I hope this isn't considered petty as it really is a tiny, tiny bug which has little or no impact, but nonetheless doesn't mean it shouldn't be corrected...

I've spotted an unnecessary call to uc_ssl_switch_to_non_ssl() inside some flow where it has already been determined (boolean).On a very busy site it could have a small performance impact.

If we enter the else{} branch at line 238 then uc_ssl_switch_to_non_ssl() must have evaluated to TRUE in the if{} at line 230. So at line 250 there is no need to test uc_ssl_switch_to_non_ssl() as has to be TRUE to reach this point in the logic.

FILE: modules/uc_ssl/uc_ssl.module

228 //Switch to SSL no matter what the link is if Switch isnt on.
229 if (!uc_ssl_switch_to_non_ssl())
230 {
...
235 }
236 else
237 {
...
239 if (uc_ssl_page_is_in_https_mode())
240 {
...
247 if (!uc_ssl_path_match($current_path, $include_secured_urls) && uc_ssl_switch_to_non_ssl())
248 {
...
250 }

251 }
252 }

N.B Yeah I know I've blockquoted it, but the <code> tags are literal and stop me emboldening the two main points I'm talking about.

Comments

crystaldawn’s picture

Seems logical enough, it has already returned true, no need to have it lingering around again. I'll remove that one.

crystaldawn’s picture

Status: Active » Closed (fixed)