Re: GC5NXHH 3x name in Östergötland in a year

Not my challenge, but an old one that can use a checker in my opinion.

It requires to have found caches that start with all the letters in your geocaching nick within the Swedish region Östergötland, within 1 calender year, 3 times.

The rules explain the following exceptions:
- letters (A-Z, Ä, Å, Ö) and numbers count, other symbols can be ignored.
- capital or small letter isn't important
- for numbers in the geonick:
- the cache name may start with that number
- or the number may be replaced as follows: 1 -> i, 2 -> s, 3 -> b, 4 -> y, 5 -> s, 6 -> g, 7 -> t, 8 -> b, 9 -> g, 0 -> o
- X may count as Q and vice versa

So, I did eventually manage to shoehorn this into my multi-test checker script. It really took some mindbending (and several modifications), but I think it's working as expected.

https://project-gc.com/Challenges/GC5NXHH/82907

Test it out and let us know.

@bmuzzin I ran a couple of people through it and I believe you're giving false positives for people that have the same letter multiple times in their nickname. For example you would get green for both your Z's if you had 3 Z-caches but according to the challenge you would need 6 of those since you have two Z's.

Aj. Thank you for your alertness! Bmuzzin, is this possible to fix?

Probably fixable... Although, if I read the requirements on the cache page, doesn't really address duplicate letters. The example uses multiple C's, and uses different caches for those, but it doesn't explicitly say they need to be different. It could be really hard to pass the challenge of you have a lot duplicated letters in your name.

Being a native speaker I'd say it's implied that the caches needs to be different. If you've named yourself zzzzzzzzzzzzzzzzzzz I'd suggest skipping the challenge or changing your name. :)

Alright, I'll disable it until I can fix it.

TL;DR. Okay, I think I've fixed it, and it's been re-enabled. Test it out and let me know.

The 'same' operation can now collect the caches of previous operations, and excludes them from the search results for the subsequent ones. It makes the output a little ugly if you have a really really long caching name... but, I think it works. This option is only going to work for the 'number' test right now, and would only work in certain situations where the sub-operations can use any matching result to satisfy (ie. in this case, where the test is for the same letter). More complicated situations probably wouldn't work, for example, if there are multiple possible "paths" to satisfy subtests, but caches require in an earlier step would satisfy one in a later step, but not vice-versa.

Looks good! Thank you!