Column Characteristics
When a column is added to a database table, you set several parameters that define the characteristics of that column. These include the name of the column, its data type, its length, and a default value. In this section, you will learn all about data types, what they are and how you should use them. In addition, you will see how empty or null columns are treated by the SQL processor.
Data Types
In a database, every column, variable, expression, and parameter has a related data type associated with it. A data type is an attribute that specifies the type of data (integer, string, date, and so on) that object can contain. All the data that a specific column holds must be of the same data type. A data type also determines how the data for a particular column is accessed, indexed, and physically stored on the server.
Tables 3.1 and 3.2 describe each of the available data types in Microsoft SQL Server 2000 and provide an example of what each might be used for. If the data you are working with is of different lengths, such as names, addresses, and other text, you should use variable-length data types. Fixed-length data types are best used for data, such as phone numbers, Social Security numbers, and ZIP Codes.
| Long Name | Syntax | Example | Description |
|---|---|---|---|
| Variable | varchar(6) | "John" | Variable-length character fields are best for character most strings. |
| Character | char(6) | "John" | Fixed-length character fields are best for most strings. |
| National variable characters. | nvarchar(6) | "John" | Variable-length Unicode data with a maximum length of 4,000 characters |
| Datetime | datetime | Jan 1, 2000 12:15:00.000 pm | Datetime fields are used for precise storage of dates and times. Datetimes can range from Jan 1, 1753 to Dec 31, 9999. Values outside this range must be stored as character. |
| Small datetime | smalldatetime | Jan 2, 2000 12:15pm | Small datetimes are half the size of datetimes. They use increments of one minute and represent dates from Jan 1, 1900 to Jun 6, 2079. |
| Precise decimal | decimal(4,2) or numeric (4,2) | 13.22 | Decimal/numeric data types store fractional numerics precisely. The first parameter specifies how many digits are allowed in the field. The second parameter specifies how many digits may come after the decimal. In this example, I could represent numbers from –99.99 to 99.99. |
| Big Floating point | float(15) | 64023.0134 | floating point numbers are not guaranteed to be stored precisely. SQL Server rounds up numbers that binary math can't handle. Floats take a parameter specifying the total number of digits. |
| Little float | real(7) | 16.3452 | Half the size of a float; the same rules apply. |
| Integer | int | 683423 | Integers are four bytes wide and store numbers between plus or minus two billion. |
| Small integer | smallint | 12331 | Small integers are half the size of integers, ranging from –32,768 through 32,767. |
| Tiny integer | tinyint | 5 | Tiny integers are half again the size of small integers, a single byte, and may not be negative. Values run from 0 to 255. Perfect for an age column. |
| Bit | bit | 1 | Bits are the smallest data type available today. They are one bit in size, one-eighth of a byte. Bits may not be null and can have a value of 0 or 1. This is the actual language of all computers. |
| Binary | binary | 0x00223FE2… | Fixed-length binary data with a maximum length of 8,000 bytes. |
| Money | money | $753.1132 | Money types range from +/- 922 trillion. Money types store four digits to the right of the decimal and are stored as fixed-point integers. |
| Small money | smallmoney | $32.50 | Small money can handle about +/ $214,000, with four digits to the right of the decimal. Half the size of Money. |
| Text | text | "We the people…" | Text fields can be up to 2GB in size. Text fields are treated at Binary Large Objects (BLOBs) and are subject to a great many limitations They cannot be used in an ORDER BY, indexed, or grouped, and handling the inside an application program takes some extra work. (BLOBs will be discussed on Day 21, "Handling BLOBs in T-SQL." ) |
| Image | image | 0x00223FE2… | Image data can be used to store any type of binary data, including images (gif, jpg, and so on), executables, or anything else you can store on your disk drive. Images are also BLOBs, subject to the same limitations. |
| Long Name | Use | Example | Description |
|---|---|---|---|
| Big integer | bigint | 983422348 | A large integer that can hold a number +/- 2 raised to the 63rd power. Twice as large as an integer. |
| Sql_variant | sql_variant | A data type that stores values of other supported data types, except text, ntext, and sql_variant. You can use this to hold data from any other data type without having to know the data type in advance. | |
| Table | table | This is a special data type that can be used to store a result set for later processing. It is primarily used for temporary storage of a set of rows. |
Data Type Precedence
When two expressions of different data types are combined using one or more operators or functions, the data type precedence rules specify which data type is converted to the other. The data type with the lower precedence is converted to the data type with the higher precedence. If the conversion is not a supported implicit conversion, an error is returned. If both expressions have the same data type, the resulting object has the same data type. The order of precedence for the data types are shown in Table 3.3.
| Precedence Number | Data Type |
|---|---|
| 1 | sql_variant |
| 2 | datetime |
| 3 | smalldatetime |
| 4 | float |
| 5 | real |
| 6 | decimal |
| 7 | money |
| 8 | smallmoney |
| 9 | bigint |
| 10 | int |
| 11 | smallint |
| 12 | tinyint |
| 13 | bit |
| 14 | ntext |
| 15 | image |
| 16 | timestamp |
| 17 | uniqueidentifier |
| 18 | nvarchar |
| 19 | nchar |
| 20 | varchar |
| 21 | char |
| 22 | varbinary |
| 23 | binary |
An example of the precedence can be seen when you add two unlike numbers together as shown here:
Select quantity * price as Sale_Total from "Order Details"
The quantity column is an integer data type, whereas the price column is a money data type. When this calculation is performed, the result would be a money data type.
Note
Don't worry about how to use the multiplication operator, or what the as Sale_Total means. We will cover these issues later in this lesson.
Using Null Data
When a column is empty, it is treated differently than when a column is blank or zero. These might sound the same to you, but to the computer, they are very different. A blank is an actual character that takes a position in the column. Of course, zero is its own explanation. A null means that the column actually contains nothing. To see how a null is displayed, try executing the following SQL statement in the Query Analyzer:
use pubs select title_id, advance from titles order by title_id
Results:
title_id advance -------- --------------------- BU1032 5000.0000 BU1111 5000.0000 BU2075 10125.0000 BU7832 5000.0000 MC2222 .0000 MC3021 15000.0000 MC3026 NULL PC1035 7000.0000 PC8888 8000.0000 PC9999 NULL … TC4203 4000.0000 TC7777 8000.0000 (18 row(s) affected)
You should see that null values are displayed using the string NULL. However, most standard comparisons will not recognize the null value. The next example shows you what happens when I add a where clause to the SELECT statement. I am looking for all title IDs where the advance amount is less than $5,000. You might think that 'null'or 'empty'are identical, but they're not. Although you will be covering functions later in this lesson, I want to cover one in this section.
The isnull() function permits a way to include null values in aggregate calculations. The function requires two arguments and its syntax is shown here:
Isnull(<expression>, <value>)
The first argument is the expression on which the calculation will be performed; usually this is a column from the database. The second argument is the value that will replace the null value for display or calculation purposes. If the expression contains a null, the second parameter is returned by this function. If the expression is not null, the value in the expression is returned.
The following SQL query shows the effect of the isnull() function on the titles table in the pubs database.
use pubs select title_id, price, isnull(price, $45) from titles order by price
The output of this query shows which prices were null:
title_id price -------- --------------------- --------------------- MC3026 NULL 45.0000 PC9999 NULL 45.0000 MC3021 2.9900 2.9900 BU2075 2.9900 2.9900 PS2106 7.0000 7.0000 … PS3333 19.9900 19.9900 PC8888 20.0000 20.0000 TC3218 20.9500 20.9500 PS1372 21.5900 21.5900 PC1035 22.9500 22.9500 (18 row(s) affected)
As you can see in the output, there are two titles in the table that have null in their price column. When the isnull() function evaluates those columns, it returns $45 for the third column in the result set. For the other columns, it simply returns the value of the column.
You can use the isnull() function inside an aggregate function. If you want to know the average price of a book, you would ask for the avg(price) on the titles table. However, the two books that have null for a price would be excluded from the calculation. Suppose that you know that those books will be priced at $29 each. You could use the isnull() function to include those books in the calculation. The example shows you the query without the isnull() function and then with the isnull() function.
select avg(price) from titles
Results:
--------------------- 14.7662 (1 row(s) affected) Warning: Null value is eliminated by an aggregate or other SET operation.
As you can see, the server displays a warning message informing you that there were null values found and they were not used.
select avg(isnull(price, $29)) from titles
Results:
--------------------- 16.3477 (1 row(s) affected)
In the second version of the query, you can see that the average returned is different because the two books with nulls were included in the calculation. The isnull() function doesn't change the value of the row in the table; it only assumes a value for the purposes of a single query.