Thanks for your interest in contributing to the Fuel TypeScript SDK!
This document outlines the process for installing dependencies, setting up for development and conventions for contributing.
There are many ways in which you may contribute to the project, some of which involve coding knowledge and some which do not. A few examples include:
- Reporting bugs
- Adding new features or bugfixes for which there is already an open issue
- Making feature requests
Check out our Help Wanted or Good First Issues to find a suitable task.
If you are planning something big, for example, changes related to multiple components or changes to current behaviors, make sure to open an issue to discuss with us before starting on the implementation.
If you find a vulnerability or suspect it may be a security issue, please read our Security Policy and follow the instructions.
If you would like to create an issue, please use the relevant issue template. This will allow us to correctly triage and prioritize it. Every externally submitted issue goes through the following process:
- A new issue is created and is given the
triage
label - It is assigned to a core maintainer for investigation
- Once the assigned core maintainer has completed their investigation, they remove the
triage
label and assign the relevant label to the issue i.e.bug
,feat
,chore
,docs
- The issue is assigned a milestone (e.g.
mainnet
,post-launch
) and a prioritization label wherep0
is the highest priority andp2
is the lowest priority - The issue is assigned for development and should be moved to
In Progress
- A pull request is made ready and the issue is now
In Review
- The pull request needs approval by 3 core maintainers, these can be found in the CODEOWNERS file
- It can then be merged to
master
and included in a release - The issue is closed automatically and it's status is moved to
Done
Note: If additional information is ever required by the assigned investigator then the
awaiting
label will be added to the issue, these means they require more information from the author. Anyawaiting
issue left unanswered for 2 weeks will gostale
and will be closed.
git clone git@github.com:FuelLabs/fuels-ts.git
cd fuels-ts
pnpm install
pnpm build
For building everything in watch-mode, run:
# build all projects in watch-mode
pnpm dev
File watching is done by nodemon
for increased performance.
Check nodemon.config.json
for all settings.
Note: You can
pnpm dev
a single package:cd packages/abi-coder pnpm dev
First, we need to link our fuels
package globally in our local global pnpm store
:
cd fuels-ts/packages/fuels
pnpm link --global
Let's check it out:
pnpm list --global
Cool, now on the root directory of my-local-project
:
cd my-local-project
pnpm link --global fuels
That's it — my-local-project
is now using our local version of fuels
.
The same can be done with all other packages:
cd fuels-ts/packages/wallet
pnpm link --global
# ...
pnpm list --global # validating
# ...
cd my-local-project
pnpm link --global @fuel-ts/wallet
Warning When using local symlinked
fuels-ts
inyour-local-project
, remember topnpm build
the SDK whenever you change a source file to reflect the changes onyour-local-project
. To automate this, you can usepnpm dev
, which will keep watching and compiling everything automatically while developing.
See also:
When you need to work with unreleased versions of forc
or fuel-core
, you can specify a git branch instead of a version number in the respective VERSION
file.
To use an unreleased version of forc
:
-
Edit the
VERSION
File Open theinternal/forc/VERSION
file and replace the version number (e.g.,0.59.0
) with the desired git branch:```text git:some/branch-name ```
-
Install and Build To download and build
forc
from the specified branch after updatingVERSION
file, run:```sh pnpm install pnpm build ```
The following directory will be updated or created: internal/forc/sway-repo
Similarly, to use an unreleased version of fuel-core
:
-
Edit the
VERSION
file: Open theinternal/fuel-core/VERSION
file and replace the version number with the desired git branch:```text git:some/branch-name ```
-
Install and Build To download and build
fuel-core
from the specified branch after updatingVERSION
file, run:```sh pnpm install pnpm build ```
Note
The internal/forc/sway-repo
and internal/fuel-core/fuel-core-repo
directory will contain a local clone of the Sway and Fuel Core repository respectively, at the specific version or git branch you have specified in the internal/forc/VERSION
and internal/fuel-core/VERSION
file.
After running pnpm install
and pnpm build
, the following files and directories are created:
-
Forc Repository:
- Located at
internal/forc/sway-repo
. - Contains the source code for the version of
forc
you are using.
- Located at
-
Fuel-Core Repository:
- Located at
internal/fuel-core/fuel-core-repo
. - Contains the source code for the version of
fuel-core
you are using.
- Located at
If you switch back to using standard binaries you might encounter issues where the SDK still uses the previously downloaded unreleased binaries.
To resolve the issue and ensure the SDK uses the correct binaries:
-
Delete the downloaded repositories
rm -rf internal/forc/sway-repo rm -rf internal/fuel-core/fuel-core-repo
-
Reinstall and build
pnpm install pnpm build
In order to run tests locally, you can run the following commands:
# run pretest to ensure all test dependencies are built
pnpm pretest
# run all tests in a node environment
pnpm test
# you may also run tests in a browser environment
pnpm test:browser
# watch all tests
pnpm test:watch
# run tests for a specific package
pnpm test:filter packages/my-desired-package
# run tests for a specific file
pnpm test:filter packages/my-desired-package/src/my.test.ts
# run tests while passing other flags to sub-program
pnpm test -- --coverage --my-other-flag
We currently use vitest
's bench utility to run benchmarks. You can run them in both the browser and node environments.
pnpm bench:node
# run benchmarks for a specific package
pnpm bench:node packages/my-desired-package
We currently use clinic
to profile and debug our tooling. For instance you can run clinic's flame command to create a flamegraph for a specific package:
# creates a flamegraph for a specific package
npm_config_package_name=account pnpm clinic:flame // runs flame against the account package
During the CI process an automated end-to-end (e2e) test is executed. This test is crucial as it simulates real-world scenarios on the current test-net, ensuring that the changeset maintains the expected functionality and stability.
The e2e test can be found at:
packages/fuel-gauge/src/e2e-script.test.ts
The Bech32 address of this wallet is fuel1x33ajpj0jy5p2wcqqu45e32r75zrwfeh6hwqfv5un670rv4p0mns58enjg
. This address can be funded via the faucet.
[!NOTE] Note
Bech32
addresses likefuel1..
are now deprecated. UseB256
addresses instead. (help)
If you want to run an e2e test locally, you can provide your own wallet address and private key. For obvious security reasons, the private key should not be shared.
These can be overridden by generating an environment file:
cp .env.example .env.test
And changing the below variables:
DEVNET_WALLET_PVT_KEY=0x...
TESTNET_WALLET_PVT_KEY=0x...
This will enable you to run the e2e test locally against the live network:
pnpm test:filter e2e-script
Before you create a Pull Request, please check whether your commits comply with the commit conventions used in this repository.
When you create a commit we kindly ask you to follow the convention
category(scope or module): message
in your commit message while using one of
the following categories:
feat / feature
: all changes that introduce completely new code or new featuresfix
: changes that fix a bug (ideally you will additionally reference an issue if present)refactor
: any code related change that is not a fix nor a featuredocs
: changing existing or creating new documentation (i.e. README, docs for usage of a lib or cli usage)build
: all changes regarding the build of the software, changes to dependencies or the addition of new dependenciestest
: all changes regarding tests (adding new tests or changing existing ones)ci
: all changes regarding the configuration of continuous integration (i.e. github actions, ci system)chore
: all changes to the repository that do not fit into any of the above categories
-
Fork the fuels-ts repository and clone your fork
-
Create a new branch out of the
master
branch with the naming convention<username>/<fix|feat|chore|build|docs>/<branch-name>
. -
Make and commit your changes following the commit convention. As you develop, you can run
pnpm build
andpnpm test
to make sure everything works as expected. -
Run
pnpm changeset
to create a detailed description of your changes. This will be used to generate a changelog when we publish an update. Learn more about Changeset. Please note that you might have to rungit fetch origin master
(where origin will be your fork on GitHub) beforepnpm changeset
works. -
We adhere to the Conventional Commits specification for semantic versioning, and given that we currently do not have a major release yet, only breaking changes would require a
minor
version bump. All other API compatible changes would be apatch
version bump.
If you made minor changes like CI config, prettier, etc, you can run
pnpm changeset add --empty
to generate an empty changeset file to document your changes.
The SDK utilizes a pre-push git hook to validate your contribution before review. This is a script that will run automatically before changes are pushed to the remote repository. Within the SDK, the pre-push script will run code linting.
This can be overridden using the
--no-verify
flag when pushing.
The following script will upgrade Forc to the latest version on GitHub, remove all lockfiles so the latest stdlib can be used, and rebuild all projects:
pnpm forc:update
After this you should run tests and fix any incompatibilities.
Manually edit the internal/fuel-core/VERSION
file, add the right version, and then:
pnpm install # will download new binaries
pnpm test
If all tests pass, that's it.
Otherwise, you have work to do.
The following example is for releasing a patch for v0.69.0
-> v0.69.1
.
- Checkout the release commit via its tag and create a release branch based on it (
git checkout -b release/0.69.0 v0.69.0 && git push --set-upstream origin release/0.69.0
) - Create PRs with base set to that release branch
- When the PR is merged, a changeset PR is created
- When the changeset PR is merged into the release branch, the next patch version is released and the commit is tagged (e.g.
v0.69.1
)
- After release, the release branch will be automatically deleted
Imagine the scenario:
- We release
v0.80.0
- One day later, we have a new changesets PR that will bump things to
v0.81.0
- Before releasing
v0.81.0
, we find an issue and need to make av0.80.1
patch
We'd follow the same approach as explained in the Patching old releases section above, bearing in mind the following after the release:
- A PR merging the
latest
release's branch intomaster
will be automatically created, - The automatically-created PR must be merged as soon as possible in order to
- have the versions of packages on
master
match thelatest
released package versions, - have the released functionality on
master
as well
- have the versions of packages on
In order to make the SDK for Fuel feel familiar with those coming from the ethers.js ecosystem, this project opted for an s
at the end.
The fuels-*
family of SDKs is inspired by The Ethers Project.