website: cognitive complexity does insufficiently explain method

[matu3ba]

What it does
Checks for methods with high cognitive complexity.

What kind of methology is used in the code is completely missing. Is this method implemented? #3793

Why is this bad
Methods of high cognitive complexity tend to be hard to both read and maintain. Also LLVM will tend to optimize small methods > better.

The first statement is unfalsifiable due to missing definition. LLVM and most compilers do graph-based optimisations, so a long switch-case is no problem for LLVM to handle. Which is in contrast to what cognitive complexity applies as method (from my experience).

Known problems
Sometimes it’s hard to find a way to reduce the complexity.
Search on GitHub

It is not possible to disable unless one annotates it in every file. Many problem domains are inherently complex and splitting the code does not make the problem easier. As it works currently, it does not take into account the code graph of the control structure in a proper way and thus is insufficient.

Example
No. You’ll see it when you get the warning.

Configuration
This lint has the following configuration variables:
cognitive-complexity-threshold: u64: The maximum cognitive complexity a function can have (defaults to 25).

Great, a configuration without understanding how they work.

[matu3ba]

The same argument holds for this description https://rust-lang.github.io/rust-clippy/master/index.html#type_complexity

[porzechowski]

Since I'm looking for a good-first-issue, I assume that this issue is something to work with?

However I have a question, I saw that it was previously name cyclomatic complexity, and later was renamed to cognitive, but to be honest it doesn't look like it calculates either.

Briefly:
Cyclomatic tells us how many independent paths are there in a piece of code.

Cognitive tells us how hard (or easy) certain part of code is to understand, based on how much cognitive effort certain part of code requires. It will increment each time something is trying to challenge our limited cognitive capacity.

Right now this looks like a bit crippled Cyclomatic complexity.

Isn't it a bit misleading ?