ROWID and UROWID

Internally, every database table has a ROWID pseudo column that stores binary values known as rowids. Each rowid shows the storage address of a row. A physical rowid identify a row in an ordinary table. A logical rowid identify a row in an index-organized table. The ROWID datatype can store only physical rowids. Thgough, the UROWID (the universal rowid) datatype can store logical,physical, or foreign (non-Oracle) rowids.

Physical Rowids

The Physical rowids provide a fast access to the particular rows. As long as the row exists, its physical rowid does not change. The Efficient & stable, physical rowids are very useful for selecting a set of rows, operating on the entire set, and then updating a subset. For e.g. , you can compare a UROWID variable with the ROWID pseudocolumn in the WHERE clause of an UPDATE or DELETE statement to identify the newest row fetched from a cursor.

Logical Rowids

The Logical rowids provide the fastest access to the particular rows. The Oracle uses them to construct the secondary indexes on index-organized tables. Having no permanent physical address, a logical rowid can move across the data blocks when new rows are inserted. Though, if the physical location of a row changes, its logical rowid remain valid.

VARCHAR2

You use the VARCHAR2 datatype to store the variable-length character data. How the data is represented internally is totally depend on the database character set. The VARCHAR2 datatype takes a required parameter that specifies the maximum length of up to 32767 bytes. The syntax is as shown below:

VARCHAR2(maximum_length)

You cannot use a constant or variable to specify the highest length; you should use an integer literal in the range 1 to 32767.

The VARCHAR2 datatype includes a trade-off between efficiency and memory use. For a VARCHAR2(>= 2000) variable, the PL/SQL dynamically allocates only enough memory to hold the actual value. Though, for a VARCHAR2 (< 2000) variable, the PL/SQL pre allocates enough memory to hold a maximum-size value.

VARCHAR2 Subtypes

The VARCHAR2 subtypes shown below have the same range of values as their base type. For illustration, the VARCHAR is just another name for VARCHAR2.

STRING

VARCHAR

NLS Character Types

Though the widely used 7- or 8-bit ASCII and EBCDIC character sets are sufficient to represent the Roman alphabet, some Asian languages, like Japanese, contain many characters. These languages need 16 bits (two bytes) to represent each character. How does Oracle deal with such unlike languages?

The Oracle provides National Language Support (NLS) that process single-byte & multi-byte character data and convert between the character sets. It also lets your applications run in various language environments.

NCHAR

You use the NCHAR datatype to store the fixed-length NLS character data. How the data is represented internally, totally depends on the national character set, which might use a fixed-width encoding like US7ASCII or a variable-width encoding like JA16SJIS. The NCHAR datatype takes an optional parameter that specifies the highest length of up to 32767 bytes. The syntax is as shown below:

NCHAR[(maximum_length)]

You cannot use a variable or constant to specify the maximum length; you must use an integer literal in the range 1 to 32767.

NVARCHAR2

You use the NVARCHAR2 datatype to store the variable-length NLS character data. How the data is represented internally totally depends on the national character set, that might use a fixed-width encoding like WE8EBCDIC37C or a variable-width encoding like JA16DBCS. The NVARCHAR2 datatype takes a required parameter which specifies a maximum length of up to 32767 bytes. The syntax is as shown below:

NVARCHAR2(maximum_length)

You cannot use a variable or constant to specify the maximum length; you must use an integer literal in the range 1 to 32767.