This script is intended for utility function testing, library testing, custom assertions.
It is free and open-source.
Introduction About the idea: is not exclusive, programmers tend to use this method a lot and for a long time. The point is to ensure that parts of a software, "units" (i.e modules, functions, procedures, class methods etc), work as they should, meet they design and behave as intended. That's why we use the term "Unit testing". In PineScript we don't have a lot of entities mentioned above yet. What we have are functions. For example, a function that sums numbers should return a number, a particular sum. Or a professor wrote a function that calculates something or whatever. He and you want to be sure that the function works as expected and further code changes (refactoring) won't break its behaviour. What the professor needs to do is to write unit tests for his function/library of functions. And what you need to do is to check if the professor wrote tests or not.
No tests = No code - Total test-driven development
Ok, it is not so serious, but very important in software development. And I created a tool for that. I tried to follow the APIs of testing tools/libs/frameworks I worked or work with: Jasmine (Javascript), Mocha/Chai (Javascript), Jest (Javascript), RSpec (Ruby), unittest (Python), pytest (Python) and others. Got something workable but it would be much easier to implement (and it would look much better) if PineScript had a higher-order functions feature.
API
_describe(suiteName: string) A function to declare a test suite. Each suite with tests may have 2 statuses:
✔️ Passed
❌ Failed
A suite is considered to be failed if at least one of the specs in it has failed.
_it(specName: string, actual: any, expected: any) A function to run a test. Each test may have 3 statuses:
✔️ Passed
❌ Failed
⛔ Skipped
Some examples:
_it("is a falsey value", 1 != 2, true)
_it("is not a number", na(something), true)
_it("should sum two integers", _sum(1, 2), 1)
_it("arrays are equal", _isEqual(array.from(1, 2), array.from(1, 2)), true)
Remember that both the 'actual' and 'expected' arguments must be of the same type. And a group of _it() functions must be preceded by a _describe() declaration (see in the code).
_test(specName: string, actual: any, expected: any) An alias for _it. Does the same thing.
_xit(specName: string, actual: any, expected: any) A function to skip a particular test for a while. Doesn't make any comparisons, but the test will appear in the results as skipped. This feature is unstable and may be removed in the future releases.
_xtest(specName: string, actual: any, expected: any) An alias for _xit. Does the same thing.
_isEqual(id_1: array<any>, id_2: array<any>) A function to compare two arrays for equality. Both arrays must be of the same type. This function doesn't take into account the order of elements in each array. So arrays like (1, 2, 3) and (3, 2, 1) will be equal.
_isStrictEqual(id_1: array<any>, id_2: array<any>) A function to compare two arrays for equality. Both arrays must be of the same type. This function is a stricter version of _isEqual because it takes into account the order of elements in each array. So arrays like (1, 2, 3) and (3, 2, 1) won't be equal.
Usage
To use this script to test your library you need to do the following steps: 1) Copy all the code you see between line #5 and #282 (Unit Testing Framework Core) 2) Place the copied code at the very beginning of your script (but below study()) 3) Start to write suites and tests where your code ends. That's it.
NOTE
The current version is 0.0.1 which means that a lot of things may be changed on the way to 1.0.0 - the first stable version.
진정한 TradingView 정신에 따라, 이 스크립트의 저자는 트레이더들이 이해하고 검증할 수 있도록 오픈 소스로 공개했습니다. 저자에게 박수를 보냅니다! 이 코드는 무료로 사용할 수 있지만, 출판물에서 이 코드를 재사용하는 것은 하우스 룰에 의해 관리됩니다. 님은 즐겨찾기로 이 스크립트를 차트에서 쓸 수 있습니다.