Challenge #177: Ham Hamlet

[LifebloomX]

@RolandSchubert I just wanted point out some minor errors here.

I added an equal sign onto the filter, and then also changed the way Summary should count Vowels:

I changed the field from "Vowel" to "Count", and changed action frmo "Count" to "Sum."

I basically summed the count for each group (vowel), and this changed the order of count number for vowels.

However, this didn't influence "which" vowels got counted the most for top 5 (only order changed).

Please let me know if I made an error.

[RolandSchubert]

Hi @LifebloomX , you are absolutely right regarding the COUNT/SUM. I used "> 8" for length, because "yay" is appended to words startings with a vowel (length of encoded "yay" = 6) and I expected the "original" word to have at least 2 letters (length encoded >= 4).
Thank you for the hint, I'll correct the workflow.

[garthn555]

No extra credit for me. But I tried just throwing all combinations of A-Z plus a-z and the encoded string at the macro, and oinked at the result

[JohnJPS]

Spoiler

I split the records into two-character strings, and assumed 'y', 'a', and ' ' would be the three most frequently occuring; I also found a paper (https://link.springer.com/content/pdf/10.3758/BF03195586.pdf) with lots of character frequency information that I pulled in for a dumb first guess... from there I iterated a bit, making a bunch of corrections. It gradually became clear that it was Shakespeare's complete works, and something about Gutenberg... a quick trip to Google found this (https://www.gutenberg.org/files/100/100-0.txt), which is the final English translation desired. From there it was easy to back-fill.

[LifebloomX]

@RolandSchubert That's what I thought when I saw that > there...so you just assumed it had at least 2 letters on the front. Anyways, thanks for posting your solution! Your solution was very easy to follow 😄

[phottovy]

It looks like I took a similar approach as many of you. Thanks for the challenge!

[PhilipMannering]

Very nice.

Spoiler

I tried to find

1. The most common characters

2. The most common characters at the start of words (vowels)

3. The most common characters in one letter words (ayay and Iyay)

4. The most common characters that come in pairs.

5. The most common words beginning with vowels

...and pieced it together from there.


Workflow

[TonyA]

This was fun. Probably the first time I've used Designer where the outcome wasn't a reusable workflow but a solution to a problem. I'm attaching the workflow just to show some of the tools I used.

Spoiler

I proceeded the way most did. I didn't see that they gave us two of the codes, but they were pretty obvious since it was Pig Latin so every word would end with "ay" followed by a space. I did use the Frequency table and it was a huge help. For example, I did pick up a couple of words very quickly after figuring out the "o." One of the first was "anyone" which then led me to a two letter word starting wth "o" that was either "or" or "of" since it didn't have an "n." The frequency table pointed me to the "f" since the code was pretty far down the frequency list. 

[LordNeilLord]

I had no idea how to approach this one until @PhilipMannering gave me some tips

Spoiler

I started by looking for the most common things, space, the, I etc.... and once I had a few I started to notice a pattern in the code......like C was a capitol and c was lowercase. So I used some simple powers of deduction to manually work out some of the letters....which was enough to unlock the macro and give me the full decryption.

[ggruccio]

This one I spent extra time on for two reasons.

1. I really was trying to understand the ASCII rotation logic and looking for a pattern that would repeat for the two character codes.

2. Converting from pig-Latin to English isn't easy and is hampered by punctuation and Olde English.

However:

Spoiler

1. On the ASCII codes, I realized I only needed to get 26 correct.  And, "oh yeah I have Alteryx".  So I took all of the letters of the alphabet and ran them against every possible combination of the two digit characters.  Didn't matter how many I had wrong, as long as 26 were correct.....which had to be.

2. Then on the conversions, I had no trouble creating some basic logic for handling vowels, but the consonants have so many strings that could be possible syllables.  "S-T-R", "S-T"  etc. but my ruleset has a problem - for example how do I know without reading the context that a word should be "STRAIN" or "TRAINS"?  So I did my best to create a ruleset that controlled for consonant clusters then worked word exceptions in.  That exceptions list gets long quickly, but I don't know how else to do it.  I controlled for most of what would export into excel and paste into Word...which accounts for about 7 pages.  I could literally spend weeks on the rest of the text though.