data structure...

[huskers]

I need help in finding an efficient way of storing the following data. I just want to store all the information in one data structure. The data that will be obtained contains the following info:

The data contains server names.
Each server name will have a socket that it needs to write to.
Each server name can have any number of clients. (2 or 4 or more depending on the data obtained from database)
Each client can have any number of sub clients.

so members are as follows:

serverName char[50];
socket int
client_name char[50];
sub_client char[50];

Thanks

 

[columb]

It may be overcooking it but if you make your clients and sub clients linked lists then

stuct subclient
  {
  char *name;
  struct client * next;
  }

struct client 
  {
  char *name;
  struct subclient *first;
  struct client *next;
  }

struct server
  {
  char *name;
  int socket;
  struct client *first;
  }

but you're going to spend a lot of time using malloc!

Columb Healy

 

[huskers]

I am looking for an efficient way to search elements in the following data structure

struct x {
int item;
struct node *element;
}

struct node {
int a;
struct node *ptr;
}

There can be many possible elemets associated for each item and there can be many items. All of this are stored in database. I retrieve the data and create the above data structure.

According to the data sets I have I constructed a data structure for 5 items.
item1 -- 250,000 elements
item2 -- 75,000 elements
item3 -- 50,000
item4 -- 25,000

I have some elements in a file and I have to find out what items are associated to each element in the file. I am looking for an efficient way to traverse through this data structure to search for elements.

Presently my search is as follows: I traverse throught the linked list of elements for each item until I find the required one. Once I find it I break out of the loop and look in the next item.

Is this an efficient data structure? Is there a fastest way to search through this?

 

[columb]

You might want to look at binary trees. This link is a good place to start. In breif you're sorting your data as you load it in so that searches on the data are as fast as posible.

Columb Healy