<aside> 💬
</aside>
In order to nicely transition to a Part 2 solution, it made sense to revisit one part of the Part 1 solution, where I had a relation “this character is/is not a digit.”
**isDigit Ch is { false? } :-
val _ _ is Ch.
isDigit Ch is true :-
digit Ch is _.**
Instead of this relation, I refactored the solution to use an option type instead of a boolean. The new relation works the same way but expresses the possibility that the position Line, Col describes a number, and also contains that number if it exists.
**numberAt Line Col is { nothing? } :-
val Line Col is _.
numberAt Line Col is (just N) :-
val Line Col is Digit,
digit Digit is N.**
I like this solution better, in addition to it leading nicely to the part 2 solution. I also added the INT_TIMES builtin and used that, because it was annoying not to have that.
Here’s the refactored Day 1, Part 1 solution on dusa.rocks%20%3A-%20%0A%20%20%20%20firstNumberOf%20Line%20is%20Tens%2C%0A%20%20%20%20lastNumberOf%20Line%20is%20Ones.%20%0A%0A%23%20Deriving%20a%20single%20integer%20solution%0A%0Agoal%200%20is%200.%0Agoal%20(plus%20Line%201)%20is%20(plus%20Accum%20Value)%20%3A-%0A%20%20%20goal%20Line%20is%20Accum%2C%0A%20%20%20answer%20Line%20is%20Value.%0A)
In part 2, in addition the string one4two3 has a first number “one” and last number “3.” My goal is to represent this example as the number 1 being present at position 0, the number 4 being present at position 3, the number 2 being present at position 4, and the number 3 being present at position 7, and no other numbers being present.
one4two3
^ ^^ ^
| || |
1 42 3
If we change our numberAt relation to describe numbers at these new places, the rest of our logic continues to work
numberAtSo I enrich the set of digit facts: in addition to
**digit "1" is 1.
digit "2" is 2.
digit "3" is 3.**
I add
**digit "one" is 1.
digit "two" is 2.
digit "three" is 3.**
and so on to that set of facts.
And when does a number appear at a position? It made sense for me to do it this way: we’ll want to calculate all substrings of every line, and if a substring starting at a position matches a digit pattern, then that digit is present at that location: