I went through the code base and saw there was a lot of code duplication, particularly with argument handling and colors, so I went ahead and fixed it. A lot has changed so I'll give a brief rundown of whats new.

Colors

Every function in this project had something along the lines of this copy pasted at the top

   let clear = "\x1b[0m"; // reset the font color
   let red = "\x1b[31m";
   let bright_yellow = "\x1b[93m";
   let cyan = "\x1b[36m";

which is a massive amount of code duplication given the number of functions in this project. I opted to create a colors module that would handle all text coloring with the colorize() function, including resetting it to the normal color. It would likely make more sense to take advantage of an existing crate such as termcolor or colored for terminal colors instead, but since this crate was made using 0 dependencies I opted to not use them.

Command line arguments part 1

I next started handling command line arguments. This is also an area where using an existing crate designed for this like clap or structopt would have helped a lot, but as mentioned before I opted not to use them in this case. I began with the "Arguments anywhere" part, where I saw the "changes" section with a giant unreadable line of what has been implemented. I extracted the contents into an array and used a for loop to print and format them all easily. If you do add more in the future, I'd recommend making a second array such as all_implemented_v2 and copying the for loop to print them all.

I then saw the hand formatted "help" message and decided to do a similar thing with what I did with the "changes" section. In retrospect the hand formatted version wasn't that bad, but this new version makes it easier see and add things to the lists without worrying about the formatting as it is taken care of for you.

Command line arguments part 2

I tend moved to the actual meat of the program, parsing the diet, taxon, diameter, and potential "subgroups" arguments and calling the right functions. I created 3 new types for this, a Diet and Taxon enum and an Args struct to parse and hold the arguments we care about. The Args struct handles diet and taxon being in either order, as well as parsing if we have a subgroups flag and the diameter. And since we have these enums, we can simply match on the (Diet, Taxon) instead of having all these big if input1 == ... && input 2 == ... || ..... for every implemented algorithm. And since we are parsing diameter and the subgroups flag beforehand, we can remove all the copy pasted logic doing that as well.

Errors

With diameter being parsed to a f32 from the get go, we can change all the functions in lib.rs to accept it as a f32 instead of as a &str for no reason, which means all of those functions can't panic any more. Speaking of panics, I have changed the code to print nice error messages instead of an ugly panic message when basic expected errors occur, such as being unable to parse diameter or being given an unfamiliar diet/taxon. The messages can be changed to your liking by editing the Display impl for the Error type I made in main.rs.

Other

There are some other small changes I made along the way, such as adding a superscript_number(f32) -> String function instead of doing the big .replace(...) chain you had in your original code. This version of the code should function identically for all inputs that were valid in the original code as well. I also ran cargo fmt on the code base to standardize the formatting, which may have messed up some of the hand formatted alignment unfortunately. This also changed the indentation to 4 spaces instead of 3. If you prefer it to be 3 spaces I can change the rustfmt.toml for this project to do tab_spaces = 3

There are still some improvements that could be made to this code base but this patch brings it to a much better place than it was before and should be Good Enough™ for now