PSP22 chain extension example

[stiiifff]

Example implementation of the PSP22 Fungible Token Standard as a chain extension, supporting a multi-token system provided by the FRAME assets pallet.
It demonstrates how ink! contracts (L2) can interact with native assets (L1) from the chain runtime in a standardised way.

[xgreenx]

It is bad that we did double work=( We could cooperate together to deliver but it is too late.
We did the same in the OpenBrush.

But it is still good to have several implementations=)

We plan to create a production-ready chains extension and define its interface in the PSP. So if you have some though regarding our implementation it will be cool to hear them here or in PSP=)

I want to highlight that maybe some contract wants to transfer his tokens, so it is why we added the Origin enum into the chain extension where the developer can specify it(Origin::Caller or Origin::Self).

But we also realized the problem with Origin::Caller. Allowing anyone to transfer/approve tokens on behalf of the caller without any restriction creates a vulnerability. Any contract can be malicious and can steal all tokens. I think you need to highlight that by commenting on the implementation of transfer in the chain extension.

[paritytech-cicd-pr]

🦑 📈 ink! Example Contracts ‒ Changes Report 📉 🦑

These are the results when building the examples/* contracts from this branch with cargo-contract 1.5.0-ef06f4d and comparing them to ink! master:

	Δ Optimized Size	Δ Used Gas	Total Optimized Size	Total Used Gas
accumulator			1.00 K
adder			2.04 K
contract-terminate			0.92 K	275_000
contract-transfer			8.36 K	75_000
delegate-calls			2.89 K	76_242
delegator			6.34 K	232_138
dns			8.81 K	225_000
erc1155			17.15 K	450_000
erc20			8.42 K	225_000
erc721			11.62 K	600_000
flipper			1.24 K	75_000
forward-calls			2.87 K	151_412
incrementer			1.14 K
mother			12.24 K
multisig			25.79 K	470_483
payment-channel			7.95 K
psp22-extension	+7.16 K		7.16 K
rand-extension			3.79 K	75_000
set-code-hash			1.49 K
subber			2.06 K
trait-erc20			8.69 K	225_000
trait-flipper			0.97 K	75_000
trait-incrementer			1.12 K	150_000
updated-incrementer			9.78 K
upgradeable-flipper			1.48 K

Link to the run | Last update: Tue Aug 23 14:08:39 CEST 2022

[codecov-commenter]

Codecov Report

Merging #1244 (5dad725) into master (2b7f20e) will decrease coverage by 0.40%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1244      +/-   ##
==========================================
- Coverage   70.52%   70.11%   -0.41%     
==========================================
  Files         189      190       +1     
  Lines        5910     5940      +30     
==========================================
- Hits         4168     4165       -3     
- Misses       1742     1775      +33     

Impacted Files	Coverage Δ
crates/metadata/src/layout/mod.rs	73.10% <0.00%> (-1.69%)	⬇️
crates/lang/ir/src/ir/attrs.rs	81.99% <0.00%> (-0.28%)	⬇️
crates/allocator/src/bump.rs	0.00% <0.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[stiiifff]

@xgreenx Thanks for your feedback ! I was unaware of your implementation. This is just a simple, incomplete example. My motivation was to show a more advanced chain extension example than the existing FetchRandomExtension one, and I thought that PSP22 was a good use case. So I don't see it as double work.
Good point re. the caller in the transfer function.

[Maar-io]

hi @stiiifff
Do you still plan to add this as example?
I'd like to use it for one of my workshops

[HCastano]

I'm really sorry for taking so long to look at this! Thanks for your time though, I like the addition of this 🙇

[HCastano]

@bidzyyys Hey, are you going to address the open comments?

[Wiezzel]

@HCastano I will resolve the comments.

[HCastano]

Thanks for updating this @Wiezzel!

Can you fork the substrate-contracts-node, add this chain extension to it, and share the repo? I want to see if this code compiles. We'll need to think of a way to check with our CI in the future

[HCastano]

This won't compile due to some changes with how chain extensions work. You can read more about the breaking changes here: paritytech/substrate#11874.

There's also these two PRs which you might find useful/interesting

• contracts: Composable ChainExtension paritytech/substrate#11816
• contracts: Make chain extension state types public paritytech/substrate#12032

[Wiezzel]

@HCastano I've made some changes according to your comments.
Here's the substarte-contratcs-node code with the extension installed: https://github.com/Wiezzel/substrate-contracts-node/pull/1
CI tests are failing, but it seems to me that is not related to any changes from this PR. I've run cargo test -p ink_lang --test compile_tests on master and it fails as well.

[HCastano]

Here's the substarte-contratcs-node code with the extension installed: https://github.com/Wiezzel/substrate-contracts-node/pull/1

Thanks for that!

CI tests are failing, but it seems to me that is not related to any changes from this PR. I've run cargo test -p ink_lang --test compile_tests on master and it fails as well.

Can you merge master and see if that fixes your issues?

[Wiezzel]

@HCastano All checks passed! 😄

[HCastano]

Couple of small things left but we're almost here 💪

[HCastano]

Please don't leave FIXMEs (or TODOs) in the code. File an issue and we can address it in a follow up PR

[Wiezzel]

OK, I'll open an issue as soon as this PR is merged.

[HCastano]

Thanks for the work on this!