or
Mathgeek
string optimization counting
String parsing this time!

Input is a string of indeterminate length, composed of any number of upper or lowercase letters.

Your goal is to find the longest strictly-furthering substring. This means making sure the next letter comes later than in the alphabet than the previous.

For the purpose of this challenge, all capital letters are strictly greater than all lowercase letters. [a-z][A-Z] is our order, first-to-last, smallest-to-largest.

a can only ever be our first letter, and Z could only ever be our last.

The output should print the longest strictly-increasing substring. In the event there are more than one that ties, print out either.

Input is the entry string, output is the output substring, I/O is fairly tight.

## Sample Test Cases

| Input | Output|
| :-: | :-: |
| ajrleowple | ajr OR eow |
|flwpeAREP | flw OR eAR |
|epejikwEIOepoaWlpd | ikwEIO |
| [blank string] | [blank string] |
|[alphabet in lowercase followed by uppercase] | [Same as input]|
|zyxwvutsrqponm| [Any one of those letters] |
|f|f|
|qqqqqqqrrrrrrr| qr|
Longest Stringly-Furthering Substring
I've flagged my post because it is invalid and I don't have time to fix it soon.
Actually a good thing, because this exposes a TA issue: What do can an author do when he realises that his submission is faulty?
@Adám Your solution is incorrect as it does not perform "strictly" correctly. he qqqrrr example gives a faulty output.