Implicit Casting of Array Elements #6376

IGI-111 · 2024-08-07T09:02:47Z

CS-FSSA-010

When it comes to typing the elements of an array, the type system is not able to stick to proper types.
This leads to the following snippets all successfully compiling:

let a = [8, 8, 8u64];
let b: u32 = a[2];

// compiles even though we have given a hint
let a = [1u32, 1u64];
let b:u32 = a[1];

let mut a = [8u16, 8u16, 18446744073709551615];
log(a[2]); // Logs 2**16 - 1

Note, this issue relates to two other issue reported regarding typing of arrays and typing of literals

## Description This PR fixes: #6376 The issue is that we use the first element of the array as its element type. This can be problematic as the first element is not "concrete". For example, when it is `Numeric`. Now we iterate the array elements, search for the first "good" element, and use its type as the array type. This means: 1 - if the type is concrete, we use it; 2 - if the type is `Numeric`, we keep searching for something better; This allows consistent arrays, but where the concrete types are at the end. ```sway let a = [1, 2u8]; <- we will choose u8 let a = [None, Some(1), Some(2u8)]; <- we will choose Option<u8>, instead of Option<T> ``` ## snapshot tests This PR also brings an experiment. Now when an "e2e" test has a file `snapshot.toml` it will run as `cargo insta` snapshot tests. These tests can be run as normal: `cargo r -p test`, and in two new ways: ``` > cargo t -p test > cargo insta test ``` Snapshots can be reviewed using normal "cargo insta" workflow. The secret lies in https://github.com/LukasKalbertodt/libtest-mimic. **I also want to suggest that all our tests migrate to this model**, where we could just use "cargo test" normally. Apart from that, I opted to use the `forc` executable, which means tests will run in parallel. In the future, if we want to poke compiler internals, we can expose them using flags, or create a special binary to fit our needs. `snapshot.toml` is still empty because I have no idea to what put there at the moment. My initial idea was to configure what ends up into snapshots (multiple runs, compiler flags, script/contract calls) etc... But for the moment, empty is enough. ## Checklist - [x] I have linked to any relevant issues. - [x] I have commented my code, particularly in hard-to-understand areas. - [ ] I have updated the documentation where relevant (API docs, the reference, and the Sway book). - [ ] If my change requires substantial documentation changes, I have [requested support from the DevRel team](https://github.com/FuelLabs/devrel-requests/issues/new/choose) - [x] I have added tests that prove my fix is effective or that my feature works. - [x] I have added (or requested a maintainer to add) the necessary `Breaking*` or `New Feature` labels where relevant. - [x] I have done my best to ensure that my PR adheres to [the Fuel Labs Code Review Standards](https://github.com/FuelLabs/rfcs/blob/master/text/code-standards/external-contributors.md). - [x] I have requested a review from the relevant team or maintainers. --------- Co-authored-by: IGI-111 <igi-111@protonmail.com>

IGI-111 added bug Something isn't working P: medium audit-report Related to the audit report labels Aug 7, 2024

IGI-111 assigned esdrubal and xunilrj and unassigned esdrubal Aug 7, 2024

xunilrj mentioned this issue Aug 16, 2024

Better error array numerics #6420

Merged

8 tasks

IGI-111 closed this as completed in #6420 Aug 26, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implicit Casting of Array Elements #6376

Implicit Casting of Array Elements #6376

IGI-111 commented Aug 7, 2024 •

edited

Loading

Implicit Casting of Array Elements #6376

Implicit Casting of Array Elements #6376

Comments

IGI-111 commented Aug 7, 2024 • edited Loading

IGI-111 commented Aug 7, 2024 •

edited

Loading