How to index the flat file like mysql

[Man4ish]

In mysql values can be retrieved like less than a given value.
select * from table where number > 12. But using mysql will be slower, so planning to implement the same methodology but that should be very fast. Which data structure can be used? In my previous thread i have already asked about B Tree and got very good reply but now my planning is to implement the same methodology as used in mysql for key retrieval. Thanks in advance.

[gpraghuram]

Then you can use memorymap in which and you can maintaina map with key and position so that you can easily retreive the values.

raghu

[Man4ish]

Then you can use memorymap in which and you can maintaina map with key and position so that you can easily retreive the values.

raghu

Can i retrieve the value with less than a given value if yes then how we can implement it.

[Man4ish]

[QUOTE=Man4ish;3 475178]I have one example. I have one data of only supervisor in company for past 50 years.

Supervisor Joining_date leaving date height
A.M kadik 23/2/51 23/4/57 165
M.A. Khan 25/4/57 2/02/64 179
R. A. Sharma 3/02/64 05/09/79 169
K.Kandaswamy 6/09/79 07/07/91 174
R.A.Brown 09/07/91 09/08/2000 165


I need to get data (name ,height and data of joining) for all the supervisors which were there in period 2/02/71 to 03/05/95. This is very easy in mysql but how efficiently this can be done in C/C++.Since MySQL uses the B+ Tree structure. Can B+ Tree be used here?if yes how? I have no idea at all.Thanks in advance.

Regards
Manish

[gpraghuram]

The data can be in any format.
But if u want to retreive values with some specific logic then you have to maintain keys accordingly.
You can maintain one map with name as the key and the position.
You can maintain a second map with the age as key or the period of stay and position.
So when retreiving you have to use the right map and then get the values accordingly.

raghu

[Man4ish]

The data can be in any format.
But if u want to retreive values with some specific logic then you have to maintain keys accordingly.
You can maintain one map with name as the key and the position.
You can maintain a second map with the age as key or the period of stay and position.
So when retreiving you have to use the right map and then get the values accordingly.

raghu

Thank you very much for your answer. For small file it seems ok. but if i have a file of around 100MB then i have to read the whole file sequentially which may reduce the speed. How this can be done randomly?
Thanks in advance.