• stifle867@programming.dev
    12·
    2 years ago

    It’s great! I’ve had community members point me in the right direction after already “solving” it incorrectly. It really makes you think about it. You have to expand your test cases and really come up with a better solution.

    Not to say it isn’t difficult especially if you expect the problem to be described perfectly accurately.

    • guslipkin@sopuli.xyz
      11·
      2 years ago

      I don’t need a perfect instruction set, but dang if the examples couldn’t be better sometimes. Like sixteen was in there to show it only counted for 6, but nothing with overlapping text.

      • csh83669@programming.devEnglish
        91·
        2 years ago

        The example for me immediately showed my overlap bug with “eightwo”. There aren’t too many other ways to make this ten words overlap. 🙂

        • stifle867@programming.dev
          7·
          2 years ago

          The problem is the example is actually eightwothree which comes out as 83 so if you replace from start to finish the example passes but the solution is incorrect.

          • Turun@feddit.de
            1·
            2 years ago

            At this point you’re just complaining that the edge case is not highlighted in red.

            I think it’s the right amount of pointers to make you aware of the issue without straight up telling you.

            • stifle867@programming.dev
              5·
              2 years ago

              I’m not really sure how to interpret your comment but I’ll try my best. The edge case that causes some solutions to fail does not have any definition on how to handle it on the problem page. In other words, it does not state anywhere whether the correct interpretation of 1threeight is meant to be 18 or 13. If your solution replaces the words to numbers from left to right you end up with 13 as the value but it’s meant to be 18.

              The example answers don’t cover this but you will realise something is wrong if you run it against your full problem. Community has been very helpful on providing pointers.

  • abbadon420@lemm.ee
    9·
    2 years ago

    I’ve solved part 2 multiple times now, but small intricacies keep popping up every time.

    • Cyno@programming.dev
      2·
      2 years ago

      I’m stuck on one of these intricacies now! I saw some of the ideas on how other people did it but i honestly have no clue what i did wrong. Got any tricky examples to share?

      • abbadon420@lemm.ee
        3·
        2 years ago

        I just solved it. I printed every set of integers in the format “Result: [theTwoDigits], from string: [inputString]” and mannually checked all of them. Eventually found the edge case that was causing troubles. It was a typo.

        • Cyno@programming.dev
          4·
          2 years ago

          Oh god, sorry to hear that 😅i’m feeling desperate enough to try that, i just wrote a different implementation and i get the same (wrong) result. At this point I just want to know what i misunderstood or mistyped cuz its driving me crazy

          • dns@aussie.zone
            2·
            2 years ago

            The bug is some strings can have overlapping characters. onEight threEight fivEight. There are more cases. So if you do a search and replace your string becomes 1ight 3ight and the second number does not get found.

            Possible fixes: Search and replace and add the extra letter: oneEigh then search and replace. Search and replace words to numbers but put some extra letters in just in case.

            • Cyno@programming.dev
              1·
              2 years ago

              Thanks, I managed to find the culprit in the end however - I was using a forward-look of 5 characters for finding if it matches a word so in a string of a3two for example, it would register the two before the 3. It was an easy fix once i found the issue.

  • Matthew@programming.dev
    6·
    2 years ago

    I was having a bitch of a time with day 1’s part 2 until someone pointed out that sometimes two numbers share a first/last letter (eightwo).

    • purplemonkeymad@programming.dev
      2·
      2 years ago

      It was for sure to catch people that were taking the easy route and using a replace. I had to rework my solution to get the start positions for the matches. That method made both parts easy to get together.

  • ornery_chemist@mander.xyz
    1·
    2 years ago

    I managed the overlap logic just fine (if in a clumsy way), but spent over an hour trying to debug my crappy parser-combinator code when the issue all along was a mis-spelling in my lookup table.