Mechanism for providing efficient access to redundant number representations

• US 8,131,709 B2
• Filed: 01/14/2005
• Issued: 03/06/2012
• Est. Priority Date: 01/14/2005
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Ricard, Gary Ross, Schmidt, William Jon
• Primary Examiner(s): Rones, Charles
• Assistant Examiner(s): Khoshnoodi, Fariborz
• Application Number:
  US11/035,718
• Publication Number:
  US 20060161574A1
• Time in Patent Office: 2,608 Days
• Field of Search: 707/5, 707/2, 707/4, 707/6, 707/713, 707/715, 707/706, 707/741, 707/602, 707/780, 707/779, 707/765, 702/20, 435/6
• US Class Current: 707/713
• CPC Class Codes: G06F 16/2445 Data retrieval commands; View definitions

Abstract

Disclosed are an apparatus, method, and program product which each provide an enhanced database engine. Access to precise values is provided while permitting unfettered access to those who are not interested in precise values. This is accomplished via an enhanced database index and indexing method. The entries of the database index are normalized to exclude precision. Individuals interested in precision can then specify precision as part of their query through use of a PRECISE keyword. Results are then filtered to account for the specified precision.

Claims

1. A method of adding an entry to a database index, said method comprising the steps of:

identifying a first decimal floating point value placed into said database index, said first decimal floating point value from a database column, said database column containing at least one decimal floating point value of a first precision and at least one decimal floating point value of a second precision, said first precision being different from said second precision;

normalizing said first decimal floating point value to produce a normalized value, wherein all normalized values of decimal floating point values from said database column have a common precision; and

inserting said entry into said index, said entry including said normalized value.

2. The method of claim 1 wherein said value is a key value.

3. The method of claim 1 wherein said identifying, normalizing, and inserting steps are performed as part of creating a new database index.

4. The method of claim 1 wherein said identifying, normalizing, and inserting steps are performed as part of updating an existing database index.

5. A method for improving the performance of a client'"'"'s program, said method comprising the steps of:

modifying a database engine to perform the sub-steps of,

identifying a first decimal floating point value placed into said database index, said first decimal floating point value from a database column, said database column containing at least one decimal floating point value of a first precision and at least one decimal floating point value of a second precision, said first precision being different from said second precision;

normalizing said first decimal floating point value to produce a normalized value, wherein all normalized values of decimal floating point values from said database column have a common precision; and

inserting said entry into said index, said entry including said normalized value; and

providing said database engine to said client.

6. The method of claim 5 wherein said value is a key value.

7. The method of claim 5 wherein said identifying, normalizing, and inserting steps are performed as part of creating a new database index.

8. The method of claim 5 wherein said identifying, normalizing, and inserting steps are performed as part of updating an existing database index.

9. A method of improving the performance of a client'"'"'s program, said method comprising the steps of:

modifying a database engine to include;

a) an enhanced database index, said enhanced database index having normalized key values stored therein, said normalized key values representing non-normalized values;

b) a program for accessing said enhanced database index, said program configured to select said normalized key values as return results and configured to select said non-normalized values as return results; and

providing said database engine to said client.

10. The method of claim 9 wherein said program is capable of selecting return results based upon a query language keyword.

11. The method of claim 10 wherein the query language is structured query language (SQL).

12. The method of claim 11 wherein said keyword is a PRECISE keyword which is added to the grammar of SQL.

13. A method for providing an enhanced database engine to a customer, said method comprising the steps of:

obtaining a license to a commercially available database engine, said database engine including an indexing mechanism and a query mechanism;

modifying said indexing mechanism configured to generate a database index including normalized key values stored therein, said normalized key values representing non-normalized values;

modifying said query mechanism configured to select said normalized key values as return results and configured to select said non-normalized values as return results; and

providing said database engine to said client.

14. The method of claim 13 wherein said query mechanism configured to select return results based upon a query language keyword.

15. The method of claim 14 wherein said query language is structured query language (SQL).

16. The method of claim 15 wherein said keyword is a PRECISE keyword which is 2 added to the grammar of SQL.

17. A computer-implemented method of managing a computerized database, comprising the steps of:

maintaining a database table having a plurality of entries and a plurality of columns, wherein a first column of said table contains entries in a numerical format supporting different precision, a first plurality of said entries containing respective values in said first column represented in a first precision, and a second plurality of said entries containing respective values in said first column represented in a second precision different from said first precision;

maintaining an index for said database table, said index containing normalized values representing said values in said first column, said normalized values being represented in a common precision;

responsive to database queries of a first type, identifying values in said first column which have different precision but identical normalized representation in said index as equivalent; and

responsive to database queries of a second type, identifying values in said first column which have different precision but identical normalized representation in said index as not equivalent.

18. The method of claim 17 wherein said numerical format is a decimal floating point format.

19. The method of claim 17 each said database query contains a query language keyword determining whether the query is a query of said first type or a query of said second type.