Thread: Please help to search key and return value from text file using data structure in java

Hi all,

I am having a file contains key and value pairs separated by space.

I want to read that text file and store into a data structure in java so that, If i want to search a key in file then it should have to iterate only upto the length of search string (key) entered by end user, not to iterate through the whole file for searching the matching key.

e. g. If the keys in file are like
java, java program, java programming

and if the search string is like "Programming"

then the search should like :

if array[0][p] = "p" -> false
if array[1][r] = "pr" -> false
if array[2][o] = "pro" -> false
if array[3][g] = "prog" -> false
if array[4][r] = "progr" -> false
if array[5][a] = "progra" -> false
if array[6][m] = "program" -> true //match found
if array[7][m] = "programm" -> false
if array[8][i] = "programmi" -> false
if array[9][n] = "programmin" -> false
if array[10][g] = "programming" -> true //match found

so the longest match is 10

Then the program should have to return the value related to the key " java programming".


Please suggest how can we achieve the same feature using java.

Originally Posted by Norm

Not sure what the usage of the second dimension index in the example shows. I assume those variables are ints. How do they get assigned values? Their values appear related to the names which has no relationship in java code.
The variable named a has nothing to do with the String "a"

Have you looked at using a Map?

------------------------------------------------------------------------------------
Thanks for reply..

I checked with map searching is working fine. But because of some rule of searching I need to iterate the hash map completely, and it is not giving performance.

That's the reason, I want to implement it by using data structure in java, so that the iteration of the data structure should be upto the length of search string only.

Kindly please suggest the data structure that will be the beast for fast reading , searching string in file and return the value accordingly.

the search string we are searching in map is not a key only. It could be a string of multiple lines with one or more keys.

e. g. Search string :-
"The Longest rivers in the world is Nile and Amazon. Nile belongs to Africa and Amazon belongs to South America. etc.......... "

Suppose the above search string contains the key as:-
rivers, longest, South

so we have to search this strings into the file and returns the related urls.

And If we read the complete file and stored it into the map then to search the matching string in the map we have to iterate the map untill we found the matched key.

so want the fastest search using and data structure. Because by using hash map we got the better performance using hashmap for a shorter search string.

but if the search string is too long then hash map also taking time.

so please suggest and algorithm.. can be used

thanks

Originally Posted by Norm

Can you explain what you mean here by Search String:
Is that what you want to find
or is that a list of keys you want to search for?

Or are you separating the separate tokens in the "Search String" and doing a lookup for each token. More tokens will cause more lookups.

Can you explain this:
A Map that is given one key will very quickly tell you if it contains that key. I don't know that the length of the key will be an issue.

yes the search string can contains multiple tokens. And it's length can be of any size, so we have to split that search string and then search for the matching keys and return the url.

I have already posted the code implemented by using hash map in the same forum, but I want to implement it by using any different algorithms so that the performance will be faster for any length of string.

you can search the post on the same forum by searching:
"Fastest way to read and search a string in a large file using core java"

Please suggest accordingly.

Hi All,

I found the solution for the same.

I created own tree structure and store the keys in file as character by character.

And now the searching is very fast as compare to other solutions...