Benjamin Woosley

Thanks @luke-jr, I adapted the short reasons from c59e716. How does that look? Is it alright to put the input index in the rejection reason or better to append?

Also moved the string instantiation out of the bench loop.

@luke-jr moved the input index to the end. Now I'm questioning the need for the `bad-txns-` prefix, given that these are to become debug output, rather than the reason string,...

Removed `bad-txns-`. Note these are propagated as the debug string alongside a `bad-txns-nonstandard-inputs` `CValidationState` reject reason.

@luke-jr Good point - split the reason and debug fields, and replaced `bad-txns-nonstandard-inputs` with one of: `bad-txns-script-nonstandard`, `bad-txns-scriptsig-error`, or `bad-txns-scriptcheck-sigops`.

Rebased for #13429

Fixed rebase for #13429, dropped refactor commit, and made use of every reason recommended by @luke-jr

@practicalswift replaced existing and new uses of `&array[sizeof(array]` with `array + sizeof(array)`

Stopped shadowing `reason` var (by reusing the existing var). Limited test array access changes to only affecting the added tests.

Rebased for #14908, added test for `bad-txns-input-scriptcheck-missing`, so testing covers all failure conditions, and removed some test precondition checks that were already captured in the `AreInputsStandard` outputs.