mssql

package module
v1.7.1 Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Mar 19, 2024 License: BSD-3-Clause Imports: 39 Imported by: 216

README

Microsoft's official Go MSSQL driver

Go Reference Build status codecov

Install

Requires Go 1.17 or above.

Install with go install github.com/microsoft/go-mssqldb@latest.

Connection Parameters and DSN

The recommended connection string uses a URL format: sqlserver://username:password@host/instance?param1=value&param2=value Other supported formats are listed below.

Common parameters
  • user id - enter the SQL Server Authentication user id or the Windows Authentication user id in the DOMAIN\User format. On Windows, if user id is empty or missing Single-Sign-On is used. The user domain sensitive to the case which is defined in the connection string.
  • password
  • database
  • connection timeout - in seconds (default is 0 for no timeout), set to 0 for no timeout. Recommended to set to 0 and use context to manage query and connection timeouts.
  • dial timeout - in seconds (default is 15 times the number of registered protocols), set to 0 for no timeout.
  • encrypt
    • strict - Data sent between client and server is encrypted E2E using TDS8.
    • disable - Data send between client and server is not encrypted.
    • false/optional/no/0/f - Data sent between client and server is not encrypted beyond the login packet. (Default)
    • true/mandatory/yes/1/t - Data sent between client and server is encrypted.
  • app name - The application name (default is go-mssqldb)
  • authenticator - Can be used to specify use of a registered authentication provider. (e.g. ntlm, winsspi (on windows) or krb5 (on linux))
Connection parameters for ODBC and ADO style connection strings
  • server - host or host\instance (default localhost)
  • port - specifies the host\instance port (default 1433). If instance name is provided but no port, the driver will use SQL Server Browser to discover the port.
Less common parameters
  • keepAlive - in seconds; 0 to disable (default is 30)
  • failoverpartner - host or host\instance (default is no partner).
  • failoverport - used only when there is no instance in failoverpartner (default 1433)
  • packet size - in bytes; 512 to 32767 (default is 4096)
  • log - logging flags (default 0/no logging, 255 for full logging)
    • 1 log errors
    • 2 log messages
    • 4 log rows affected
    • 8 trace sql statements
    • 16 log statement parameters
    • 32 log transaction begin/end
    • 64 additional debug logs
    • 128 log retries
  • TrustServerCertificate
    • false - Server certificate is checked. Default is false if encrypt is specified.
    • true - Server certificate is not checked. Default is true if encrypt is not specified. If trust server certificate is true, driver accepts any certificate presented by the server and any host name in that certificate. In this mode, TLS is susceptible to man-in-the-middle attacks. This should be used only for testing.
  • certificate - The file that contains the public key certificate of the CA that signed the SQL Server certificate. The specified certificate overrides the go platform specific CA certificates. Currently, certificates of PEM type are supported.
  • hostNameInCertificate - Specifies the Common Name (CN) in the server certificate. Default value is the server host.
  • tlsmin - Specifies the minimum TLS version for negotiating encryption with the server. Recognized values are 1.0, 1.1, 1.2, 1.3. If not set to a recognized value the default value for the tls package will be used. The default is currently 1.2.
  • ServerSPN - The kerberos SPN (Service Principal Name) for the server. Default is MSSQLSvc/host:port.
  • Workstation ID - The workstation name (default is the host name)
  • ApplicationIntent - Can be given the value ReadOnly to initiate a read-only connection to an Availability Group listener. The database must be specified when connecting with Application Intent set to ReadOnly.
  • protocol - forces use of a protocol. Make sure the corresponding package is imported.
  • columnencryption or column encryption setting - a boolean value indicating whether Always Encrypted should be enabled on the connection.
  • multisubnetfailover
    • true (Default) Client attempt to connect to all IPs simultaneously.
    • false Client attempts to connect to IPs in serial.
Connection parameters for namedpipe package
  • pipe - If set, no Browser query is made and named pipe used will be \\<host>\pipe\<pipe>
  • protocol can be set to np
  • For a non-URL DSN, the server parameter can be set to the full pipe name like \\host\pipe\sql\query

If no pipe name can be derived from the DSN, connection attempts will first query the SQL Browser service to find the pipe name for the instance.

DNS Resolution through a Custom Dialer

Custom Dialers can be used to resolve DNS if the Connection's Dialer implements the HostDialer interface. This is helpful when the dialer is proxying requests to a different, private network and the DNS record is local to the private network.

Protocol configuration

To force a specific protocol for the connection there two several options:

  1. Prepend the server name in a DSN with the protocol and a colon, like np:host or lpc:host or tcp:host
  2. Set the protocol parameter to the protocol name

msdsn.ProtocolParsers can be reordered to prioritize other protocols ahead of tcp

The admin protocol will not be used for dialing unless the connection string explicitly specifies it. Note SQL Server allows only 1 admin (or DAC) connection to be active at a time.

Kerberos Active Directory authentication outside Windows

To connect with kerberos authentication from a Linux server you can use the optional krb5 package. Imported krb alongside the main driver

package main

import (
    ...
    _ "github.com/microsoft/go-mssqldb"
    _ "github.com/microsoft/go-mssqldb/integratedauth/krb5"
)

func main() {
    ...
}

It will become available for use when the connection string parameter "authenticator=krb5" is used.

The package supports authentication via 3 methods.

  • Keytabs - Specify the username, keytab file, the krb5.conf file, and realm.

    authenticator=krb5;server=DatabaseServerName;database=DBName;user id=MyUserName;krb5-realm=domain.com;krb5-configfile=/etc/krb5.conf;krb5-keytabfile=~/MyUserName.keytab
    
  • Credential Cache - Specify the krb5.conf file path and credential cache file path.

    authenticator=krb5;server=DatabaseServerName;database=DBName;krb5-configfile=/etc/krb5.conf;krb5-credcachefile=~/MyUserNameCachedCreds 
    
  • Raw credentials - Specity krb5.confg, Username, Password and Realm.

    authenticator=krb5;server=DatabaseServerName;database=DBName;user id=MyUserName;password=foo;krb5-realm=comani.com;krb5-configfile=/etc/krb5.conf;
    
Kerberos Parameters
  • authenticator - set this to krb5 to enable kerberos authentication. If this is not present, the default provider would be ntlm for unix and winsspi for windows.
  • krb5-configfile (optional) - path to kerberos configuration file. Defaults to /etc/krb5.conf. Can also be set using KRB5_CONFIG environment variable.
  • krb5-realm (required with keytab and raw credentials) - Domain name for kerberos authentication. Omit this parameter if the realm is part of the user name like username@REALM.
  • krb5-keytabfile - path to Keytab file. Can also be set using environment variable KRB5_KTNAME. If no parameter or environment variable is set, the DefaultClientKeytabName value from the krb5 config file is used.
  • krb5-credcachefile - path to Credential cache. Can also be set using environment variable KRBCCNAME.
  • krb5-dnslookupkdc - Optional parameter in all contexts. Set to lookup KDCs in DNS. Boolean. Default is true.
  • krb5-udppreferencelimit - Optional parameter in all contexts. 1 means to always use tcp. MIT krb5 has a default value of 1465, and it prevents user setting more than 32700. Integer. Default is 1.

For further information on usage:

The connection string can be specified in one of three formats
  1. URL: with sqlserver scheme. username and password appears before the host. Any instance appears as the first segment in the path. All other options are query parameters. Examples:

    • sqlserver://username:password@host/instance?param1=value&param2=value
    • sqlserver://username:password@host:port?param1=value&param2=value
    • sqlserver://sa@localhost/SQLExpress?database=master&connection+timeout=30 // `SQLExpress instance.
    • sqlserver://sa:mypass@localhost?database=master&connection+timeout=30 // username=sa, password=mypass.
    • sqlserver://sa:mypass@localhost:1234?database=master&connection+timeout=30 // port 1234 on localhost.
    • sqlserver://sa:my%7Bpass@somehost?connection+timeout=30 // password is "my{pass" A string of this format can be constructed using the URL type in the net/url package.
    
    query := url.Values{}
    query.Add("app name", "MyAppName")
    
    u := &url.URL{
    	Scheme:   "sqlserver",
    	User:     url.UserPassword(username, password),
    	Host:     fmt.Sprintf("%s:%d", hostname, port),
    	// Path:  instance, // if connecting to an instance instead of a port
    	RawQuery: query.Encode(),
    }
    db, err := sql.Open("sqlserver", u.String())
    
    
  • sqlserver://username@host/instance?krb5-configfile=path/to/file&krb5-credcachefile=/path/to/cache
    • sqlserver://username@host/instance?krb5-configfile=path/to/file&krb5-realm=domain.com&krb5-keytabfile=/path/to/keytabfile
  1. ADO: key=value pairs separated by ;. Values may not contain ;, leading and trailing whitespace is ignored. Examples:

    • server=localhost\\SQLExpress;user id=sa;database=master;app name=MyAppName
    • server=localhost;user id=sa;database=master;app name=MyAppName
    • server=localhost;user id=sa;database=master;app name=MyAppName;krb5-configfile=path/to/file;krb5-credcachefile=path/to/cache;authenticator=krb5
    • server=localhost;user id=sa;database=master;app name=MyAppName;krb5-configfile=path/to/file;krb5-realm=domain.com;krb5-keytabfile=path/to/keytabfile;authenticator=krb5

    ADO strings support synonyms for database, app name, user id, and server

    • server <= addr, address, network address, data source
    • user id <= user, uid
    • database <= initial catalog
    • app name <= application name
  2. ODBC: Prefix with odbc, key=value pairs separated by ;. Allow ; by wrapping values in {}. Examples:

    • odbc:server=localhost\\SQLExpress;user id=sa;database=master;app name=MyAppName
    • odbc:server=localhost;user id=sa;database=master;app name=MyAppName
    • odbc:server=localhost;user id=sa;password={foo;bar} // Value marked with {}, password is "foo;bar"
    • odbc:server=localhost;user id=sa;password={foo{bar} // Value marked with {}, password is "foo{bar"
    • odbc:server=localhost;user id=sa;password={foobar } // Value marked with {}, password is "foobar "
    • odbc:server=localhost;user id=sa;password=foo{bar // Literal {, password is "foo{bar"
    • odbc:server=localhost;user id=sa;password=foo}bar // Literal }, password is "foo}bar"
    • odbc:server=localhost;user id=sa;password={foo{bar} // Literal {, password is "foo{bar"
    • odbc:server=localhost;user id=sa;password={foo}}bar} // Escaped } with}}`, password is "foo}bar"
    • odbc:server=localhost;user id=sa;database=master;app name=MyAppName;krb5-configfile=path/to/file;krb5-credcachefile=path/to/cache;authenticator=krb5
    • odbc:server=localhost;user id=sa;database=master;app name=MyAppName;krb5-configfile=path/to/file;krb5-realm=domain.com;krb5-keytabfile=path/to/keytabfile;authenticator=krb5
Azure Active Directory authentication

Azure Active Directory authentication uses temporary authentication tokens to authenticate. The mssql package does not provide an implementation to obtain tokens: instead, import the azuread package and use driver name azuresql. This driver uses azidentity to acquire tokens using a variety of credential types.

To reduce friction in local development, ActiveDirectoryDefault can authenticate as the user signed into the Azure CLI.

Run the following command to sign into the Azure CLI before running your application using the ActiveDirectoryDefault connection string parameter:

az login

Azure CLI authentication isn't recommended for applications running in Azure. More details are available via the Azure authentication with the Azure Identity module for Go tutorial.

The credential type is determined by the new fedauth connection string parameter.

  • fedauth=ActiveDirectoryServicePrincipal or fedauth=ActiveDirectoryApplication - authenticates using an Azure Active Directory application client ID and client secret or certificate. Implemented using ClientSecretCredential or CertificateCredential
    • clientcertpath=<path to certificate file>;password=<certificate password> or
    • password=<client secret>
    • user id=<application id>[@tenantid] Note the @tenantid component can be omitted if the server's tenant is the same as the application's tenant.
  • fedauth=ActiveDirectoryPassword - authenticates using a user name and password.
    • user id=username@domain
    • password=<password>
    • applicationclientid=<application id> - This guid identifies an Azure Active Directory enterprise application that the AAD admin has approved for accessing Azure SQL database resources in the tenant. This driver does not have an associated application id of its own.
  • fedauth=ActiveDirectoryDefault - authenticates using a chained set of credentials. The chain is built from EnvironmentCredential -> ManagedIdentityCredential->AzureCLICredential. See DefaultAzureCredential docs for instructions on setting up your host environment to use it. Using this option allows you to have the same connection string in a service deployment as on your interactive development machine.
  • fedauth=ActiveDirectoryManagedIdentity or fedauth=ActiveDirectoryMSI - authenticates using a system-assigned or user-assigned Azure Managed Identity.
    • user id=<identity id> - optional id of user-assigned managed identity. If empty, system-assigned managed identity is used.
    • resource id=<resource id> - optional resource id of user-assigned managed identity. If empty, system-assigned managed identity or user id are used (if both user id and resource id are provided, resource id will be used)
  • fedauth=ActiveDirectoryInteractive - authenticates using credentials acquired from an external web browser. Only suitable for use with human interaction.
    • applicationclientid=<application id> - This guid identifies an Azure Active Directory enterprise application that the AAD admin has approved for accessing Azure SQL database resources in the tenant. This driver does not have an associated application id of its own.
  • fedauth=ActiveDirectoryDeviceCode - prints a message to stdout giving the user a URL and code to authenticate. Connection continues after user completes the login separately.
  • fedauth=ActiveDirectoryAzCli - reuses local authentication the user already performed using Azure CLI.

import (
  "database/sql"
  "net/url"

  // Import the Azure AD driver module (also imports the regular driver package)
  "github.com/microsoft/go-mssqldb/azuread"
)

func ConnectWithMSI() (*sql.DB, error) {
  return sql.Open(azuread.DriverName, "sqlserver://azuresql.database.windows.net?database=yourdb&fedauth=ActiveDirectoryMSI")
}

Executing Stored Procedures

To run a stored procedure, set the query text to the procedure name:


var account = "abc"
_, err := db.ExecContext(ctx, "sp_RunMe",
	sql.Named("ID", 123),
	sql.Named("Account", sql.Out{Dest: &account}),
)

Reading Output Parameters from a Stored Procedure with Resultset

To read output parameters from a stored procedure with resultset, make sure you read all the rows before reading the output parameters:


sqltextcreate := `
CREATE PROCEDURE spwithoutputandrows
	@bitparam BIT OUTPUT
AS BEGIN
	SET @bitparam = 1
	SELECT 'Row 1'
END
`
var bitout int64
rows, err := db.QueryContext(ctx, "spwithoutputandrows", sql.Named("bitparam", sql.Out{Dest: &bitout}))
var strrow string
for rows.Next() {
	err = rows.Scan(&strrow)
}
fmt.Printf("bitparam is %d", bitout)

Caveat for local temporary tables

Due to protocol limitations, temporary tables will only be allocated on the connection as a result of executing a query with zero parameters. The following query will, due to the use of a parameter, execute in its own session, and #mytemp will be de-allocated right away:

conn, err := pool.Conn(ctx)
defer conn.Close()
_, err := conn.ExecContext(ctx, "select @p1 as x into #mytemp", 1)
// at this point #mytemp is already dropped again as the session of the ExecContext is over

To work around this, always explicitly create the local temporary table in a query without any parameters. As a special case, the driver will then be able to execute the query directly on the connection-scoped session. The following example works:

conn, err := pool.Conn(ctx)

// Set us up so that temp table is always cleaned up, since conn.Close()
// merely returns conn to pool, rather than actually closing the connection.
defer func() {
	_, _ = conn.ExecContext(ctx, "drop table #mytemp")  // always clean up
	conn.Close() // merely returns conn to pool
}()


// Since we not pass any parameters below, the query will execute on the scope of
// the connection and succeed in creating the table.
_, err := conn.ExecContext(ctx, "create table #mytemp ( x int )")

// #mytemp is now available even if you pass parameters
_, err := conn.ExecContext(ctx, "insert into #mytemp (x) values (@p1)", 1)

Return Status

To get the procedure return status, pass into the parameters a *mssql.ReturnStatus. For example:


var rs mssql.ReturnStatus
_, err := db.ExecContext(ctx, "theproc", &rs)
log.Printf("status=%d", rs)

or

var rs mssql.ReturnStatus
_, err := db.QueryContext(ctx, "theproc", &rs)
for rows.Next() {
	err = rows.Scan(&val)
}
log.Printf("status=%d", rs)

Limitation: ReturnStatus cannot be retrieved using QueryRow.

Parameters

The sqlserver driver uses normal MS SQL Server syntax and expects parameters in the sql query to be in the form of either @Name or @p1 to @pN (ordinal position).


db.QueryContext(ctx, `select * from t where ID = @ID and Name = @p2;`, sql.Named("ID", 6), "Bob")

Parameter Types

To pass specific types to the query parameters, say varchar or date types, you must convert the types to the type before passing in. The following types are supported:

  • string -> nvarchar
  • mssql.VarChar -> varchar
  • time.Time -> datetimeoffset or datetime (TDS version dependent)
  • mssql.DateTime1 -> datetime
  • mssql.DateTimeOffset -> datetimeoffset
  • "github.com/golang-sql/civil".Date -> date
  • "github.com/golang-sql/civil".DateTime -> datetime2
  • "github.com/golang-sql/civil".Time -> time
  • mssql.TVP -> Table Value Parameter (TDS version dependent)

Using an int parameter will send a 4 byte value (int) from a 32bit app and an 8 byte value (bigint) from a 64bit app. To make sure your integer parameter matches the size of the SQL parameter, use the appropriate sized type like int32 or int8.

// If this is passed directly as a parameter, 
// the SQL parameter generated would be nvarchar
name := "Bob"
// If the user_name is defined as varchar,
// it needs to be converted like this:
db.QueryContext(ctx, `select * from t2 where user_name = @p1;`, mssql.VarChar(name))
// Note: Mismatched data types on table and parameter may cause long running queries

Using Always Encrypted

The protocol and cryptography details for AE are detailed elsewhere.

Enablement

To enable AE on a connection, set the ColumnEncryption value to true on a config or pass columnencryption=true in the connection string.

Decryption and encryption won't succeed, however, without also including a decryption key provider. To avoid code size impacts on non-AE applications, key providers are not included by default.

Include the local certificate providers:

 import (
  "github.com/microsoft/go-mssqldb/aecmk/localcert"
 )

You can also instantiate a key provider directly in code and hand it to a Connector instance.

c := mssql.NewConnectorConfig(myconfig)
c.RegisterCekProvider(providerName, MyProviderType{})
Decryption

If the correct key provider is included in your application, decryption of encrypted cells happens automatically with no extra server round trips.

Encryption

Encryption of parameters passed to Exec and Query variants requires an extra round trip per query to fetch the encryption metadata. If the error returned by a query attempt indicates a type mismatch between the parameter and the destination table, most likely your input type is not a strict match for the SQL Server data type of the destination. You may be using a Go string when you need to use one of the driver-specific aliases like VarChar or NVarCharMax.

*** NOTE *** - Currently char and varchar types do not include a collation parameter component so can't be used for inserting encrypted values. https://github.com/microsoft/go-mssqldb/issues/129

Local certificate AE key provider

Key provider configuration is managed separately without any properties in the connection string. The pfx provider exposes its instance as the variable PfxKeyProvider. You can give it passwords for certificates using SetCertificatePassword(pathToCertificate, path). Use an empty string or "*" as the path to use the same password for all certificates.

The MSSQL_CERTIFICATE_STORE provider exposes its instance as the variable WindowsCertificateStoreKeyProvider.

Both providers can be constrained to an allowed list of encryption key paths by appending paths to provider.AllowedLocations.

Azure Key Vault (AZURE_KEY_VAULT) key provider

Import this provider using github.com/microsoft/go-mssqldb/aecmk/akv

Constrain the provider to an allowed list of key vaults by appending vault host strings like "mykeyvault.vault.azure.net" to akv.KeyProvider.AllowedLocations.

Important Notes

  • LastInsertId should not be used with this driver (or SQL Server) due to how the TDS protocol works. Please use the OUTPUT Clause or add a select ID = convert(bigint, SCOPE_IDENTITY()); to the end of your query (ref SCOPE_IDENTITY). This will ensure you are getting the correct ID and will prevent a network round trip.
  • NewConnector may be used with OpenDB.
  • Connector.SessionInitSQL may be set to set any driver specific session settings after the session has been reset. If empty the session will still be reset but use the database defaults in Go1.10+.

Features

  • Can be used with SQL Server 2005 or newer
  • Can be used with Microsoft Azure SQL Database
  • Can be used on all go supported platforms (e.g. Linux, Mac OS X and Windows)
  • Supports new date/time types: date, time, datetime2, datetimeoffset
  • Supports string parameters longer than 8000 characters
  • Supports encryption using SSL/TLS
  • Supports SQL Server and Windows Authentication
  • Supports Single-Sign-On on Windows
  • Supports connections to AlwaysOn Availability Group listeners, including re-direction to read-only replicas.
  • Supports query notifications
  • Supports Kerberos Authentication
  • Supports handling the uniqueidentifier data type with the UniqueIdentifier and NullUniqueIdentifier go types
  • Pluggable Dialer implementations through msdsn.ProtocolParsers and msdsn.ProtocolDialers
  • A namedpipe package to support connections using named pipes (np:) on Windows
  • A sharedmemory package to support connections using shared memory (lpc:) on Windows
  • Dedicated Administrator Connection (DAC) is supported using admin protocol
  • Always Encrypted
    • MSSQL_CERTIFICATE_STORE provider on Windows
    • pfx provider on Linux and Windows

Tests

go test is used for testing. A running instance of MSSQL server is required. Environment variables are used to pass login information.

Example:

    env SQLSERVER_DSN=sqlserver://user:pass@hostname/instance?database=test1 go test

AZURESERVER_DSN environment variable provides the connection string for Azure Active Directory-based authentication. If it's not set the AAD test will be skipped.

Deprecated

These features still exist in the driver, but they are are deprecated.

Query Parameter Token Replace (driver "mssql")

If you use the driver name "mssql" (rather then "sqlserver") the SQL text will be loosly parsed and an attempt to extract identifiers using one of

  • ?
  • ?nnn
  • :nnn
  • $nnn

will be used. This is not recommended with SQL Server. There is at least one existing won't fix issue with the query parsing.

Use the native "@Name" parameters instead with the "sqlserver" driver name.

Known Issues

  • SQL Server 2008 and 2008 R2 engine cannot handle login records when SSL encryption is not disabled. To fix SQL Server 2008 R2 issue, install SQL Server 2008 R2 Service Pack 2. To fix SQL Server 2008 issue, install Microsoft SQL Server 2008 Service Pack 3 and Cumulative update package 3 for SQL Server 2008 SP3. More information: http://support.microsoft.com/kb/2653857

  • Bulk copy does not yet support encrypting column values using Always Encrypted. Tracked in #127

Contributing

This project is a fork of https://github.com/denisenkom/go-mssqldb and welcomes new and previous contributors. For more informaton on contributing to this project, please see Contributing.

For more information on the roadmap for go-mssqldb, project plans are available for viewing and discussion.

Microsoft Open Source Code of Conduct

This project has adopted the Microsoft Open Source Code of Conduct.

Resources:

Documentation

Overview

package mssql implements the TDS protocol used to connect to MS SQL Server (sqlserver) database servers.

This package registers the driver:

sqlserver: uses native "@" parameter placeholder names and does no pre-processing.

If the ordinal position is used for query parameters, identifiers will be named "@p1", "@p2", ... "@pN".

Please refer to the README for the format of the DSN. There are multiple DSN formats accepted: ADO style, ODBC style, and URL style. The following is an example of a URL style DSN:

sqlserver://sa:mypass@localhost:1234?database=master&connection+timeout=30

Index

Examples

Constants

View Source
const (
	CertificateStoreKeyProvider = "MSSQL_CERTIFICATE_STORE"
	CspKeyProvider              = "MSSQL_CSP_PROVIDER"
	CngKeyProvider              = "MSSQL_CNG_STORE"
	AzureKeyVaultKeyProvider    = "AZURE_KEY_VAULT"
	JavaKeyProvider             = "MSSQL_JAVA_KEYSTORE"
	KeyEncryptionAlgorithm      = "RSA_OAEP"
)
View Source
const (
	// FedAuthLibraryLiveIDCompactToken specifies the Microsoft Live ID Compact Token authentication scheme
	FedAuthLibraryLiveIDCompactToken = 0x00

	// FedAuthLibrarySecurityToken specifies a token-based authentication where the token is available
	// without additional information provided during the login sequence.
	FedAuthLibrarySecurityToken = 0x01

	// FedAuthLibraryADAL specifies a token-based authentication where a token is obtained during the
	// login sequence using the server SPN and STS URL provided by the server during login.
	FedAuthLibraryADAL = 0x02

	// FedAuthLibraryReserved is used to indicate that no federated authentication scheme applies.
	FedAuthLibraryReserved = 0x7F
)

Federated authentication library affects the login data structure and message sequence.

View Source
const (
	// FedAuthADALWorkflowPassword uses a username/password to obtain a token from Active Directory
	FedAuthADALWorkflowPassword = 0x01

	// fedAuthADALWorkflowPassword uses the Windows identity to obtain a token from Active Directory
	FedAuthADALWorkflowIntegrated = 0x02

	// FedAuthADALWorkflowMSI uses the managed identity service to obtain a token
	FedAuthADALWorkflowMSI = 0x03

	// FedAuthADALWorkflowNone does not need to obtain token
	FedAuthADALWorkflowNone = 0x04
)

Federated authentication ADAL workflow affects the mechanism used to authenticate.

Variables

View Source
var (
	ErrorEmptyTVPTypeName = errors.New("TypeName must not be empty")
	ErrorTypeSlice        = errors.New("TVP must be slice type")
	ErrorTypeSliceIsEmpty = errors.New("TVP mustn't be null value")
	ErrorSkip             = errors.New("all fields mustn't skip")
	ErrorObjectName       = errors.New("wrong tvp name")
	ErrorWrongTyping      = errors.New("the number of elements in columnStr and tvpFieldIndexes do not align")
)

Functions

func CopyIn

func CopyIn(table string, options BulkOptions, columns ...string) string
Example

This example shows how to perform bulk imports

//go:build go1.10
// +build go1.10

package main

import (
	"database/sql"
	"log"
	"strings"
	"unicode/utf8"

	mssql "github.com/microsoft/go-mssqldb"
)

const (
	createTestTable = `CREATE TABLE test_table(
		id int IDENTITY(1,1) NOT NULL,
		test_nvarchar nvarchar(50) NULL,
		test_varchar varchar(50) NULL,
		test_float float NULL,
		test_datetime2_3 datetime2(3) NULL,
		test_bitn bit NULL,
		test_bigint bigint NOT NULL,
		test_geom geometry NULL,
	CONSTRAINT PK_table_test_id PRIMARY KEY CLUSTERED
	(
		id ASC
	) ON [PRIMARY]);`
	dropTestTable = "IF OBJECT_ID('test_table', 'U') IS NOT NULL DROP TABLE test_table;"
)

// This example shows how to perform bulk imports
func main() {

	connString := makeConnURL().String()

	db, err := sql.Open("sqlserver", connString)
	if err != nil {
		log.Fatal("Open connection failed:", err.Error())
	}
	defer db.Close()

	txn, err := db.Begin()
	if err != nil {
		log.Fatal(err)
	}

	// Create table
	_, err = db.Exec(createTestTable)
	if err != nil {
		log.Fatal(err)
	}
	defer db.Exec(dropTestTable)

	// mssqldb.CopyIn creates string to be consumed by Prepare
	stmt, err := txn.Prepare(mssql.CopyIn("test_table", mssql.BulkOptions{}, "test_varchar", "test_nvarchar", "test_float", "test_bigint"))
	if err != nil {
		log.Fatal(err.Error())
	}

	for i := 0; i < 10; i++ {
		_, err = stmt.Exec(generateString(0, 30), generateStringUnicode(0, 30), i, i)
		if err != nil {
			log.Fatal(err.Error())
		}
	}

	result, err := stmt.Exec()
	if err != nil {
		log.Fatal(err)
	}

	err = stmt.Close()
	if err != nil {
		log.Fatal(err)
	}

	err = txn.Commit()
	if err != nil {
		log.Fatal(err)
	}
	rowCount, _ := result.RowsAffected()
	log.Printf("%d row copied\n", rowCount)
	log.Printf("bye\n")
}

func generateString(x int, n int) string {
	letters := "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"
	b := make([]byte, n)
	for i := range b {
		b[i] = letters[(x+i)%len(letters)]
	}
	return string(b)
}
func generateStringUnicode(x int, n int) string {
	letters := []byte("ab©💾é?ghïjklmnopqЯ☀tuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ")
	b := &strings.Builder{}
	for i := 0; i < n; i++ {
		r, sz := utf8.DecodeRune(letters[x%len(letters):])
		x += sz
		b.WriteRune(r)
	}
	return b.String()
}
Output:

func IsSkipField

func IsSkipField(tvpTagValue string, isTvpValue bool, jsonTagValue string, isJsonTagValue bool) bool

func NewAccessTokenConnector

func NewAccessTokenConnector(dsn string, tokenProvider func() (string, error)) (driver.Connector, error)

NewAccessTokenConnector creates a new connector from a DSN and a token provider. The token provider func will be called when a new connection is requested and should return a valid access token. The returned connector may be used with sql.OpenDB.

func SetContextLogger

func SetContextLogger(ctxLogger ContextLogger)

SetContextLogger sets a ContextLogger for both driver instances ("mssql" and "sqlserver"). Use this to get callbacks from go-mssqldb with additional information and extra details that you can log in the format of your choice. You can set either a ContextLogger or a Logger, but not both. Calling SetContextLogger will overwrite any Logger you set with SetLogger.

func SetLogger

func SetLogger(logger Logger)

SetLogger sets a Logger for both driver instances ("mssql" and "sqlserver"). Use this to have go-msqldb log additional information in a format it picks. You can set either a Logger or a ContextLogger, but not both. Calling SetLogger will overwrite any ContextLogger you set with SetContextLogger.

Types

type Bulk

type Bulk struct {
	Options BulkOptions
	Debug   bool
	// contains filtered or unexported fields
}

func (*Bulk) AddRow

func (b *Bulk) AddRow(row []interface{}) (err error)

AddRow immediately writes the row to the destination table. The arguments are the row values in the order they were specified.

func (*Bulk) Done

func (b *Bulk) Done() (rowcount int64, err error)

type BulkOptions

type BulkOptions struct {
	CheckConstraints  bool
	FireTriggers      bool
	KeepNulls         bool
	KilobytesPerBatch int
	RowsPerBatch      int
	Order             []string
	Tablock           bool
}

type ColumnEncryptionType added in v1.6.0

type ColumnEncryptionType int
var (
	ColumnEncryptionPlainText     ColumnEncryptionType = 0
	ColumnEncryptionDeterministic ColumnEncryptionType = 1
	ColumnEncryptionRandomized    ColumnEncryptionType = 2
)

type Conn

type Conn struct {
	// contains filtered or unexported fields
}

func (*Conn) Begin

func (c *Conn) Begin() (driver.Tx, error)

func (*Conn) BeginTx

func (c *Conn) BeginTx(ctx context.Context, opts driver.TxOptions) (driver.Tx, error)

BeginTx satisfies ConnBeginTx.

func (*Conn) CheckNamedValue

func (c *Conn) CheckNamedValue(nv *driver.NamedValue) error

func (*Conn) Close

func (c *Conn) Close() error

func (*Conn) Commit

func (c *Conn) Commit() error

func (*Conn) CreateBulk

func (cn *Conn) CreateBulk(table string, columns []string) (_ *Bulk)

func (*Conn) CreateBulkContext

func (cn *Conn) CreateBulkContext(ctx context.Context, table string, columns []string) (_ *Bulk)

func (*Conn) IsValid

func (c *Conn) IsValid() bool

IsValid satisfies the driver.Validator interface.

func (*Conn) Ping

func (c *Conn) Ping(ctx context.Context) error

Ping is used to check if the remote server is available and satisfies the Pinger interface.

func (*Conn) Prepare

func (c *Conn) Prepare(query string) (driver.Stmt, error)

func (*Conn) PrepareContext

func (c *Conn) PrepareContext(ctx context.Context, query string) (driver.Stmt, error)

func (*Conn) ResetSession

func (c *Conn) ResetSession(ctx context.Context) error

func (*Conn) Rollback

func (c *Conn) Rollback() error

type Connector

type Connector struct {

	// SessionInitSQL is executed after marking a given session to be reset.
	// When not present, the next query will still reset the session to the
	// database defaults.
	//
	// When present the connection will immediately mark the session to
	// be reset, then execute the SessionInitSQL text to setup the session
	// that may be different from the base database defaults.
	//
	// For Example, the application relies on the following defaults
	// but is not allowed to set them at the database system level.
	//
	//    SET XACT_ABORT ON;
	//    SET TEXTSIZE -1;
	//    SET ANSI_NULLS ON;
	//    SET LOCK_TIMEOUT 10000;
	//
	// SessionInitSQL should not attempt to manually call sp_reset_connection.
	// This will happen at the TDS layer.
	//
	// SessionInitSQL is optional. The session will be reset even if
	// SessionInitSQL is empty.
	SessionInitSQL string

	// Dialer sets a custom dialer for all network operations, except DNS resolution unless
	// the dialer implements the HostDialer.
	//
	// If Dialer is not set, normal net dialers are used.
	Dialer Dialer
	// contains filtered or unexported fields
}

Connector holds the parsed DSN and is ready to make a new connection at any time.

In the future, settings that cannot be passed through a string DSN may be set directly on the connector.

Example

This example shows the usage of Connector type

//go:build go1.10
// +build go1.10

package main

import (
	"context"
	"database/sql"
	"flag"
	"fmt"
	"log"
	"net/url"
	"strconv"

	mssql "github.com/microsoft/go-mssqldb"
)

var (
	debug         = flag.Bool("debug", false, "enable debugging")
	password      = flag.String("password", "", "the database password")
	port     *int = flag.Int("port", 1433, "the database port")
	server        = flag.String("server", ".", "the database server")
	user          = flag.String("user", "", "the database user")
)

const (
	createTableSql      = "CREATE TABLE TestAnsiNull (bitcol bit, charcol char(1));"
	dropTableSql        = "IF OBJECT_ID('TestAnsiNull', 'U') IS NOT NULL DROP TABLE TestAnsiNull;"
	insertQuery1        = "INSERT INTO TestAnsiNull VALUES (0, NULL);"
	insertQuery2        = "INSERT INTO TestAnsiNull VALUES (1, 'a');"
	selectNullFilter    = "SELECT bitcol FROM TestAnsiNull WHERE charcol = NULL;"
	selectNotNullFilter = "SELECT bitcol FROM TestAnsiNull WHERE charcol <> NULL;"
)

func makeConnURL() *url.URL {
	flag.Parse()
	if *debug {
		fmt.Printf(" password:%s\n", *password)
		fmt.Printf(" port:%d\n", *port)
		fmt.Printf(" server:%s\n", *server)
		fmt.Printf(" user:%s\n", *user)
	}

	params, err := mssql.GetConnParams()
	if err == nil && params != nil {
		return params.URL()
	}
	var userInfo *url.Userinfo
	if *user != "" {
		userInfo = url.UserPassword(*user, *password)
	}
	return &url.URL{
		Scheme: "sqlserver",
		Host:   *server + ":" + strconv.Itoa(*port),
		User:   userInfo,
	}
}

// This example shows the usage of Connector type
func main() {

	connString := makeConnURL().String()
	if *debug {
		fmt.Printf(" connString:%s\n", connString)
	}

	// Create a new connector object by calling NewConnector
	connector, err := mssql.NewConnector(connString)
	if err != nil {
		log.Println(err)
		return
	}

	// Use SessionInitSql to set any options that cannot be set with the dsn string
	// With ANSI_NULLS set to ON, compare NULL data with = NULL or <> NULL will return 0 rows
	connector.SessionInitSQL = "SET ANSI_NULLS ON"

	// Pass connector to sql.OpenDB to get a sql.DB object
	db := sql.OpenDB(connector)
	defer db.Close()

	// Create and populate table
	_, err = db.Exec(createTableSql)
	if err != nil {
		log.Println(err)
		return
	}
	defer db.Exec(dropTableSql)
	_, err = db.Exec(insertQuery1)
	if err != nil {
		log.Println(err)
		return
	}
	_, err = db.Exec(insertQuery2)
	if err != nil {
		log.Println(err)
		return
	}

	var bitval bool
	ctx, cancel := context.WithCancel(context.Background())
	defer cancel()

	// (*Row) Scan should return ErrNoRows since ANSI_NULLS is set to ON
	err = db.QueryRowContext(ctx, selectNullFilter).Scan(&bitval)
	if err != nil {
		if err.Error() != "sql: no rows in result set" {
			log.Println(err)
			return
		}
	} else {
		log.Println("Expects an ErrNoRows error. No error is returned")
		return
	}

	// (*Row) Scan should return ErrNoRows since ANSI_NULLS is set to ON
	err = db.QueryRowContext(ctx, selectNotNullFilter).Scan(&bitval)
	if err != nil {
		if err.Error() != "sql: no rows in result set" {
			log.Println(err)
			return
		}
	} else {
		log.Println("Expects an ErrNoRows error. No error is returned")
		return
	}

	// Set ANSI_NULLS to OFF
	connector.SessionInitSQL = "SET ANSI_NULLS OFF"

	// (*Row) Scan should copy data to bitval
	err = db.QueryRowContext(ctx, selectNullFilter).Scan(&bitval)
	if err != nil {
		log.Println(err)
		return
	}
	if bitval != false {
		log.Println("Incorrect value retrieved.")
		return
	}

	// (*Row) Scan should copy data to bitval
	err = db.QueryRowContext(ctx, selectNotNullFilter).Scan(&bitval)
	if err != nil {
		log.Println(err)
		return
	}
	if bitval != true {
		log.Println("Incorrect value retrieved.")
		return
	}
}
Output:

func NewActiveDirectoryTokenConnector

func NewActiveDirectoryTokenConnector(config msdsn.Config, adalWorkflow byte, tokenProvider func(ctx context.Context, serverSPN, stsURL string) (string, error)) (*Connector, error)

newADALTokenConnector creates a new connector from a Config and a Active Directory token provider. Token provider implementations are called during federated authentication login sequences where the server provides a service principal name and security token service endpoint that should be used to obtain the token. Implementations should contact the security token service specified and obtain the appropriate token, or return an error to indicate why a token is not available.

The returned connector may be used with sql.OpenDB.

func NewConnector

func NewConnector(dsn string) (*Connector, error)

NewConnector creates a new connector from a DSN. The returned connector may be used with sql.OpenDB.

func NewConnectorConfig

func NewConnectorConfig(config msdsn.Config) *Connector

NewConnectorConfig creates a new Connector for a DSN Config struct. The returned connector may be used with sql.OpenDB.

func NewConnectorWithAccessTokenProvider added in v0.15.0

func NewConnectorWithAccessTokenProvider(dsn string, tokenProvider func(ctx context.Context) (string, error)) (*Connector, error)

NewConnectorWithAccessTokenProvider creates a new connector from a DSN using the given access token provider. The returned connector may be used with sql.OpenDB.

func NewSecurityTokenConnector

func NewSecurityTokenConnector(config msdsn.Config, tokenProvider func(ctx context.Context) (string, error)) (*Connector, error)

newSecurityTokenConnector creates a new connector from a Config and a token provider. When invoked, token provider implementations should contact the security token service specified and obtain the appropriate token, or return an error to indicate why a token is not available. The returned connector may be used with sql.OpenDB.

func (*Connector) Connect

func (c *Connector) Connect(ctx context.Context) (driver.Conn, error)

Connect to the server and return a TDS connection.

func (*Connector) Driver

func (c *Connector) Driver() driver.Driver

Driver underlying the Connector.

func (*Connector) RegisterCekProvider added in v1.6.0

func (c *Connector) RegisterCekProvider(name string, provider aecmk.ColumnEncryptionKeyProvider)

RegisterCekProvider associates the given provider with the named key store. If an entry of the given name already exists, that entry is overwritten

type ContextLogger

type ContextLogger interface {
	Log(ctx context.Context, category msdsn.Log, msg string)
}

ContextLogger is an interface that provides more information than Logger and lets you log messages yourself. This gives you more information to log (e.g. trace IDs in the context), more control over the logging activity (e.g. log it, trace it, or log and trace it, depending on the class of message), and lets you log in exactly the format you want.

type DataValue

type DataValue interface{}

type DateTime1

type DateTime1 time.Time

DateTime1 encodes parameters to original DateTime SQL types.

type DateTimeOffset

type DateTimeOffset time.Time

DateTimeOffset encodes parameters to DateTimeOffset, preserving the UTC offset.

Example

This example shows how to insert and retrieve date and time types data

//go:build go1.10
// +build go1.10

package main

import (
	"database/sql"
	"fmt"
	"log"
	"time"

	"github.com/golang-sql/civil"
	mssql "github.com/microsoft/go-mssqldb"
)

// This example shows how to insert and retrieve date and time types data
func main() {

	connString := makeConnURL().String()

	db, err := sql.Open("sqlserver", connString)
	if err != nil {
		log.Fatal("Open connection failed:", err.Error())
	}
	defer db.Close()

	insertDateTime(db)
	retrieveDateTime(db)
	retrieveDateTimeOutParam(db)
}

func insertDateTime(db *sql.DB) {
	_, err := db.Exec("CREATE TABLE datetimeTable (timeCol TIME, dateCol DATE, smalldatetimeCol SMALLDATETIME, datetimeCol DATETIME, datetime2Col DATETIME2, datetimeoffsetCol DATETIMEOFFSET)")
	if err != nil {
		log.Fatal(err)
	}
	stmt, err := db.Prepare("INSERT INTO datetimeTable VALUES(@p1, @p2, @p3, @p4, @p5, @p6)")
	if err != nil {
		log.Fatal(err)
	}
	defer stmt.Close()

	tin, err := time.Parse(time.RFC3339, "2006-01-02T22:04:05.787-07:00")
	if err != nil {
		log.Fatal(err)
	}
	var timeCol civil.Time = civil.TimeOf(tin)
	var dateCol civil.Date = civil.DateOf(tin)
	var smalldatetimeCol string = "2006-01-02 22:04:00"
	var datetimeCol mssql.DateTime1 = mssql.DateTime1(tin)
	var datetime2Col civil.DateTime = civil.DateTimeOf(tin)
	var datetimeoffsetCol mssql.DateTimeOffset = mssql.DateTimeOffset(tin)
	_, err = stmt.Exec(timeCol, dateCol, smalldatetimeCol, datetimeCol, datetime2Col, datetimeoffsetCol)
	if err != nil {
		log.Fatal(err)
	}
}

func retrieveDateTime(db *sql.DB) {
	rows, err := db.Query("SELECT timeCol, dateCol, smalldatetimeCol, datetimeCol, datetime2Col, datetimeoffsetCol FROM datetimeTable")
	if err != nil {
		log.Fatal(err)
	}
	defer rows.Close()
	var c1, c2, c3, c4, c5, c6 time.Time
	for rows.Next() {
		err = rows.Scan(&c1, &c2, &c3, &c4, &c5, &c6)
		if err != nil {
			log.Fatal(err)
		}
		fmt.Printf("c1: %+v; c2: %+v; c3: %+v; c4: %+v; c5: %+v; c6: %+v;\n", c1, c2, c3, c4, c5, c6)
	}
}

func retrieveDateTimeOutParam(db *sql.DB) {
	CreateProcSql := `
	CREATE PROCEDURE OutDatetimeProc
		@timeOutParam TIME OUTPUT,
		@dateOutParam DATE OUTPUT,
		@smalldatetimeOutParam SMALLDATETIME OUTPUT,
		@datetimeOutParam DATETIME OUTPUT,
		@datetime2OutParam DATETIME2 OUTPUT,
		@datetimeoffsetOutParam DATETIMEOFFSET OUTPUT
	AS
		SET NOCOUNT ON
		SET @timeOutParam = '22:04:05.7870015'
		SET @dateOutParam = '2006-01-02'
		SET @smalldatetimeOutParam = '2006-01-02 22:04:00'
		SET @datetimeOutParam = '2006-01-02 22:04:05.787'
		SET @datetime2OutParam = '2006-01-02 22:04:05.7870015'
		SET @datetimeoffsetOutParam = '2006-01-02 22:04:05.7870015 -07:00'`
	_, err := db.Exec(CreateProcSql)
	if err != nil {
		log.Fatal(err)
	}
	var (
		timeOutParam, datetime2OutParam, datetimeoffsetOutParam mssql.DateTimeOffset
		dateOutParam, datetimeOutParam                          mssql.DateTime1
		smalldatetimeOutParam                                   string
	)
	_, err = db.Exec("OutDatetimeProc",
		sql.Named("timeOutParam", sql.Out{Dest: &timeOutParam}),
		sql.Named("dateOutParam", sql.Out{Dest: &dateOutParam}),
		sql.Named("smalldatetimeOutParam", sql.Out{Dest: &smalldatetimeOutParam}),
		sql.Named("datetimeOutParam", sql.Out{Dest: &datetimeOutParam}),
		sql.Named("datetime2OutParam", sql.Out{Dest: &datetime2OutParam}),
		sql.Named("datetimeoffsetOutParam", sql.Out{Dest: &datetimeoffsetOutParam}))
	if err != nil {
		log.Fatal(err)
	}
	fmt.Printf("timeOutParam: %+v; dateOutParam: %+v; smalldatetimeOutParam: %s; datetimeOutParam: %+v; datetime2OutParam: %+v; datetimeoffsetOutParam: %+v;\n", time.Time(timeOutParam), time.Time(dateOutParam), smalldatetimeOutParam, time.Time(datetimeOutParam), time.Time(datetime2OutParam), time.Time(datetimeoffsetOutParam))
}
Output:

type Dialer

type Dialer interface {
	DialContext(ctx context.Context, network string, addr string) (net.Conn, error)
}

type Driver

type Driver struct {
	// contains filtered or unexported fields
}

func (*Driver) Open

func (d *Driver) Open(dsn string) (driver.Conn, error)

func (*Driver) OpenConnection

func (d *Driver) OpenConnection(dsn string) (*Conn, error)

func (*Driver) OpenConnector

func (d *Driver) OpenConnector(dsn string) (*Connector, error)

OpenConnector opens a new connector. Useful to dial with a context.

func (*Driver) SetContextLogger

func (d *Driver) SetContextLogger(ctxLogger ContextLogger)

SetContextLogger sets a ContextLogger for the driver instance on which you call it. Use this to get callbacks from go-mssqldb with additional information and extra details that you can log in the format of your choice. You can set either a ContextLogger or a Logger, but not both. Calling SetContextLogger will overwrite any Logger you set with SetLogger.

func (*Driver) SetLogger

func (d *Driver) SetLogger(logger Logger)

SetLogger sets a Logger for the driver instance on which you call it. Use this to have go-msqldb log additional information in a format it picks. You can set either a Logger or a ContextLogger, but not both. Calling SetLogger will overwrite any ContextLogger you set with SetContextLogger.

type Error

type Error struct {
	Number     int32
	State      uint8
	Class      uint8
	Message    string
	ServerName string
	ProcName   string
	LineNo     int32
	// All lists all errors that were received from first to last.
	// This includes the last one, which is described in the other members.
	All []Error
}

Error represents an SQL Server error. This type includes methods for reading the contents of the struct, which allows calling programs to check for specific error conditions without having to import this package directly.

func (Error) Error

func (e Error) Error() string

func (Error) SQLErrorClass

func (e Error) SQLErrorClass() uint8

func (Error) SQLErrorLineNo

func (e Error) SQLErrorLineNo() int32

func (Error) SQLErrorMessage

func (e Error) SQLErrorMessage() string
Example
package main

import (
	"fmt"
)

func main() {
	// call a function that might return a mssql error
	err := callUsingMSSQL()

	type SQLError interface {
		SQLErrorNumber() int32
		SQLErrorMessage() string
	}

	if sqlError, ok := err.(SQLError); ok {
		if sqlError.SQLErrorNumber() == 1205 {
			fmt.Println("deadlock error", sqlError.SQLErrorMessage())
		}
	}
}

func callUsingMSSQL() error {
	return nil
}
Output:

func (Error) SQLErrorNumber

func (e Error) SQLErrorNumber() int32

SQLErrorNumber returns the SQL Server error number.

Example
package main

import (
	"fmt"
)

func main() {
	// call a function that might return a mssql error
	err := callUsingMSSQL()

	type ErrorWithNumber interface {
		SQLErrorNumber() int32
	}

	if errorWithNumber, ok := err.(ErrorWithNumber); ok {
		if errorWithNumber.SQLErrorNumber() == 1205 {
			fmt.Println("deadlock error")
		}
	}
}

func callUsingMSSQL() error {
	return nil
}
Output:

func (Error) SQLErrorProcName

func (e Error) SQLErrorProcName() string

func (Error) SQLErrorServerName

func (e Error) SQLErrorServerName() string

func (Error) SQLErrorState

func (e Error) SQLErrorState() uint8

func (Error) String added in v0.14.0

func (e Error) String() string

type HostDialer added in v1.4.0

type HostDialer interface {
	Dialer
	HostName() string
}

HostDialer should be used if the dialer is proxying requests to a different network and DNS should be resolved in that other network

type Logger

type Logger interface {
	Printf(format string, v ...interface{})
	Println(v ...interface{})
}

Logger is an interface you can implement to have the go-msqldb driver automatically log detailed information on your behalf

type MssqlBulk

type MssqlBulk = Bulk // Deprecated: users should transition to the new name when possible.

type MssqlBulkOptions

type MssqlBulkOptions = BulkOptions // Deprecated: users should transition to the new name when possible.

type MssqlConn

type MssqlConn = Conn // Deprecated: users should transition to the new name when possible.

type MssqlDriver

type MssqlDriver = Driver // Deprecated: users should transition to the new name when possible.

type MssqlProtocolDialer added in v0.18.0

type MssqlProtocolDialer interface {
	// DialSqlConnection creates a net.Conn from a Connector based on the Config
	DialSqlConnection(ctx context.Context, c *Connector, p *msdsn.Config) (conn net.Conn, err error)
}

type MssqlResult

type MssqlResult = Result // Deprecated: users should transition to the new name when possible.

type MssqlRows

type MssqlRows = Rows // Deprecated: users should transition to the new name when possible.

type MssqlStmt

type MssqlStmt = Stmt // Deprecated: users should transition to the new name when possible.

type NChar added in v1.6.0

type NChar string

NChar is used to encode a string parameter as NChar instead of a sized NVarChar

type NVarCharMax

type NVarCharMax string

NVarCharMax is used to encode a string parameter as NVarChar(max) instead of a sized NVarChar

type NullUniqueIdentifier added in v1.7.0

type NullUniqueIdentifier struct {
	UUID  UniqueIdentifier
	Valid bool // Valid is true if UUID is not NULL
}

func (NullUniqueIdentifier) MarshalText added in v1.7.0

func (n NullUniqueIdentifier) MarshalText() (text []byte, err error)

func (*NullUniqueIdentifier) Scan added in v1.7.0

func (n *NullUniqueIdentifier) Scan(v interface{}) error

func (NullUniqueIdentifier) String added in v1.7.0

func (n NullUniqueIdentifier) String() string

func (*NullUniqueIdentifier) UnmarshalJSON added in v1.7.0

func (n *NullUniqueIdentifier) UnmarshalJSON(b []byte) error

func (NullUniqueIdentifier) Value added in v1.7.0

func (n NullUniqueIdentifier) Value() (driver.Value, error)

type RWCBuffer added in v1.6.0

type RWCBuffer struct {
	// contains filtered or unexported fields
}

func (RWCBuffer) Close added in v1.6.0

func (R RWCBuffer) Close() error

func (RWCBuffer) Read added in v1.6.0

func (R RWCBuffer) Read(p []byte) (n int, err error)

func (RWCBuffer) Write added in v1.6.0

func (R RWCBuffer) Write(p []byte) (n int, err error)

type Result

type Result struct {
	// contains filtered or unexported fields
}

func (*Result) LastInsertId

func (r *Result) LastInsertId() (int64, error)

func (*Result) RowsAffected

func (r *Result) RowsAffected() (int64, error)

type RetryableError

type RetryableError struct {
	// contains filtered or unexported fields
}

RetryableError is returned when an error was caused by a bad connection at the start of a query and can be safely retried using database/sql's automatic retry logic.

In many cases database/sql's retry logic will transparently handle this error, the retried call will return successfully, and you won't even see this error. However, you may see this error if the retry logic cannot successfully handle the error. In that case you can get the underlying error by calling this error's UnWrap function.

func (RetryableError) Error

func (r RetryableError) Error() string

func (RetryableError) Is

func (r RetryableError) Is(err error) bool

func (RetryableError) Unwrap

func (r RetryableError) Unwrap() error

type ReturnStatus

type ReturnStatus int32

ReturnStatus may be used to return the return value from a proc.

var rs mssql.ReturnStatus
_, err := db.Exec("theproc", &rs)
log.Printf("return status = %d", rs)

type Rows

type Rows struct {
	// contains filtered or unexported fields
}

Rows represents the non-experimental data/sql model for Query and QueryContext

Example (Usingmessages)

This example shows the usage of sqlexp/Messages

//go:build go1.10
// +build go1.10

package main

import (
	"context"
	"database/sql"
	"fmt"
	"log"

	"github.com/golang-sql/sqlexp"
	mssql "github.com/microsoft/go-mssqldb"
)

const (
	msgQuery = `select 'name' as Name
PRINT N'This is a message'
select 199
RAISERROR (N'Testing!' , 11, 1)
select 300
`
)

// This example shows the usage of sqlexp/Messages
func main() {

	connString := makeConnURL().String()

	// Create a new connector object by calling NewConnector
	connector, err := mssql.NewConnector(connString)
	if err != nil {
		log.Println(err)
		return
	}

	// Pass connector to sql.OpenDB to get a sql.DB object
	db := sql.OpenDB(connector)
	defer db.Close()
	retmsg := &sqlexp.ReturnMessage{}
	ctx := context.Background()
	rows, err := db.QueryContext(ctx, msgQuery, retmsg)
	if err != nil {
		log.Fatalf("QueryContext failed: %v", err)
	}
	active := true
	for active {
		msg := retmsg.Message(ctx)
		switch m := msg.(type) {
		case sqlexp.MsgNotice:
			fmt.Println(m.Message)
		case sqlexp.MsgNext:
			inresult := true
			for inresult {
				inresult = rows.Next()
				if inresult {
					cols, err := rows.Columns()
					if err != nil {
						log.Fatalf("Columns failed: %v", err)
					}
					fmt.Println(cols)
					var d interface{}
					if err = rows.Scan(&d); err == nil {
						fmt.Println(d)
					}
				}
			}
		case sqlexp.MsgNextResultSet:
			active = rows.NextResultSet()
		case sqlexp.MsgError:
			fmt.Println("Error:", m.Error)
		case sqlexp.MsgRowsAffected:
			fmt.Println("Rows affected:", m.Count)
		}
	}
}
Output:

func (*Rows) Close

func (rc *Rows) Close() error

func (*Rows) ColumnTypeDatabaseTypeName

func (r *Rows) ColumnTypeDatabaseTypeName(index int) string

RowsColumnTypeDatabaseTypeName may be implemented by Rows. It should return the database system type name without the length. Type names should be uppercase. Examples of returned types: "VARCHAR", "NVARCHAR", "VARCHAR2", "CHAR", "TEXT", "DECIMAL", "SMALLINT", "INT", "BIGINT", "BOOL", "[]BIGINT", "JSONB", "XML", "TIMESTAMP".

func (*Rows) ColumnTypeLength

func (r *Rows) ColumnTypeLength(index int) (int64, bool)

RowsColumnTypeLength may be implemented by Rows. It should return the length of the column type if the column is a variable length type. If the column is not a variable length type ok should return false. If length is not limited other than system limits, it should return math.MaxInt64. The following are examples of returned values for various types:

TEXT          (math.MaxInt64, true)
varchar(10)   (10, true)
nvarchar(10)  (10, true)
decimal       (0, false)
int           (0, false)
bytea(30)     (30, true)

func (*Rows) ColumnTypeNullable

func (r *Rows) ColumnTypeNullable(index int) (nullable, ok bool)

The nullable value should be true if it is known the column may be null, or false if the column is known to be not nullable. If the column nullability is unknown, ok should be false.

func (*Rows) ColumnTypePrecisionScale

func (r *Rows) ColumnTypePrecisionScale(index int) (int64, int64, bool)

It should return the precision and scale for decimal types. If not applicable, ok should be false. The following are examples of returned values for various types:

decimal(38, 4)    (38, 4, true)
int               (0, 0, false)
decimal           (math.MaxInt64, math.MaxInt64, true)

func (*Rows) ColumnTypeScanType

func (r *Rows) ColumnTypeScanType(index int) reflect.Type

It should return the value type that can be used to scan types into. For example, the database column type "bigint" this should return "reflect.TypeOf(int64(0))".

func (*Rows) Columns

func (rc *Rows) Columns() (res []string)

func (*Rows) HasNextResultSet

func (rc *Rows) HasNextResultSet() bool

func (*Rows) Next

func (rc *Rows) Next(dest []driver.Value) error

func (*Rows) NextResultSet

func (rc *Rows) NextResultSet() error

type Rowsq

type Rowsq struct {
	// contains filtered or unexported fields
}

Rowsq implements the sqlexp messages model for Query and QueryContext Theory: We could also implement the non-experimental model this way

func (*Rowsq) Close

func (rc *Rowsq) Close() error

func (*Rowsq) ColumnTypeDatabaseTypeName

func (r *Rowsq) ColumnTypeDatabaseTypeName(index int) string

RowsColumnTypeDatabaseTypeName may be implemented by Rows. It should return the database system type name without the length. Type names should be uppercase. Examples of returned types: "VARCHAR", "NVARCHAR", "VARCHAR2", "CHAR", "TEXT", "DECIMAL", "SMALLINT", "INT", "BIGINT", "BOOL", "[]BIGINT", "JSONB", "XML", "TIMESTAMP".

func (*Rowsq) ColumnTypeLength

func (r *Rowsq) ColumnTypeLength(index int) (int64, bool)

RowsColumnTypeLength may be implemented by Rows. It should return the length of the column type if the column is a variable length type. If the column is not a variable length type ok should return false. If length is not limited other than system limits, it should return math.MaxInt64. The following are examples of returned values for various types:

TEXT          (math.MaxInt64, true)
varchar(10)   (10, true)
nvarchar(10)  (10, true)
decimal       (0, false)
int           (0, false)
bytea(30)     (30, true)

func (*Rowsq) ColumnTypeNullable

func (r *Rowsq) ColumnTypeNullable(index int) (nullable, ok bool)

The nullable value should be true if it is known the column may be null, or false if the column is known to be not nullable. If the column nullability is unknown, ok should be false.

func (*Rowsq) ColumnTypePrecisionScale

func (r *Rowsq) ColumnTypePrecisionScale(index int) (int64, int64, bool)

It should return the precision and scale for decimal types. If not applicable, ok should be false. The following are examples of returned values for various types:

decimal(38, 4)    (38, 4, true)
int               (0, 0, false)
decimal           (math.MaxInt64, math.MaxInt64, true)

func (*Rowsq) ColumnTypeScanType

func (r *Rowsq) ColumnTypeScanType(index int) reflect.Type

It should return the value type that can be used to scan types into. For example, the database column type "bigint" this should return "reflect.TypeOf(int64(0))".

func (*Rowsq) Columns

func (rc *Rowsq) Columns() (res []string)

ProcessSingleResponse queues MsgNext for every columns token. data/sql calls Columns during the app's call to Next.

func (*Rowsq) HasNextResultSet

func (rc *Rowsq) HasNextResultSet() bool

In Message Queue mode, we always claim another resultset could be on the way to avoid Rows being closed prematurely

func (*Rowsq) Next

func (rc *Rowsq) Next(dest []driver.Value) error

func (*Rowsq) NextResultSet

func (rc *Rowsq) NextResultSet() error

Scans to the end of the current statement being processed Note that the caller may not have read all the rows in the prior set

type ServerError

type ServerError struct {
	// contains filtered or unexported fields
}

ServerError is returned when the server got a fatal error that aborts the process and severs the connection.

To get the errors returned before the process was aborted, unwrap this error or call errors.As with a pointer to an mssql.Error variable.

func (ServerError) Error

func (e ServerError) Error() string

func (ServerError) Unwrap

func (e ServerError) Unwrap() error

type Stmt

type Stmt struct {
	// contains filtered or unexported fields
}

func (*Stmt) Close

func (s *Stmt) Close() error

func (*Stmt) Exec

func (s *Stmt) Exec(args []driver.Value) (driver.Result, error)

func (*Stmt) ExecContext

func (s *Stmt) ExecContext(ctx context.Context, args []driver.NamedValue) (driver.Result, error)

func (*Stmt) NumInput

func (s *Stmt) NumInput() int

func (*Stmt) Query

func (s *Stmt) Query(args []driver.Value) (driver.Rows, error)

func (*Stmt) QueryContext

func (s *Stmt) QueryContext(ctx context.Context, args []driver.NamedValue) (driver.Rows, error)

func (*Stmt) SetQueryNotification

func (s *Stmt) SetQueryNotification(id, options string, timeout time.Duration)

type StreamError

type StreamError struct {
	InnerError error
}

func (StreamError) Error

func (e StreamError) Error() string

type TSQLQuoter added in v1.6.0

type TSQLQuoter struct {
}

TSQLQuoter implements sqlexp.Quoter

func (TSQLQuoter) ID added in v1.6.0

func (TSQLQuoter) ID(name string) string

ID quotes identifiers such as schema, table, or column names. This implementation handles multi-part names.

func (TSQLQuoter) Value added in v1.6.0

func (TSQLQuoter) Value(v interface{}) string

Value quotes database values such as string or []byte types as strings that are suitable and safe to embed in SQL text. The returned value of a string will include all surrounding quotes.

If a value type is not supported it must panic.

type TVP

type TVP struct {
	//TypeName mustn't be default value
	TypeName string
	//Value must be the slice, mustn't be nil
	Value interface{}
}

TVP is driver type, which allows supporting Table Valued Parameters (TVP) in SQL Server

Example

This example shows how to use tvp type

const (
	createTable = "CREATE TABLE Location (Name VARCHAR(50), CostRate INT, Availability BIT, ModifiedDate DATETIME2)"

	createTVP = `CREATE TYPE LocationTableType AS TABLE
		(LocationName VARCHAR(50),
		CostRate INT)`

	dropTVP = "IF type_id('LocationTableType') IS NOT NULL DROP TYPE LocationTableType"

	createProc = `CREATE PROCEDURE dbo.usp_InsertProductionLocation
		@TVP LocationTableType READONLY
		AS
		SET NOCOUNT ON
		INSERT INTO Location
		(
			Name,
			CostRate,
			Availability,
			ModifiedDate)
			SELECT *, 0,GETDATE()
			FROM @TVP`

	dropProc = "IF OBJECT_ID('dbo.usp_InsertProductionLocation', 'P') IS NOT NULL DROP PROCEDURE dbo.usp_InsertProductionLocation"

	execTvp = "exec dbo.usp_InsertProductionLocation @TVP;"
)
type LocationTableTvp struct {
	LocationName    string
	LocationCountry string `tvp:"-"`
	CostRate        int64
	Currency        string `json:"-"`
}

connString := makeConnURL().String()

db, err := sql.Open("sqlserver", connString)
if err != nil {
	log.Fatal("Open connection failed:", err.Error())
}
defer db.Close()

_, err = db.Exec(createTable)
if err != nil {
	log.Fatal(err)
}
_, err = db.Exec(createTVP)
if err != nil {
	log.Fatal(err)
}
defer db.Exec(dropTVP)
_, err = db.Exec(createProc)
if err != nil {
	log.Fatal(err)
}
defer db.Exec(dropProc)

locationTableTypeData := []LocationTableTvp{
	{
		LocationName:    "Alberta",
		LocationCountry: "Canada",
		CostRate:        0,
		Currency:        "CAD",
	},
	{
		LocationName:    "British Columbia",
		LocationCountry: "Canada",
		CostRate:        1,
		Currency:        "CAD",
	},
}

tvpType := mssql.TVP{
	TypeName: "LocationTableType",
	Value:    locationTableTypeData,
}

_, err = db.Exec(execTvp, sql.Named("TVP", tvpType))
if err != nil {
	log.Fatal(err)
} else {
	for _, locationData := range locationTableTypeData {
		fmt.Printf("Data for location %s, %s has been inserted.\n", locationData.LocationName, locationData.LocationCountry)
	}
}
Output:

type UniqueIdentifier

type UniqueIdentifier [16]byte

func (UniqueIdentifier) MarshalText

func (u UniqueIdentifier) MarshalText() (text []byte, err error)

MarshalText converts Uniqueidentifier to bytes corresponding to the stringified hexadecimal representation of the Uniqueidentifier e.g., "AAAAAAAA-AAAA-AAAA-AAAA-AAAAAAAAAAAA" -> [65 65 65 65 65 65 65 65 45 65 65 65 65 45 65 65 65 65 45 65 65 65 65 65 65 65 65 65 65 65 65]

func (*UniqueIdentifier) Scan

func (u *UniqueIdentifier) Scan(v interface{}) error

func (UniqueIdentifier) String

func (u UniqueIdentifier) String() string

func (*UniqueIdentifier) UnmarshalJSON added in v1.4.0

func (u *UniqueIdentifier) UnmarshalJSON(b []byte) error

Unmarshals a string representation of a UniqueIndentifier to bytes "01234567-89AB-CDEF-0123-456789ABCDEF" -> [48, 49, 50, 51, 52, 53, 54, 55, 45, 56, 57, 65, 66, 45, 67, 68, 69, 70, 45, 48, 49, 50, 51, 45, 52, 53, 54, 55, 56, 57, 65, 66, 67, 68, 69, 70]

func (UniqueIdentifier) Value

func (u UniqueIdentifier) Value() (driver.Value, error)

type VarChar

type VarChar string

VarChar is used to encode a string parameter as VarChar instead of a sized NVarChar

type VarCharMax

type VarCharMax string

VarCharMax is used to encode a string parameter as VarChar(max) instead of a sized NVarChar

Directories

Path Synopsis
akv
bach splits a single script containing multiple batches separated by a keyword into multiple scripts.
bach splits a single script containing multiple batches separated by a keyword into multiple scripts.
examples
tvp
krb5
Package krb5 implements the integratedauth.IntegratedAuthenticator interface in order to provide kerberos/active directory (Windows) based authentication.
Package krb5 implements the integratedauth.IntegratedAuthenticator interface in order to provide kerberos/active directory (Windows) based authentication.
internal
cp
gopkg.in/natefinch/npipe.v2
Package npipe provides a pure Go wrapper around Windows named pipes.
Package npipe provides a pure Go wrapper around Windows named pipes.
np
querytext
Package querytext is the old query parser and parameter substitute process.
Package querytext is the old query parser and parameter substitute process.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL