Do you like writing your tests in a tree of describe
, and it
, but also have
a hard requirement on stable rust? This crate is for you!
If you don't know what is Rust, are confused by the terms BDD, TDD, or just want a gently beginner introduction, please go to the Beginner Section.
The last stable documentation is available for consultation at mackwic.github.io/rspec.
All rspec releases are garanteed to compile against the latest stable rust and are tested on all rust versions from the 1.9.
Add this in your Cargo.toml
:
toml
[dev_dependencies]
rspec = "1.0"
and add this to your src/lib.rs
or src/main.rs
:
```rust
extern crate rspec; ```
You can see complete examples in the examples/
directory.
You can now use rspec in your unit tests, this example will use the cargo test
runner:
```rust fn add(x: u32, y: u32) -> u64 { x + y }
fn testadd() { rdescribe("add", |ctx| { ctx.describe("0 <= x + y <= u32::MAX", |ctx| { ctx.it("2 + 4 = 6", || { asserteq!(6, add(2, 4)); Ok(()) });
ctx.it("4 + 4 = 8", || {
assert_eq!(8, add(4, 4)); Ok(())
});
});
ctx.it("is associative", || {
assert_eq!(add(2, 1), add(1, 2));
assert_eq!(add(4, 1), add(1, 4));
assert_eq!(add(4, 5), add(5, 4));
assert_eq!(add(12, 1), add(1, 12));
Ok(())
});
});
} ```
You can also use rspec in integration tests, this example uses the rspec runner:
```rust extern crate rspec; use rspec::context::describe; use std::io;
pub fn main() { let stdout = &mut io::stdout(); let mut formatter = rspec::formatter::simple::Simple::new(stdout); let mut runner = describe("rspec is a classic BDD testing", |ctx| {
ctx.it("can define tests", || Ok(()));
ctx.describe("rspec use results for tests results", |ctx| {
ctx.it("passed if the return is_ok()", || Ok(()));
ctx.it("failed if the return is_err()", || Err(()));
ctx.it("is executed so you can use dynamic values", || {
if (42 % 37 + 2) > 3 { Ok(()) } else { Err(()) }
})
});
ctx.describe("rspec also supports asserts", |ctx| {
ctx.it("is a simple test", || {
assert_eq!(true, false);
Ok(()) // don't forget to return a Result
});
ctx.it("can also pass", || {
assert_eq!(true, true);
Ok(())
});
});
});
runner.add_event_handler(&mut formatter);
runner.run().unwrap();
}
```
Again, you can see complete examples in the examples/
directory.
The last stable documentation is available for consultation at mackwic.github.io/rspec.
... are greatly welcome! Contributions follow the standard Github workflow, which is:
[WIP]
in the title and we'll gladly help you.Take a look at the issues if you want to help without knowing how. Some issues are mentored!
Welcome in the Rust community! Here are some links which can hopefully help:
TDD, short for Tests Driven Development, is a methodology of development where your code is always working, where refactoring is easy, and you know exactly what piece of code do and what it doesn't.
With TDD, legacy code is limited, your build is always green and you don't have regresssions.
This is a wonderful and magical land, a well-keeped secret where only the best of the best are admitted, people who don't compromise on quality because they know the cost of the absence of quality. People who know that over-quality is a non-sense.
Here are some useful links:
BDD, short for Behavior Driven Development, is a variation on TDD; some would say that BDD is simply TDD but refined.
BDD states that tests are not the center of the methodology. They are one of the most usefull tool available, but we should not look at them in too high regard. What matters is the contract they seal, the described behavior.
Thus, there is enough tests when the behavior of the struct
is sufficiently
described. Thinking in term of behavior has two benefits:
When doing TDD, it helps to make incremential steps. Just write examples of
how to use the functions, and make this example pass, then go to the next one.
Your tests will naturally have one clear intent, and so will be easy to debug
/ rely on when refactoring.
This is the describe/it approach, which this crate hopes to fill.
By describing behavior, we are doing an analysis of our program. This
analysis can be very useful! Say... an User Story, for example. Given the
formalism As a X, When Y, I want Z, you can assign scenarii describing the
high-level behavior of your units. The Gherkin formalism is often employed for
this, it use a Given X, When Y, Then Z structure.
This project does not aim to help on high-level BDD, see the cucumber for
Rust port
for that.
The foundation of BDD is well explained here and also here.
BDD written with the Gherkin formalism can really gain from a layer of DDD (Domain Driven Development), but this is another story....
Mozilla Public Licence 2.0. See the LICENCE file at the root of the repository.
In non legal terms it means that: - if you fix a bug, you MUST give back the code of the fix (it's only fair, see the Contributing Section). - if you change/extend the API, you MUST give back the code you changed in the files under MPL2. The Contributing Section can help there. - you CAN'T sue the authors of this project for anything about this code - appart from that, you can do almost whatever you want. See the LICENCE file for details.
This section DOES NOT REPLACE NOR COMPLETE the LICENCE files. The LICENCE file is the only place where the licence of tis project is defined.