in reply to Re^4: how to define key for a hash
in thread how to define key for a hash

Even an expert cannot tell you that without knowing your system, your environment and so on. I already told you this. If you can't do the analysis yourself, you will have to hire somebody who does the analysis for you.

Replies are listed 'Best First'.
Re^6: how to define key for a hash
by Anonymous Monk on Jun 25, 2009 at 12:54 UTC
    :( thaxs
[reply]
      Let me try and rephrase what has been said in a way that will hopefully be of more help to you.

      The type of information you are attempting is best implemented with a database. Databases are designed exactly for the problem you are attempting to solve and using one will make your lookups much easier. A properly installed, configured, and implemented database on an adequate system will have no problem meeting your performance needs. Whether your database would be properly installed, configured, and implemented on an adequate system is not a question this forum can answer.

      If you feel that you cannot use the database solution, it is possible to solve the problem with a hash. A hash can also easily handle the problem set (100k records), assuming the system has adequate resources. You would need to key the hash by whatever unique value you have (group id?), possibly using a HoH (hash of hashes). Because your lookup values (phone numbers) are not unique, you will have to deal with the complexities of duplicate matches, etc, in your code (instead of letting a database do this for you).

      Once you have chosen one of these options, this forum can help you to progress. It cannot help you to choose which of them is appropriate, since your unique environment is not known (and that's really beyond the scope of perl help anyway).
[reply]
        let me explain my work flow it may help in a better way. I have a set of calldetails which is coming to my system unix file area which needs to be lookup with the incoming phone numbers and replace a perticular value in the call details file and then use our core api to calculate the rate of call and upload to our database. So my intension is to do the lookup before the call rates are getting calculated.since my sore api which calculates the rate can only identify input as flat file,so i have to get the file as flat file after this lookup. So how can i bring database in b/w these? that y i thought of a perl script which can do the lookup activity whose performance is comparable with database. But since this perl is very new to me..i am really confused and am not getting any encouraging reply from this site also.... I have alreadt done this logic with sql loader and shell scripting ,but im not quite satisfied with the performance.That y i turned to perl.....
[reply]

In Section Seekers of Perl Wisdom