(I'm all ears for anybody else's thoughts on that, as well. I tagged those who wrote answers for it and are here.)
Aw, I didn't want Giuseppe's SNOBOL4 answer to be deleted, I just wanted the bugs fixed, or at least an explanation of why they can't be fixed. And there's a reply to my comment but I can't read it, because the whole answer was deleted.
I used to have deleted-answer-reading privilege, but that's been bumped up from 2,000 to 10,000 rep now. I'd appreciate if anybody would read me what that comment reply says, please.
@Bubbler so would it be acceptable etiquette to put in a brief comment in everybody's answer, so that each of them are alerted their answer needs to be edited to be correct?
So, if we want to use an ordinary regex to solve that challenge properly, we should delete +s and replace runs of *s to single *s and _s to .{0}s before actually running the given regex.
Reverse Hashing
Background
You have invented a new hashing algorithm which is equiprobable to generate a number between \$1\$ and \$n\$. To show off how good your hash function is to your friends you have decided to tell them the amount of time (measured in units of time given below) it will ta...
(((0+a)|_)|0) part is the result of transformation from ((_+a)|_).
And _ can't accept any next char, so it is transformed to 0.
The code doesn't do any logical reasoning to reduce the size of the resulting regex; it simply traverses the recursive structure and mechanically replaces parts of it.
There's actually a problem I'd like to aim at tackling in APL. Do you think it'd be suitable for porting this algorithm to? tinyurl.com/rudolph-Java-v0-62 - with the dictionary passed to it in some way that doesn't count in bytes
I'm hoping it's possible to break 135 bytes, maybe even 131.
I think I get it now, eat Morse tokens one-by-one and attempt to decode at each step until it is valid Morse and matches a dictionary word, then continue. Right?
Well, I tried a larger dictionary, and each word took more than 6 bits.
(The maximum "choice" exceeded 63)
And it'd be cheating to specifically add just the needed words without counting it towards the byte size of the solution :)
Adding the words in a way that counts would negate the benefit of using the algorithm, I think.
Of course, a separate version of this challenge in which the spaces are required to be present, and in the correct places, would be interesting for that.
As long as it takes. I've found that for the mathematically inclined, 3 hours (not necessarily consecutive) gives a very solid foundation for continuing on one's own with increasing one's vocabulary and solving general tasks.
My mathematical knowledge is rather spotty. In terms of formal study haven't gone beyond AP Calculus, and although I know many of the concepts of proofs, I wouldn't consider myself all that good at them.
I've been programming for a very long time though, since age 8 (so 34 years)
QuickBASIC, Turbo Pascal, x86 asm, dabbled in other asm (6502, Z80, 68000), C, C++, dabbling in many others
Well I know the basics of solving a system of linear equations, and have worked with some code doing that... (though I'm not intimately familiar with it). I reverse-engineered a hard drive's ECC algorithm and wrote code to reconstruct the missing parts of bad sectors, adapting a matrix inversion function to Galois fields.
Then later puzzled through a document on Reed-Solomon encoding and figured out how to actually detect which portion is bad (if it doesn't exceed half the number of bits devoted to the ECC).
I still don't know how to adapt that code to do a mixture of both things though: tag some bad bits, and then figure out which of the remaining bits are bad, if possible. I have a hunch it's possible.
Is it time?
Input
Any integer, as a string or an integer type, in the range 0..240000
Output
Truthy/Falsy value
The Challenge
Given the input integer, figure out if the number represents a time (truthy) or not (falsy).
A number represents a time if a time (hhMMss) with leading 0s removed i...
CMC If we have n binary strings, each of length n, we can compute the smallest hamming distance between any pair. If the strings are chosen independently and uniformly at random the expected smallest hamming distance between any pair has some value. E.g . if n = 4 it is 3/4. What is the smallest n so that the value is greater than 1?
@Anush you've got an algorithm that calculates it exactly? i tried different sample sizes and different prng seeds and i still get around 0.75, so theory and empirical evidence are clashing :)
i'll try to rewrite it in a different language, in case i've made a mistake here
Lots of developers import all 69.2kb of Lodash into their apps.
Often to use one function. I thought it would be useful and interesting to code golf all the lodash functions and collate them in one place. So that developers have no excuse.
Lodash Open Round 1 _.snakeCase
https://lodash.co...
Reverse Hashing
Background
You have invented a new hashing algorithm which is equiprobable to generate a number between \$1\$ and \$n\$. To show off how good your hash function is to your friends you have decided to tell them the amount of time (measured in units of time given below) it will ta...
Is it time?
Input
Any integer, as a string or an integer type, in the range 0..240000
Output
Truthy/Falsy value
The Challenge
Given the input integer, figure out if the number represents a time (truthy) or not (falsy).
A number represents a time if a time (hhMMss) with leading 0s removed i...
Lots of developers import all 69.2kb of Lodash into their apps.
Often to use one function. I thought it would be useful and interesting to code golf all the lodash functions and collate them in one place. So that developers have no excuse.
Lodash Open Round 1 _.snakeCase
https://lodash.co...
