Is loading unused constants a bad thing?

schwim

Hi there guys,

I'm creating a language system for my script, and I was wondering:

When I look at large scripts like PHP-Nuke, or phpBB, I see that they're calling a file with 100's, if not thousands of lang constants in it, to load a page that may only use 20 of the definitions.

Does the script only load the definitions it needs? Is there any ill impact at all when loading a page that large just to use a couple of them?

I'm considering using separate lang files(like I've seen in Zen-Cart) so the lang file being loaded is smaller. My script is pretty tiny to begin with, but I'm trying to do these things to learn how to code smart. If there's absolutely no use in doing it, then I guess it's not so smart 🙂

thanks,
json

void_function

The less you load, the faster it goes and less memory it uses. Although, hundreds of lines more or less won't make that much of a visible impact, while thousands may. Personally, I write my scripts to load as less as possible.

Weedpacket

I also prefer to have individual pages for this, just to cut down on the work involved in locating particular strings; if the same strings are required on multiple pages, a few judicious include()s in the language files can do nice things.

Plus it mirrors the fact that translated content isn't all lumped into one big ... lump.

However, I usually don't use constants. Once a constant is defined, it can't be altered; so you can't have default values for everything defined to fall back on if something hasn't been translated, nor can you define, say, "en-us" in terms of "en-gb plus these differences."

schwim

Hi there guys, and thanks very much for the replies.

Weedpacket, could you explain:

1) what type of system you do use instead of constants for language systems

2) an example of why it's better

Weedpacket

1) An array, indexed by a default: $lang['Welcome'] = 'Bienvenue';. There's also the gettext extension which automates most of this.
2) See the previous post.

void_function

I've had problems with UTF-8 encoded translations if I used a PHP parseable include file. UTF-8 encoded files have certain control marks at the beginning, so I learned that if I use an ini file, and parse it with parse_ini_file() that the control marks will be ignored, and so I could use UTF-8 translated files.