Home/ Questions/Q 7587693
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-30T19:44:13+00:00

Consider a huge CSV with the following structure (modified for simplicity): ID, NAME, ADDRESS,

  • 0

Consider a huge CSV with the following structure (modified for simplicity):

ID, NAME,  ADDRESS, PHONE, MAIL
1,  Jon,   UK,      403,  jon@skeet.com
2,  Marc,  UK,      292,  marc@gravel.com
3,  Darin, France,  291,  darin@dimitrov.com
...
(Some million records)

The natural data structure for quick fetch is a Hash Table, where every ID is a key and NAME, ADDRESS, PHONE, MAIL are the value. My dillema is the data structure of the values.

Storing it in a HashMap where every row header is the key is a waste of space, because the row headers are exactly the same for each row. Storing it as an Array would lose the metadata for each item, because the reader

I was thinking of two approaches:

  • Overload Java’s Hashmap. The row headers will be stored once, and every ID will be associated with a String Array. The get() method will be overloaded so that it will return a map between the header rows and the corresponding fields in the row.

  • Create a dumb class which stores the data for each row using getters and setters (row.getMail(), row.getAddress(),…)

What’s the right way to go, in terms of memory efficiency, type safety and speed?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    I’d go for the “dumb” class instead of overloading a collection.

    I don’t know about type safety or speed, but I would say your code will be more readable. Those values go together; encapsulate them in an object to emphasize the point. Is there any behavior associated with them besides get/set? If yes, then so much the better.

    • 0