ok, I did it

ok, I'll do it

Because it won't lex correctly otherwise. Fork the translation and try it.

Here are the changes I made:

• added <stdlib.h> in full tests
• broke cr_assert_float_eq and it's arguments on multiple lines
• changed all %lf to %.17g
• made solution setup return 0.

Thank you!

approved, though can't mark suggestion as resolved

@trashy_incel: could you look at it again?

• all points addressed, thanks

• took a different approach for generating random positions

• in the original ref sol (still int there, commented out) why would it matter if invalid / unallocated memory was accessed if the Boolean logic of the return statement returns early anyway if the position is not within bounds? It just seems like holding onto some potential garbage until then.

Merged.

I will try to get someone to take a look into DB, maybe it's a problem with some data consistency.

Merged.

Does this happen also when applying changes with a fork?

1. removed do_test, fixed a minor bug in generate_test_case
2. i don't see any problem with the current test case generator. making a more complicated one would probably not be worth the effort.

Removed in python.

Approved

C Fork

• indices made const
• comment added to set-up about return
• return calloc(1, 1); added to set-up
• tester moved out of preloaded
• tester made static
• improved error feedback on strings
• improverd titling for tests