cano091197
asked on
Info about Hash method
I need use VB 4 to managment a big database (maybe a 800 hundred to one million records), and someone said me that is better use the hash method than a access database.
What is the hash method, where I can get info about it, and why and must select it before a access format.
What is the hash method, where I can get info about it, and why and must select it before a access format.
ASKER
Adjusted points to 200
ASKER
Edited text of question
If you wish to make your own "database engine", you can use a hash table. A hash table uses some sort of a hash function to assign a number to every record in the database. Then, you can make a hash table which is an array whose indices correspond to the hash numbers, and whose elements contain pointers/file offsets/whatever which point to the records.
Example: suppose you have a database of strings, and your hash function assigns a number to every letter of the alphabet (A=1,B=2, etc.), taking the first character of every string and returning the number. So, if your database contains the following records:
File offset Record contents
1 Apple
20 Book
63 Doom
48 Zargon
your hash table will look like this:
Array Index Contents (Implied record)
[1] 1 Apple
[2] 20 Book
[3] Nothing
[4] 63 Doom
[5] Nothing
...
[26] 48 Zargon
Now, if the user wants to find "Zargon" in the database, all you have to do is run it through the hash function (which will give you 26), and go to the file offset specified in the 26th position of the array.
Problems arise when "collisions" occur, that is, your hash function returns the same number for two different records (in my example, this would occur if the database contained both "Book" and "Bread").
This is really all I know. A good, cheap hash method would be to compute a checksum of every record (XORing it in some way), but I have read that this is not very efficient... You will have to ask the real experts for more info.
Example: suppose you have a database of strings, and your hash function assigns a number to every letter of the alphabet (A=1,B=2, etc.), taking the first character of every string and returning the number. So, if your database contains the following records:
File offset Record contents
1 Apple
20 Book
63 Doom
48 Zargon
your hash table will look like this:
Array Index Contents (Implied record)
[1] 1 Apple
[2] 20 Book
[3] Nothing
[4] 63 Doom
[5] Nothing
...
[26] 48 Zargon
Now, if the user wants to find "Zargon" in the database, all you have to do is run it through the hash function (which will give you 26), and go to the file offset specified in the 26th position of the array.
Problems arise when "collisions" occur, that is, your hash function returns the same number for two different records (in my example, this would occur if the database contained both "Book" and "Bread").
This is really all I know. A good, cheap hash method would be to compute a checksum of every record (XORing it in some way), but I have read that this is not very efficient... You will have to ask the real experts for more info.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
I ask for source information and nobody wrote about it
ASKER