0} WHERE {1}", srcTableName, deleteSrcSqlWhere.Trim().Substring(3)), srcSqlDataAdapter.SelectCommand.Connection);
                    deleteCommand.Parameters.AddRange(srcPrimarykeyParams.ToArray());
                    deleteCommand.UpdatedRowSource = UpdateRowSource.None;
                    deleteCommand.CommandTimeout = 200;

                    srcSqlDataAdapter.DeleteCommand = deleteCommand;
                    foreach (DataRow row in srcTable.Rows)
                    {
                        row.Delete();
                    }

                    srcSqlDataAdapter.UpdateBatchSize = 1000;
                    srcSqlDataAdapter.Update(srcTable);
                    srcTable.AcceptChanges();

                    return true;
                }
                catch (Exception ex)
                {
                    //LogUtil.Error("SqlDapperUtil.BatchMoveData:"   ex.ToString(), "SqlDapperUtil.BatchMoveData");
                    return false;
                }
                finally
                {
                    if (deleteCommand != null)
                    {
                        deleteCommand.Parameters.Clear();
                    }
                }
            }

        }

        /// <summary>
        /// 批量复制数据(把源DB中根据SQL语句查出的结果批量COPY插入到目的DB的目的表中)
        /// </summary>
        public TResult BatchCopyData<TResult>(string srcSelectSql, string destConnName, string destTableName, IDictionary<string, string> colMappings, Func<IDbConnection, TResult> afterCoppyFunc)
        {

            using (SqlDataAdapter srcSqlDataAdapter = new SqlDataAdapter(srcSelectSql, GetDbConnectionString(dbConnectionName, out dbProviderName)))
            {
                DataTable srcTable = new DataTable();
                TResult copyResult = default(TResult);
                try
                {
                    srcSqlDataAdapter.AcceptChangesDuringFill = true;
                    srcSqlDataAdapter.AcceptChangesDuringUpdate = false;
                    srcSqlDataAdapter.Fill(srcTable);

                    if (srcTable == null || srcTable.Rows.Count <= 0) return copyResult;


                    string dbProviderName2 = null;
                    using (var destConn = new SqlConnection(GetDbConnectionString(destConnName, out dbProviderName2)))
                    {
                        destConn.Open();
                        string tempDestTableName = "#temp_"   destTableName;
                        destConn.Execute(string.Format("select top 0 * into {0} from {1}", tempDestTableName, destTableName));
                        bool bcpResult = false;
                        using (var destSqlBulkCopy = new SqlBulkCopy(destConn))
                        {
                            try
                            {
                                destSqlBulkCopy.BulkCopyTimeout = 120;
                                destSqlBulkCopy.DestinationTableName = tempDestTableName;
                                foreach (var col in colMappings)
                                {
                                    destSqlBulkCopy.ColumnMappings.Add(col.Key, col.Value);
                                }

                                destSqlBulkCopy.BatchSize = 1000;
                                destSqlBulkCopy.WriteToServer(srcTable);
                                bcpResult = true;
                            }
                            catch (Exception ex)
                            {
                                //LogUtil.Error("SqlDapperUtil.BatchMoveData.SqlBulkCopy:"   ex.ToString(), "SqlDapperUtil.BatchMoveData");
                            }
                        }

                        if (bcpResult)
                        {
                            copyResult = afterCoppyFunc(destConn);
                        }

                        destConn.Close();
                    }

                    return copyResult;
                }
                catch (Exception ex)
                {
                    //LogUtil.Error("SqlDapperUtil.BatchCopyData:"   ex.ToString(), "SqlDapperUtil.BatchCopyData");
                    return copyResult;
                }
            }

        }


        /// <summary>
        /// 当使用了事务,则最后需要调用该方法以提交所有操作
        /// </summary>
        /// <param name="dbTransaction"></param>
        public void Commit()
        {
            try
            {
                if (dbTransaction.Connection != null && dbTransaction.Connection.State != ConnectionState.Closed)
                {
                    dbTransaction.Commit();
                }
            }
            catch
            {
                throw;
            }
            finally
            {
                if (dbTransaction.Connection != null)
                {
                    CloseDbConnection(dbTransaction.Connection);
                }
                dbTransaction.Dispose();
                dbTransaction = null;
                useDbTransaction = false;

                if (dbConnection != null)
                {
                    CloseDbConnection(dbConnection);
                }
            }
        }

        /// <summary>
        /// 当使用了事务,如果报错或需要中断执行,则需要调用该方法执行回滚操作
        /// </summary>
        /// <param name="dbTransaction"></param>
        public void Rollback()
        {
            try
            {
                if (dbTransaction.Connection != null && dbTransaction.Connection.State != ConnectionState.Closed)
                {
                    dbTransaction.Rollback();
                }
            }
            catch
            {
                throw;
            }
            finally
            {
                if (dbTransaction.Connection != null)
                {
                    CloseDbConnection(dbTransaction.Connection);
                }

                dbTransaction.Dispose();
                dbTransaction = null;
                useDbTransaction = false;
            }
        }

        ~SqlDapperUtil()
        {
            try
            {
                CloseDbConnection(dbConnection, true);
            }
            catch
            { }
        }

    }
}

图片 8图片 9

  具体的作业数据接口承袭至基础接口就好,此中说明式部分是本人要好做了一个打包,前面会简介。

using System.Data.Common;using System.Data;namespace HY.DataAccess{    /// <summary>    /// 提供对数据库的基本操作,连接字符串需要在数据库配置。    /// </summary>    public interface IDBHelper    {        /// <summary>        /// 生成分页SQL语句        /// </summary>        /// <param name="pageIndex"></param>        /// <param name="pageSize"></param>        /// <param name="selectSql"></param>        /// <param name="sqlCount"></param>        /// <param name="orderBy"></param>        /// <returns></returns>        string GetPagingSql(int pageIndex, int pageSize, string selectSql, string sqlCount, string orderBy);        /// <summary>        /// 开始一个事务        /// </summary>        /// <returns></returns>        DbTransaction BeginTractionand();        /// <summary>        /// 开始一个事务        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        DbTransaction BeginTractionand(string connKey);        /// <summary>        /// 回滚事务        /// </summary>        /// <param name="dbTransaction">要回滚的事务</param>        void RollbackTractionand(DbTransaction dbTransaction);        /// <summary>        /// 结束并确认事务        /// </summary>        /// <param name="dbTransaction">要结束的事务</param>        void CommitTractionand(DbTransaction dbTransaction);        #region DataSet        /// <summary>        /// 执行sql语句,ExecuteDataSet 返回DataSet        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        DataSet ExecuteDataSet(string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteDataSet 返回DataSet        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        DataSet ExecuteDataSet(string connKey, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteDataSet 返回DataSet        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        DataSet ExecuteDataSet(string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,ExecuteDataSet 返回DataSet        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        DataSet ExecuteDataSet(string connKey, string commandText, CommandType commandType, params DbParameter[] parameterValues);        #endregion        #region ExecuteNonQuery        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        int ExecuteNonQuery(string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        int ExecuteNonQuery(string connKey, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="trans">事务对象</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        int ExecuteNonQuery(DbTransaction trans, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        int ExecuteNonQuery(string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        int ExecuteNonQuery(string connKey, string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,返回影响的行数        /// </summary>        /// <param name="trans">事务对象</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        int ExecuteNonQuery(DbTransaction trans, string commandText, CommandType commandType, params DbParameter[] parameterValues);        #endregion        #region IDataReader        /// <summary>        /// 执行sql语句,ExecuteReader 返回IDataReader        /// </summary>           /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        IDataReader ExecuteReader(string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteReader 返回IDataReader        /// </summary>         /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        IDataReader ExecuteReader(string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,ExecuteReader 返回IDataReader        /// </summary>        /// <param name="connKey">数据库连接字符key</param>                /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        IDataReader ExecuteReader(string connKey, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteReader 返回IDataReader        /// </summary>        /// <param name="connKey">数据库连接字符key</param>                /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        IDataReader ExecuteReader(string connKey, string commandText, CommandType commandType, params DbParameter[] parameterValues);        #endregion        #region ExecuteScalar        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        object ExecuteScalar(string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        object ExecuteScalar(string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="trans">事务</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        object ExecuteScalar(DbTransaction trans, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        object ExecuteScalar(string connKey, string commandText, CommandType commandType);        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="connKey">数据库连接字符key</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        object ExecuteScalar(string connKey, string commandText, CommandType commandType, params DbParameter[] parameterValues);        /// <summary>        /// 执行sql语句,ExecuteScalar 返回第一行第一列的值        /// </summary>        /// <param name="trans">事务</param>        /// <param name="commandText">sql语句</param>        /// <param name="commandType"></param>        /// <param name="parameterValues">参数</param>        /// <returns></returns>        object ExecuteScalar(DbTransaction trans, string commandText, CommandType commandType, params DbParameter[] parameterValues);        #endregion    }}

  1. 接口设计

3、Dapper原理通过Emit反射IDataReader的行列队列来火速的获取和产生对象。性能平日很牛逼的样子

三. OSS.Core仓库储存层设计完成

IDataServiceRepository.cs(提供业务层使用,里面包车型地铁艺术不援助传递sql,满含sql的讲话最佳依然放在数据层操作的好)

  同期一个系列中数据库访谈代码好多都会以CRUD为主,所以那边小编定义了七个基础接口(IBaseRep),其富含的秘技主要有(表明式部分在后头介绍):

using System.Collections.Generic;using System.Data;using DapperExtensions;using HY.DataAccess;namespace HY.ORM{    public interface IDataServiceRepository    {        IDBSession DBSession { get; }        T GetById<T>(dynamic primaryId) where T : class;        IEnumerable<T> GetByIds<T>(IList<dynamic> ids) where T : class;        IEnumerable<T> GetAll<T>() where T : class;        int Count<T>(object predicate, bool buffered = false) where T : class;        //lsit        IEnumerable<T> GetList<T>(object predicate = null, IList<ISort> sort = null, bool buffered = false) where T : class;        IEnumerable<T> GetPageList<T>(int pageIndex, int pageSize, out long allRowsCount, object predicate = null, IList<ISort> sort = null, bool buffered = true) where T : class;        dynamic Insert<T>(T entity, IDbTransaction transaction = null) where T : class;        bool InsertBatch<T>(IEnumerable<T> entityList, IDbTransaction transaction = null) where T : class;        bool Update<T>(T entity, IDbTransaction transaction = null) where T : class;        bool UpdateBatch<T>(IEnumerable<T> entityList, IDbTransaction transaction = null) where T : class;        int Delete<T>(dynamic primaryId, IDbTransaction transaction = null) where T : class;        int DeleteList<T>(object predicate, IDbTransaction transaction = null) where T : class;        bool DeleteBatch<T>(IEnumerable<dynamic> ids, IDbTransaction transaction = null) where T : class;    }}

图片 10

RepositoryServiceBase.cs(IData瑟维斯Repository的达成类)

  3. 基于Connection的扩展

HY.Web.DAO

概念基类 BaseRepository.cs (能够设置暗中同意的DBsession,方便增加此外东东)

图片 11View Code

定义数据访谈层 Sys_UsersRepository.cs (代码里能够打包任何须求写sql 的代码)

图片 12View Code

  既然是多个完好无缺的体系,数据访问是其最大旨的一些,同不日常候,数据访谈也是一切项目最轻易并发瓶颈的地点。在自己的剪切中,其承受的剧中人物是担任整个数据的输入输出,不仅是指向单数据库(一时以至多库),不时还必要完结一级缓存的兑现,给逻辑层提供最基础的数目援助。

View Code

 

using System;using System.Collections.Generic;using System.Data;using Dapper;using HY.DataAccess;namespace HY.ORM{    public interface IDataRepository : IDataServiceRepository    {        IDBSession DBSession { get; }                    IEnumerable<T> Get<T>(string sql, dynamic param = null, bool buffered = true) where T : class;        IEnumerable<dynamic> Get(string sql, dynamic param = null, bool buffered = true);        IEnumerable<TReturn> Get<TFirst, TSecond, TReturn>(string sql, Func<TFirst, TSecond, TReturn> map,            dynamic param = null, IDbTransaction transaction = null, bool buffered = true,            string splitOn = "Id", int? commandTimeout = null);         IEnumerable<TReturn> Get<TFirst, TSecond,TThird, TReturn>(string sql, Func<TFirst, TSecond,TThird, TReturn> map,            dynamic param = null, IDbTransaction transaction = null, bool buffered = true,            string splitOn = "Id", int? commandTimeout = null);        SqlMapper.GridReader GetMultiple(string sql, dynamic param = null, IDbTransaction transaction = null,            int? commandTimeout = null, CommandType? commandType = null);         IEnumerable<T> GetPage<T>(int pageIndex, int pageSize, out long allRowsCount, string sql, dynamic param = null, string allRowsCountSql=null,  dynamic allRowsCountParam = null, bool buffered = true) where T : class;              Int32 Execute(string sql, dynamic param = null, IDbTransaction transaction = null);    }}

图片 13

View Code

   通过Dapper能够实现在数据库访问一些一层简易的包裹,不过本人仍然须要手动编写不菲的sql语句,同不经常候还要开展参数化的管理,富含数据的读写分离等。那么这个成效的达成自个儿将要OSS.Core.RepDapper中实现,为了方便理解,先贴出五个简便的卷入后的格局调用传输流程:

图片 14

   业务恒久是在变幻莫测的,那么项目也要享有急迅多变的才具,所以本身希望数据层能够维持相对的粗略,在结构上尽量减弱复杂的耦合查询,在性质上尽量减弱不供给的损耗,比如反射的汪洋行使。同时针对各种业务对象实现数据库层面为主的CRUD统一封装完毕。假诺有亟待的时候还能够在最少的更改下参与缓存的立异。(对于怎么着兑现差异模块不相同缓存存款和储蓄战术,像Redis,Memcached会在后头小说介绍)

就此SqlHelper作为基础,ORM作为援助,一箭穿心啊。

  3. 概念落成类  UserInfoRep : BaseRep, IUserInfoRep

dapper-dot-net源码: (更新频率快,项目包涵了各类除了Dapper-Extensions的增加类型)

设若你还应该有任何难题,迎接关心大伙儿号(OSSCoder)

Dapper对DB的操作注重于Connection,为了帮衬多库,大家用IDbConnection conn

图片 15

3、无法而且援助各类数据库

  最后,大家也兑现了针对性基础接口方法的实际贯彻,举一示范:

图片 16

  

            //实体类            DemoEntity entity = new DemoEntity();            //根据实体主键删除            this.Delete<DemoEntity>;            //根据主键ID删除            this.Delete<DemoEntity>(1);            //增加            this.Insert<DemoEntity>;            //更新            bool result = this.Update<DemoEntity>;            //根据主键返回实体            entity = this.GetById<DemoEntity>(1);            //返回 行数            this.Count<DemoEntity>(new { ID = 1 });            //查询所有            IEnumerable<DemoEntity> list = this.GetAll<DemoEntity>();            IList<ISort> sort = new List<ISort>();            sort.Add(new Sort { PropertyName = "ID", Ascending = false });            //条件查询            list = this.GetList<DemoEntity>(new { ID = 1, Name = "123" }, sort);            //orm 拼接条件 查询            IList<IPredicate> predList = new List<IPredicate>();            predList.Add(Predicates.Field<DemoEntity>(p => p.Name, Operator.Like, "不知道%"));            predList.Add(Predicates.Field<DemoEntity>(p => p.ID, Operator.Eq, 1));            IPredicateGroup predGroup = Predicates.Group(GroupOperator.And, predList.ToArray;                        list = this.GetList<DemoEntity>(predGroup);            //分页查询            long allRowsCount = 0;            this.GetPageList<DemoEntity>(1, 10, out allRowsCount, new { ID = 1 }, sort);

  能够看来在这么些办法提供了三个针对性IDbConnection的信托,提供调用层自由使用Dapper方法的还要,统一了多少访问方法入口,便于日志记录,和各种核实。

IDataRepository.cs(提供数据层使用,承继了上边包车型客车IDataServiceRepository,援救传入sql)

  在不增加任何代码的根基上,大家就能够实现下边包车型地铁调用:

Dapper-Extensions源码:

 

实际的选择方发

下图是自家要介绍实现的品类截图:

图片 17

实际也是三层,只是名字不等同而已。

HY.Web

HY.Web.Iservice

HY.Web.Service(服务层,HY.Web.Iservice的落到实处类, 你也能够知晓为专门的学业逻辑层BLL)

HY.Web.DAO(数据访谈层, 你也可知为DAL)

HY.Web.Entity(实体层, 近期只定义了数码实体, 借让你的系统供给给app提供数据, 那么传输的数目要言简意赅,就供给单独定义DTO了。 )

就那顾客表来做个实例吧,表结构如下:(下图是用代码生成器截图效果,能够向来退换数据库的叙说音讯,开采利器。要求的仇敌点这里【CodeBuilder-RazorEngine】)

  1. 其结构轻便,整个封装首要集中Dapper.cs文件中,体量十分的小

IDBSession.cs 对数据访谈对象的概念

  2. 概念接口  IUserInfoRep: IBaseRep

View Code

  同期,对于二个有些有一点点范围的项目以来,化解数据库访问的最便捷做法正是落到实处读写分离,所以,我梦想以此框架能够在一早先在底层就落实了读写分离的支持,以幸免后期再重头对作业代码的豁达退换。  

优点:

=============================

HY.Web.Service

定义BaseService.cs,(能够设置暗中同意的DBsession,方便扩展别的东东)

图片 18View Code

定义Sys_UsersService.cs, 去实现ISys_UsersService。

图片 19View Code

  当前在.NET种类下,开源的ORM框架非常多,如:Entityframework,NHibernate,iBATIS.NET,Dapper等等,各有特色,基于后面小编说的,保障效能的还要,兼顾简单还是能最大程度裁减质量的损耗,并且提供.net standard标准库下的支撑。这里比较之后小编选拔Dapper那么些半自动化的ORM作为仓库储存层的功底框架,选择原因如下:

Dapper-Extensions的缺点:

    那点差非常少大胜其余框架,不需求任何多余的安装,同时基本上你可调用全体原生ADO.NET的功能,sql语句完全本人掌控,却又不供给关怀command的参数赋值,以致结果实体调换等。

接下去配置实体类和DB的照射:

    近期趁着不忙,在考虑二个搭建一个开源的完整项目,至于原因以致全部项目框架前面作品小编再作证。既然要起八个总体的体系,那么数量存款和储蓄访谈就必需,那篇小说作者根本介绍这几个新项目(OSS.Core)中自己对仓库储存层的简短思维和落到实处进程(当前项目还处在搭建阶段),首要聚焦在以下几个地点:

咱俩那边介绍的是DapperExtensions

  1. 数码存储层的须要

  2. ORM框架选用

  3. OSS.Core积攒层设计完成

  4. 调用示例

2、不支持oracle(木有oracle的白话,已经解决)

四. 调用示例

今昔成熟的ORM数不清,这里只介绍Dapper的利用(最最少小编在采取它,已经运用到品种中,小朋侪们反映仍是能够)。

  3. 品质上的便捷

RepositoryBase.cs(IDataRepository的兑现类)

 

缺点:

 

1、开源

  图中得以见到接口层独立于达成部分,笔者将具体育赛事情实人体模型型和接口 单独放在了OSS.Core.DomainMos 类库中,一方面是为着实人体模型型在各模块中的共用,另一方面解耦业务逻辑层(Services)和存款和储蓄层(Reps)之间的重视关系。

2、补助的数据库还蛮多的, Mysql,SqlLite,Sqlserver,Oracle等一体系的数据库。

二. ORM 框架选拔

如此那般就能够在类中 实现 this.Get<DemoEntity>("select * from 德姆o where ID=@ID", new { ID = 1 }); 那样的语法了。

  在此个图里显示一个粗略的办法调用流程,围绕那张图的多少个大旨部分,笔者分别介绍下:

using System;using System.Collections.Generic;using System.Configuration;using System.Data;namespace HY.DataAccess{    public enum DatabaseType    {            SqlServer,        MySql,        Oracle,        DB2    }    public class SqlConnectionFactory    {        public static IDbConnection CreateSqlConnection(DatabaseType dbType, string strKey)        {            IDbConnection connection = null;            string strConn = ConfigurationManager.ConnectionStrings[strKey].ConnectionString;            switch             {                case DatabaseType.SqlServer:                    connection = new System.Data.SqlClient.SqlConnection;                    break;                case DatabaseType.MySql:                    //connection = new MySql.Data.MySqlClient.MySqlConnection;                    //break;                case DatabaseType.Oracle:                    //connection = new Oracle.DataAccess.Client.OracleConnection;                    connection = new System.Data.OracleClient.OracleConnection;                    break;                case DatabaseType.DB2:                    connection = new System.Data.OleDb.OleDbConnection;                    break;            }            return connection;        }    }}

图片 20

using System;using System.Data;namespace HY.DataAccess{    /// <summary>    /// 数据库接口    /// </summary>    public interface IDatabase    {        IDbConnection Connection { get; }        DatabaseType DatabaseType { get; }        string ConnKey { get; }    }    /// <summary>    /// 数据库类对象    /// </summary>    public class Database : IDatabase    {        public IDbConnection Connection { get; private set; }        public DatabaseType DatabaseType { get; private set; }        public string ConnKey { get; set; }        public Database(IDbConnection connection)        {            Connection = connection;        }        public Database(DatabaseType dbType, string connKey)        {            DatabaseType = dbType;            ConnKey = connKey;            Connection = SqlConnectionFactory.CreateSqlConnection(dbType, connKey);        }    }    /// <summary>    /// 数据连接事务的Session接口    /// </summary>    public interface IDBSession : IDisposable    {        string ConnKey { get; }        DatabaseType DatabaseType { get; }        IDbConnection Connection { get; }        IDbTransaction Transaction { get; }        IDbTransaction Begin(IsolationLevel isolation = IsolationLevel.ReadCommitted);        void Commit();        void Rollback();    }}

 

View Code

  同有的时候候,为了保险子类中可以见到参与缓存管理,所以利用了虚方法(virtual)的花样,保险子类能够重写。

各个艺术都重载了业务的操作,日常的数据库操作都匡助。但是每一回试行都需求传递sql,况兼每回都要利用Using,望着不爽啊, 这……

图片 21

    public class DemoEntity    {                public int ID { get; set; }        public string Name { get; set; }    }    [Serializable]    public class DomoEntityORMMapper : ClassMapper<DemoEntity>    {        public DomoEntityORMMapper()        {            base.Table("Demo");            //Map(f => f.UserID).Ignore();//设置忽略            //Map(f => f.Name).Key(KeyType.Identity);//设置主键  (如果主键名称不包含字母“ID”,请设置)                       AutoMap();        }    }

图片 22

View Code

  因为本身盼望以此是全体的演示项目,所在此之前面希望能够协作分歧数据库,因而对外的存款和储蓄访谈都基于接口调用。当然假如你的连串一贯未有切换数据库的急需,作者更提出去掉这一环节,直接在基类中完成单例情势,业务逻辑层直接调用。

ps:已经更新版本了, 参预了对lambda的恢宏,点击这里步向

相关小说:

搭建一套本身实用的.net架构

搭建一套自身实用的.net架构【日志模块-log4net】

搭建一套本人实用的.net架构【ORM-Dapper DapperExtensions】

搭建一套自身实用的.net架构续 【ORM Dapper DapperExtensions Lambda】

搭建一套本身实用的.net架构【CodeBuilder-RazorEngine】

原稿链接:

  b. 有了表明式的强盛之后,就能够获得相应的sql和参数,通过this扩大Connection方法就能够,代码见ConnoctionExtention.cs

下边是带参数的语法

  

using System.Collections.Generic;using System.Data;using System.Linq;using Dapper;using DapperExtensions;using HY.DataAccess;namespace HY.ORM{    public class RepositoryServiceBase : IDataServiceRepository    {        public RepositoryServiceBase()        {        }        public RepositoryServiceBase(IDBSession dbSession)        {            DBSession = dbSession;        }        public IDBSession DBSession { get; private set; }        public void SetDBSession(IDBSession dbSession)        {            DBSession = dbSession;        }        /// <summary>        /// 根据Id获取实体        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="primaryId"></param>        /// <returns></returns>        public T GetById<T>(dynamic primaryId) where T : class        {            return DBSession.Connection.Get<T>(primaryId as object, databaseType: DBSession.DatabaseType);        }        /// <summary>        /// 根据多个Id获取多个实体        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="ids"></param>        /// <returns></returns>        public IEnumerable<T> GetByIds<T>(IList<dynamic> ids) where T : class        {            var tblName = string.Format("dbo.{0}", typeof;            var idsin = string.Join(",", ids.ToArray<dynamic>;            var sql = "SELECT * FROM @table WHERE Id in ";            IEnumerable<T> dataList = DBSession.Connection.Query<T>(sql, new { table = tblName, ids = idsin });            return dataList;        }        /// <summary>        /// 获取全部数据集合        /// </summary>        /// <typeparam name="T"></typeparam>        /// <returns></returns>        public IEnumerable<T> GetAll<T>() where T : class        {            return DBSession.Connection.GetList<T>(databaseType: DBSession.DatabaseType);        }        /// <summary>        /// 统计记录总数        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="predicate"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public int Count<T>(object predicate, bool buffered = false) where T : class        {            return DBSession.Connection.Count<T>(predicate, databaseType: DBSession.DatabaseType);        }        /// <summary>        /// 查询列表数据        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="predicate"></param>        /// <param name="sort"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public IEnumerable<T> GetList<T>(object predicate = null, IList<ISort> sort = null,            bool buffered = false) where T : class        {            return DBSession.Connection.GetList<T>(predicate, sort, null, null, buffered, databaseType: DBSession.DatabaseType);        }        /// <summary>        /// 分页        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="pageIndex"></param>        /// <param name="pageSize"></param>        /// <param name="allRowsCount"></param>        /// <param name="predicate"></param>        /// <param name="sort"></param>        /// <param name="buffered"></param>        /// <returns></returns>        public IEnumerable<T> GetPageList<T>(int pageIndex, int pageSize, out long allRowsCount,            object predicate = null, IList<ISort> sort = null, bool buffered = true) where T : class        {            if (sort == null)            {                sort = new List<ISort>();            }            IEnumerable<T> entityList = DBSession.Connection.GetPage<T>(predicate, sort, pageIndex, pageSize, null, null, buffered, databaseType: DBSession.DatabaseType);            allRowsCount = DBSession.Connection.Count<T>(predicate, databaseType: DBSession.DatabaseType);            return entityList;        }        /// <summary>        /// 插入单条记录        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="entity"></param>        /// <param name="transaction"></param>        /// <returns></returns>        public dynamic Insert<T>(T entity, IDbTransaction transaction = null) where T : class        {            dynamic result = DBSession.Connection.Insert<T>(entity, transaction, databaseType: DBSession.DatabaseType);            return result;        }        /// <summary>        /// 更新单条记录        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="entity"></param>        /// <param name="transaction"></param>        /// <returns></returns>        public bool Update<T>(T entity, IDbTransaction transaction = null) where T : class        {            bool isOk = DBSession.Connection.Update<T>(entity, transaction, databaseType: DBSession.DatabaseType);            return isOk;        }        /// <summary>        /// 删除单条记录        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="primaryId"></param>        /// <param name="transaction"></param>        /// <returns></returns>        public int Delete<T>(dynamic primaryId, IDbTransaction transaction = null) where T : class        {            var entity = GetById<T>(primaryId);            var obj = entity as T;            int isOk = DBSession.Connection.Delete<T>(obj, databaseType: DBSession.DatabaseType);            return isOk;        }        /// <summary>        /// 删除单条记录        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="predicate"></param>        /// <param name="transaction"></param>        /// <returns></returns>         public int DeleteList<T>(object predicate = null, IDbTransaction transaction = null) where T : class        {            return DBSession.Connection.Delete<T>(predicate, transaction, databaseType: DBSession.DatabaseType);        }        /// <summary>        /// 批量插入功能        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="entityList"></param>        /// <param name="transaction"></param>        public bool InsertBatch<T>(IEnumerable<T> entityList, IDbTransaction transaction = null) where T : class        {            bool isOk = false;            foreach (var item in entityList)            {                Insert<T>(item, transaction);            }            isOk = true;            return isOk;        }        /// <summary>        /// 批量更新()        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="entityList"></param>        /// <param name="transaction"></param>        /// <returns></returns>        public bool UpdateBatch<T>(IEnumerable<T> entityList, IDbTransaction transaction = null) where T : class        {            bool isOk = false;            foreach (var item in entityList)            {                Update<T>(item, transaction);            }            isOk = true;            return isOk;        }        /// <summary>        /// 批量删除        /// </summary>        /// <typeparam name="T"></typeparam>        /// <param name="ids"></param>        /// <param name="transaction"></param>        /// <returns></returns>        public bool DeleteBatch<T>(IEnumerable<dynamic> ids, IDbTransaction transaction = null) where T : class        {            bool isOk = false;            foreach (var id in ids)            {                Delete<T>(id, transaction);            }            isOk = true;            return isOk;        }    }}

  能够看见,如若子类中定义了团结的连接串,则以子类自定义为主,不然走暗中同意的再而三消息。

HY.Web.IService

概念接口 ISys_UsersService.cs ,提供给UI访问。

图片 23View Code

HY.Web

1、定义相关的Controller

2、ISys_UsersService iSys_UsersService = new Sys_UsersService(); (那块其实能够选择 IoC, 相关内容且听后续分解)

3、调用接口

图片 24View Code

**下载:**

HY.DataAccess

修改后的DapperExtensions:Dapperextensions.RAR

4、部分代码有个别bug

int xoID=666; //变量主键
 using (IDbConnection conn = GetConnection    {        const string query = "select * from XO where Id=@MyID";        return conn.Query<XOEntity>(query, new { MyID = xoID});                            }

DBHelper 都连任自IDBHelper.cs

图片 25图片 26

图片 27图片 28

Dapper就三个.cs文件,可以停放项目代码中向来编写翻译,也足以直接援引DLL文件。

用作一款ORM太过分轻量级了,依据目的自动生成sql的效能依旧一贫如洗,须要团结来扩大,

上述代码就能够编写翻译成 HY.ORM.DLL文件了。

图片 29图片 30

View Code

在说ORM从前,如故要说一下HY.DataAccess那个模块

图片 31图片 32

1、开源、轻量、小巧、上手轻便。

所以针对Dapper已经有为数不菲老奸巨滑的扩大项目了,Dapper.Rainbow、Dapper.Contrib,DapperExtensions。

下面先简单介绍一下Dapper的为主语法。

图片 33

由此各自的构造函数可能,SetDBSession(Helper.CreateDBSession; 进行数据连接开首化。

Dapper-Extensions的优点

3、须求映射的实体类自个儿0配置,不需求加天性什么的。是经过独立的映射类来管理,能够安装类映射到DB的小名,字段的外号等等。

好呢上边简单介绍下使用Dapper-Extensions的骨干语法(在Dapper-Extensions 的功底上用了Repository方式,代码效果如下)。

下面就能够在 本人业务层传承HY.ORM中的RepositoryServiceBase类 ,数据层承继HY.ORM中的 RepositoryBase类。

那几个模块是对数据访谈提供的八个Helper的作用,里面包括了 各类DB的SqlHelper,分页。

View Code

本文由时时app平台注册网站发布于编程知识,转载请注明出处:基于Dapper二次封装了一个易用的ORM工具类:SqlD

关键词: