It is what it says it is.
However, this bundler aims to be a monolithic does-it-all type of bundler; with zero configuration needed. (however, optional configuration will be supported)
- A bundler should not need excessive attention.
- A bundler should not get in the way of a developers workflow.
- A bundler should be fast.
- A bundler should not be written in Javascript / Node.
- Commonly used features among JS developers should be supported out of the box.
- The output of the bundler should work everywhere. (to an extent which is possible of course)
Most bundlers allows you to select an output format of the code, such as:
"CommonJS", "AMD", "System", "UMD", "ES6", "ES2015" or "ESNext"
FJB does not implement this concept.
Instead, FJB adapts the code automatically through static analysis.
FJB aims to generate code that works everywhere (when possible).
A bundler should be smart enough so that the developer can focus on other things. π§
fjb index.js
- Being fast
- ES6 module imports / exports
- Aliased imports
- Wildcard imports
- JSX support
- Importing JSON
- Importing CSS
- Removing dead code (Tree shaking)
- Supports package.json
- Importing anything, even when it's not exported, no matter how deeply nested it is
- Plugin support
- TypeScript support
- Lazy imports
- Better formatted output. (kind of ugly right now)
Below is an example of some code that is 100% supported.
(Without any plugins or configuration)
For more examples, have a look at:
To see benchmarks, have a look at benchmarks.md π₯
To build
fjb
you will need the following:
- A C compiler (I would suggest gcc)
- Git
- Clone down the repository
- Run:
./build.sh
This step will generate a
fjb.out
executable file.
You can start using this right away. π©
To continue and install
fjb
globally: Run:
make install
Also make sure you add
$HOME/.local/bin
to your PATH.
https://unix.stackexchange.com/questions/26047/how-to-correctly-add-a-path-to-path
Have a look at benchmarks.md π₯
Not sure, it's just a hobby project at this point.
Frustration, frustration with existing tools.
The existing alternatives are not fast enough, and gets in the way a bit more than they should.
At some point, I want this project to get there.
But right now, it lacks some features.
For example, fjb currently supports JSX; but react-jsx is not implemented
right now.
I will not encourage anyone to use it, but feel free to do it.
Yes
Probably not, but feel free to make a pull-request with the required changes for it to work.
Probably. I haven't tried it, but I would definitely think so.
Yes you can. Simply fork it and start making pull-requests.
It's not possible at this point.
This is a feature I plan to implement.
If you want some other behaviour right now, simply just change the source code.
I'm thinking about shared library files, such as
.so
,.a
,.dll
... etc.
No. I don't think Javascript belongs in a bundler.
The reason why most bundlers today are slow, is because they are written in Javascript.
If you find a bug, please report it here: https://github.com/sebbekarlsson/fjb/issues
The process of implementing this will come soon,
it's important that we have a stable foundation first.
GPL-3.0, LICENSE.md
Feel free to create an issue here, and I'll be able to help you.
I would prefer if you make pull-requests through forking at first.
Once I've seen that you've made valuable contributions, you might be able to be added as a collaborator.