Chapter III data type (II)

The third chapter data type （ Two ）

SQL System data type mapping

In the above table is DDL and IRIS The syntax of the data type expression is SQL.SystemDataTypes Default mapping for configuration . For the system data type and user data type provided , A separate mapping table is available .

To view and modify the current data type mapping , Please go to the administration portal , Choose system management 、 To configure 、SQL And object settings 、 System DDL mapping .

understand DDL Data type mapping

Change the data type from DDL Mapping to IRIS when , General parameters and function parameters follow the following rules ：

• Conventional parameters - These in DDL Data types and IRIS The data type is marked with %# Format logo . for example ：

     VARCHAR(%1)

Mapping to ：

     %String(MAXLEN=%1)

therefore ,DDL The data type is ：

     VARCHAR(10)

Mapping to ：

     %String(MAXLEN=10)

• Function parameter — When DDL The parameters in the data type must undergo some conversion before they can be put into IRIS When the data type is , Use these parameters . An example of this would be DDL The numeric precision and scale parameters of the data type are converted to IRIS Data type MAXVAL、MINVAL and SCALE Parameters . for example ：

     DECIMAL(%1,%2)

Mapping to ：

     %Numeric(MAXVAL=<|'$$maxval^%apiSQL(%1,%2)'|>,
              MINVAL=<|'$$minval^%apiSQL(%1,%2)'|>,
              SCALE=%2)

DDL data type DECIMAL Using parameters Precision (%1) and Scale (%2), but IRIS data type %Numeric No precision parameters . therefore , To put DECIMAL Convert to %Numeric, Must be Precision The parameter is converted to the appropriate %Numeric Parameters , under these circumstances , By way of IRIS Function format 、maxval and minval be applied to DECIMAL Provided parameters . special <|'xxx'|> grammar （ As shown above ） instructions DDL The processor performs parameter replacement , Then call the function with the supplied value . <|'xxx'|> The expression is then replaced by the value returned by the function call .

Consider this example with actual values , There may be an accuracy of 4 position 、 The number of decimal places is 2 Of DECIMAL data type ：

     DECIMAL(4,2)

Mapping to ：

     %Numeric(MAXVAL=<|'$$maxval^%apiSQL(4,2)'|>,
              MINVAL=<|'$$minval^%apiSQL(4,2)'|>,
              SCALE=2)

The calculation for the ：

     %Numeric(MAXVAL=99.99,MINVAL=-99.99,SCALE=2)

• Additional parameters —— Data type classes can be defined and cannot be used DDL Additional data definition parameters for data type definitions . These include data validation operations , For example, an enumerated list of allowed data values 、 Allowed data value pattern matching and exceeding MAXLEN Automatic truncation of maximum length data value .

Data type priority

When an operation can return multiple different values , And these values may have different data types ,IRIS Assign the return value to the data type with the highest priority . for example ,NUMERIC The data type can contain all possible INTEGER Data type value , but INTEGER The data type cannot contain all possible NUMERIC Data type value . therefore NUMERIC Higher priority （ More inclusive ）.

for example , If CASE Statement has a data type of INTEGER Possible result values for , And a data type of NUMERIC Possible result values for , No matter which of the two cases is adopted , The actual result is always NUMERIC type .

The priority of data types is as follows , From the highest （ Including the most ） To the lowest ：

LONGVARBINARY
LONGVARCHAR
VARBINARY
VARCHAR
GUID
TIMESTAMP
DOUBLE
NUMERIC
BIGINT
INTEGER
DATE
TIME
SMALLINT
TINYINT
BIT

Normalization and verification

%Library.DataType Superclasses contain classes of specific data types . These data type classes provide Normalize() Method to normalize the input value to a data type format , And provide IsValid() Method to determine whether the input value is valid for the data type , And various mode conversion methods , for example LogicalToDisplay( ) and DisplayToLogical().

The following example shows %TimeStamp Data type Normalize() Method ：

/// d ##class(PHA.TEST.SQLFunction).NormalizeValidate()
ClassMethod NormalizeValidate()
{
    
	s indate = 64701
	s tsdate = ##class(%Library.TimeStamp).Normalize(indate)
	w "%TimeStamp date: ",tsdate
}

DHC-APP>d ##class(PHA.TEST.SQLFunction).NormalizeValidate()
%TimeStamp date: 2018-02-22 00:00:00

/// d ##class(PHA.TEST.SQLFunction).NormalizeValidate2()
ClassMethod NormalizeValidate2()
{
	s indate = "2018-2-22"
	s tsdate = ##class(%Library.TimeStamp).Normalize(indate)
	w "%TimeStamp date: ",tsdate
}

DHC-APP> d ##class(PHA.TEST.SQLFunction).NormalizeValidate2()
%TimeStamp date: 2018-02-22 00:00:00

The following example shows %TimeStamp Data type IsValid() Method ：

/// d ##class(PHA.TEST.SQLFunction).NormalizeValidate3()
ClassMethod NormalizeValidate3()
{
    
	s datestr = "July 4, 2018"
	s stat = ##class(%Library.TimeStamp).IsValid(datestr)
	if stat = 1 {
    
		w datestr," is a valid %TimeStamp",! 
	} else {
    
		w datestr," is not a valid %TimeStamp",!
	}
}

DHC-APP> d ##class(PHA.TEST.SQLFunction).NormalizeValidate3()
July 4, 2018 is not a valid %TimeStamp

/// d ##class(PHA.TEST.SQLFunction).NormalizeValidate4()
ClassMethod NormalizeValidate4()
{
    
	s leapdate = "2016-02-29 00:00:00"
	s noleap = "2018-02-29 00:00:00"
	s stat = ##class(%Library.TimeStamp).IsValid(leapdate)
	if stat = 1 {
    
		w leapdate," is a valid %TimeStamp",! 
	} else {
    
		w leapdate," is not a valid %TimeStamp",!
	}
	s stat = ##class(%Library.TimeStamp).IsValid(noleap)
	if stat = 1 {
    
		w noleap," is a valid %TimeStamp",! 
	} else {
    
		w noleap," is not a valid %TimeStamp",!
	}
}

DHC-APP> d ##class(PHA.TEST.SQLFunction).NormalizeValidate4()
2016-02-29 00:00:00 is a valid %TimeStamp
2018-02-29 00:00:00 is not a valid %TimeStamp