Example of tables within a table - sql, PL-SQL Programming

Example of Tables within a Table - SQL

Example: Obtaining C_ER from COURSE and EXAM_MARK

SELECT CourseId, CAST (TABLE (SELECT DISTINCT StudentId, Mark FROM EXAM_MARK AS EM WHERE EM.CourseId = C.CourseId)

AS ROW (StudentId SID, Mark INTEGER) MULTISET)

AS ExamResult FROM COURSE AS C

Explanation

  • The SELECT clause operates on each row of the result of the FROM clause-i.e., on each row of the COURSE table, deriving two columns, CourseId and ExamResult.
  • CourseId is self-explanatory, merely carrying forward the column values from the column of that name in COURSE.
  • TABLE ( SELECT DISTINCT StudentId, Mark FROM EXAM_MARK AS EM WHERE EM.CourseId = C.CourseId ) denotes a multiset whose elements are rows, obtained by taking the StudentId and Mark values from those rows of EXAM_MARK that match the current row of COURSE on CourseId. Note very carefully, however, that this multiset does not necessarily inherit the column names, StudentId and Mark, from the table that is the operand to the invocation of TABLE. The SQL standard allows the column names to be "implementation-dependent" (i.e., undefined) so long as no two columns have the same name. An implementation that nevertheless carried forward the unique names StudentId and Mark would be both sensible and conforming, and would obviate the need for the CAST invocation explained in the next bullet. The same multiset would result if the word DISTINCT had been omitted, thanks to the WHERE condition, but I include it because the example in the theory book uses COMPOSE, which is defined as a projection of a join, and SQL's counterpart of projection uses SELECT DISTINCT.
  • CAST (t AS ROW ( StudentId SID, Mark INTEGER ) MULTISET ), where t is the above TABLE expression, addresses the aforementioned possible problem by assigning the required column names. Note that we need to know and write down the declared types of those columns as well as their names. The "type conversion" operator CAST. Here it is being used to convert a value of some incompletely defined multiset type to one whose multiset type is explicitly defined.
  • AS ExamResult then gives the resulting column the name ExamResult. Note that here the name comes after AS and the expression defining it comes before, in the same style as the use of AS to define the range variables C and EM in the example.

The values for columns such as ExamResult in this example have sometimes been referred to informally as nested tables, being "tables within a table", so to speak. Unfortunately, however, they are not actually tables, but rather multisets of rows. Because of that fact, a column such as ExamResult cannot appear as an element in a FROM clause.

Posted Date: 1/18/2013 6:08:33 AM | Location : United States







Related Discussions:- Example of tables within a table - sql, Assignment Help, Ask Question on Example of tables within a table - sql, Get Answer, Expert's Help, Example of tables within a table - sql Discussions

Write discussion on Example of tables within a table - sql
Your posts are moderated
Related Questions
Create a procedure named STATUS_SHIP_SP that allows a company to employee in the Shipping Department to update the status of an order to add shipping information. The BB_BASKETSTAT

Iterative Control: LOOP  Statements The LOOP statement executes a series of statements multiple times. There are 3 forms of LOOP statements: LOOP, WHILE-LOOP, & FOR-LOOP. LOOP

Assignments in pl/sql The Variables and constants are initialized every time a block or subprogram is entered. By default, the variables are initialized to NULL. Therefore, unle

I have a Pascal Source file that needs to be compiled into a Service. In addition, there are various functions (Pascal Procedures I guess) that need to be created to Read and Write

Providing Results of Queries Expressing queries in SQL is the (big) subject. Here I present just a simple example to give you the flavour of things to come in those chapters.

Named Notation The second procedure call uses the named notation. An arrow (=>) serve as the relationship operator that associates the formal parameter to the left of the arro

Collection Methods:   The collection method is a built-in function or procedure which operates on the collections and is called using the dot notation. The methods like the C

What Is a Collection The collection is an ordered group of elements, all of similar type (for e.g. the grades for a class of students). Each element has a unique subscript whic

Positional and Named Notation You can write the actual parameters when calling a subprogram, using either positional or named notation. That is, you can point to the relationsh

NULL Statement The NULL statement clearly specifies in action; it does nothing other than to pass control to the next statement. It can, though, improve the readability. In a