A toy web rendering engine written in the Rust language, by Matt Brubeck ([email protected]).
I'm writing this code purely for educational purposes. My goal is to create an incomplete but extremely simple engine as a way to learn more about basic implementation techniques, without worrying about complications like:
Real-world usabilityStandards compliancePerformance and efficiencyInteroperability
These are all important goals, but there are other projects working on them. By ignoring them completely, this project can focus on being as simple and easy-to-understand as possible.
Why create a simple—but useless—toy rendering engine? Mostly because I personally want to learn how to do it. If I succeed, I also hope that other people can learn from my code by reading or modifying it, or learn from my experience as they set out to build their own toy browser engines.
For more details see Let's build a browser engine!, a series of how-to articles based on this project.
Currently implemented:
- Parse a small subset of HTML and build a DOM tree.
- Parse a small subset of CSS.
- Perform selector matching to apply styles to elements.
- Basic block layout.
Coming soon, I hope:
- Inline layout.
- Paint text and boxes.
- Load resources from network or filesystem.
-
Clone the robinson source code from https://github.com/mbrubeck/robinson
-
Run
cargo build
to build robinson, andcargo run
to run it.
To build and run with optimizations enabled, use cargo build --release
and
cargo run --release
.
By default, robinson will load test.html and test.css from the examples
directory. You can use the --html
and --css
arguments to the robinson
executable to change the input files:
./target/debug/robinson --html examples/test.html --css examples/test.css
The rendered page will be saved to a file named output.png
. To change the
output filename, use the -o
option. To switch to PDF output, use add
--format pdf
.