>任务：将 JSON 响应插入多个 mssql 表中
问题：这项工作需要很长时间。
现状：
我从 API 收到以下内容响应：
[{
"About": "",
"Name": "Employee_1",
"EmployeeId": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
"Address1": "Empstreet 4, 3. floor",
"Phone": "90909090",
"Birthdate": "1980-01-21",
"City": "Empcity",
"Zip": "8989",
"Competencies": [{
"Name": "Tablet",
"ShortCode": "TAB",
"Status": false
}, {
"Name": "Flex",
"ShortCode": "FLX",
"Status": false
}],
"CustomColumns": [{
"Description": "",
"Name": "",
"Type": "CheckBox",
"Value": "False"
}],
"CustomRoles": [{
"CustomRoleModels": [{
"IsAdminRole": false,
"IsEmployeeRole": false,
"IsPlannerRole": true,
"Name": "Shop Manager"
}],
"DepartmentId": "49664884-8382-4712-9953-fb36ca13fd1b"
}],
"HistoricalWages": [{
"StartDate": "2000-01-01",
"Wage": 120
}],
"NotAvailable": [{
"From": "12:00",
"To": "05:00",
"WeekDayIndex": 1
}, {
"From": "12:00",
"To": "04:00",
"WeekDayIndex": 0
}, {
"From": "12:00",
"To": "03:00",
"WeekDayIndex": 4
}, {
"From": "12:00",
"To": "05:00",
"WeekDayIndex": 3
}],
"Notes": [{
"Author": "4a408a58-f9d6-4898-80a5-c3ef5667d2c6",
"Comment": "",
"Date": "2016-03-17"
}],
"Roles": [{
"Admin": false,
"DepartmentId": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
"Employee": true,
"Planner": false
}]
}, {
"About": "",
"Name": "Employee_2",
"EmployeeId": "aaaaaaaa-bbbb-cccc-dddd-ffffffffffff",
"Address1": "Empstreet 5, 3. floor",
"Phone": "80808080",
"Birthdate": "1981-01-21",
"City": "Empcity",
"Zip": "8987",
"Competencies": [{
"Name": "Tablet",
"ShortCode": "TAB",
"Status": false
}, {
"Name": "Flex",
"ShortCode": "FLX",
"Status": false
}],
"CustomColumns": [{
"Description": "",
"Name": "",
"Type": "CheckBox",
"Value": "False"
}],
"CustomRoles": [{
"CustomRoleModels": [{
"IsAdminRole": false,
"IsEmployeeRole": false,
"IsPlannerRole": true,
"Name": "Shop Manager"
}],
"DepartmentId": "49664884-8382-4712-9953-fb36ca13fd1b"
}],
"HistoricalWages": [{
"StartDate": "2000-01-01",
"Wage": 120
}],
"NotAvailable": [{
"From": "12:00",
"To": "05:00",
"WeekDayIndex": 1
}, {
"From": "12:00",
"To": "04:00",
"WeekDayIndex": 0
}, {
"From": "12:00",
"To": "03:00",
"WeekDayIndex": 4
}, {
"From": "12:00",
"To": "05:00",
"WeekDayIndex": 3
}],
"Notes": [{
"Author": "4a408a58-f9d6-4898-80a5-c3ef5667d2c6",
"Comment": "",
"Date": "2016-03-17"
}],
"Roles": [{
"Admin": false,
"DepartmentId": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
"Employee": true,
"Planner": false
}]
}]
由此我制作了以下课程
public class Competency
{
public string Name { get; set; }
public string ShortCode { get; set; }
public bool Status { get; set; }
}
public class CustomColumn
{
public string Description { get; set; }
public string Name { get; set; }
public string Type { get; set; }
public string Value { get; set; }
}
public class CustomRoleModel
{
public bool IsAdminRole { get; set; }
public bool IsEmployeeRole { get; set; }
public bool IsPlannerRole { get; set; }
public string Name { get; set; }
}
public class CustomRole
{
public List<CustomRoleModel> CustomRoleModels { get; set; }
public string DepartmentId { get; set; }
}
public class HistoricalWage
{
public string StartDate { get; set; }
public int Wage { get; set; }
}
public class NotAvailable
{
public string From { get; set; }
public string To { get; set; }
public int WeekDayIndex { get; set; }
}
public class Note
{
public string Author { get; set; }
public string Comment { get; set; }
public string Date { get; set; }
}
public class Role
{
public bool Admin { get; set; }
public string DepartmentId { get; set; }
public bool Employee { get; set; }
public bool Planner { get; set; }
}
public class Employees
{
public string About { get; set; }
public string Name { get; set; }
public string EmployeeId { get; set; }
public string Address1 { get; set; }
public string Phone { get; set; }
public string Birthdate { get; set; }
public string City { get; set; }
public string Zip { get; set; }
public List<Competency> Competencies { get; set; }
public List<CustomColumn> CustomColumns { get; set; }
public List<CustomRole> CustomRoles { get; set; }
public List<HistoricalWage> HistoricalWages { get; set; }
public List<NotAvailable> NotAvailable { get; set; }
public List<Note> Notes { get; set; }
public List<Role> Roles { get; set; }
}
JSON 内容存储在名为"json"的变量中，并反序列化为上述类类，如下所示：
List<Employees> employees = JsonConvert.DeserializeObject<List<Employees>>(json);
使用下面新创建的对象调用函数
private static void InsertDataToEmployees(List<Employees> allEmployees)
{
string connectionString = ConfigurationManager.ConnectionStrings["msEmployees"].ConnectionString;
List<Employees> records = allEmployees;
string insert_employee = "INSERT INTO dbo.dtEmployees (Name, EmployeeId, Address1, Phone, Birthdate, City, Zip) " +
" VALUES (@Name, @EmployeeId, @Address1, @Phone, @Birthdate, @City, @Zip)";
string insert_competencies = "INSERT INTO dbo.dtCompetencies (EmployeeId, Name, ShortCode, Status) " +
" VALUES (@EmployeeId, @Name, @ShortCode, @Status)";
string insert_available = "INSERT INTO dbo.dtAvailable (EmployeeId, [From], [To], WeekDayIndex) " +
" VALUES (@EmployeeId, @From, @To, @WeekDayIndex)";
string insert_roles = "INSERT INTO dbo.dtRoles (EmployeeId, Admin, DepartmentId, Employee, Planner) " +
" VALUES (@EmployeeId, @Admin, @DepartmentId, @Employee, @Planner)";
string insert_wages = "INSERT INTO dbo.dtWages (EmployeeId, StartDate, Wage) " +
" VALUES (@EmployeeId, @StartDate, @Wage)";
using (SqlConnection conn = new SqlConnection(connectionString))
{
conn.Open();
foreach (var employee in allEmployees)
{
string empID = employee.EmployeeId;
using (SqlCommand cmd_employees = new SqlCommand(insert_employee, conn))
{
cmd_employees.Parameters.AddWithValue("@Name", (object)employee.Name ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@EmployeeId", (object)employee.EmployeeId ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@Address1", (object)employee.Address1 ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@Phone", (object)employee.Phone ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@Birthdate", (object)employee.Birthdate ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@Zip", (object)employee.Zip ?? DBNull.Value);
cmd_employees.Parameters.AddWithValue("@City", (object)employee.City ?? DBNull.Value);
cmd_employees.ExecuteNonQuery();
cmd_employees.Parameters.Clear();
}
foreach (var c in employee.Competencies)
{
using (SqlCommand cmd_competencies = new SqlCommand(insert_competencies, conn))
{
cmd_competencies.Parameters.AddWithValue("@EmployeeId", (object)empID ?? DBNull.Value);
cmd_competencies.Parameters.AddWithValue("@Name", (object)c.Name ?? DBNull.Value);
cmd_competencies.Parameters.AddWithValue("@ShortCode", (object)c.ShortCode ?? DBNull.Value);
cmd_competencies.Parameters.AddWithValue("@Status", (object)c.Status ?? DBNull.Value);
cmd_competencies.ExecuteNonQuery();
cmd_competencies.Parameters.Clear();
}
}
foreach (var a in employee.NotAvailable)
{
using (SqlCommand cmd_available = new SqlCommand(insert_available, conn))
{
cmd_available.Parameters.AddWithValue("@EmployeeId", (object)empID ?? DBNull.Value);
cmd_available.Parameters.AddWithValue("@From", (object)a.From ?? DBNull.Value);
cmd_available.Parameters.AddWithValue("@To", (object)a.To ?? DBNull.Value);
cmd_available.Parameters.AddWithValue("@WeekDayIndex", (object)a.WeekDayIndex ?? DBNull.Value);
cmd_available.ExecuteNonQuery();
cmd_available.Parameters.Clear();
}
}
foreach (var r in employee.Roles)
{
using (SqlCommand cmd_roles = new SqlCommand(insert_roles, conn))
{
cmd_roles.Parameters.AddWithValue("@EmployeeId", (object)empID ?? DBNull.Value);
cmd_roles.Parameters.AddWithValue("@Admin", (object)r.Admin ?? DBNull.Value);
cmd_roles.Parameters.AddWithValue("@DepartmentId", (object)r.DepartmentId ?? DBNull.Value);
cmd_roles.Parameters.AddWithValue("@Employee", (object)r.Employee ?? DBNull.Value);
cmd_roles.Parameters.AddWithValue("@Planner", (object)r.Planner ?? DBNull.Value);
cmd_roles.ExecuteNonQuery();
cmd_roles.Parameters.Clear();
}
}
foreach (var w in employee.HistoricalWages)
{
using (SqlCommand cmd_wages = new SqlCommand(insert_wages, conn))
{
cmd_wages.Parameters.AddWithValue("@EmployeeId", (object)empID ?? DBNull.Value);
cmd_wages.Parameters.AddWithValue("@StartDate", (object)w.StartDate ?? DBNull.Value);
cmd_wages.Parameters.AddWithValue("@Wage", (object)w.Wage ?? DBNull.Value);
cmd_wages.ExecuteNonQuery();
cmd_wages.Parameters.Clear();
}
}
}
conn.Close();
}
}
上面的函数可以工作并将数据传送到SQL服务器，但实际上需要很长时间。
所以我的问题是，由于我绝不是这里的专家，那么完成 JSON 到 MSSQL 传输的最有效方法（就时间而言）是什么。
延迟是由于代码每行执行一个命令。考虑到每一行都很小，几乎所有时间都浪费在网络往返和单个操作的日志记录上。
您可以改用 SqlBulkCopy 将所有行作为记录流发送，这些记录流将使用最少日志记录插入到目标表中。
SqlBulkCopy 需要一个 DataTable 或 DbDataReader。无需创建自己的 DataTable 并对所有行进行双重缓冲，或者编写自己的数据读取器，您可以使用 FastMember 包从集合中创建 IDataReader：
using(var bcp = new SqlBulkCopy(connection))
using(var reader = ObjectReader.Create(allEmployees, "Name", "EmployeeId", "Address1", "Phone", "Birthdate", "City", "Zip"))
{
bcp.DestinationTableName = "dtEmployees";
bcp.WriteToServer(reader);
}
如果属性名称和表字段不匹配，则必须使用源名称和目标名称填充 ColumnMappings 属性。
您必须使用 LINQ 平展嵌套集合才能获得可与 SqlBulkCopy 一起使用的集合，例如：
var allCompetencies = from employee in allEmployees
from competency in employee.Competencies
select new { employee.EmpoyeeID,
competency.Name,
competency.ShortCode,
competency.Status};
using(var cmpBcp = new SqlBulkCopy(connection))
using(var var cmpReader=ObjectReader.Create(allCompetencies,"EmployeeID","Name","ShortCode","Status"))
{
cmpBcp.DestinationTableName = "dtCompetencies";
cmpBcp.WriteToServer(cmpReader);
}