Expand description
§tabulon
A high-performance, JIT-compiled expression evaluation engine for Rust, built on Cranelift.
tabulon parses simple arithmetic and boolean expressions and compiles them to native machine code at runtime. It is designed for applications like game servers, configuration scripts, or anywhere you need to safely and repeatedly evaluate user-provided expressions with maximum performance.
§Features
- High-Performance JIT Compilation: Expressions are compiled to native machine code for near-native performance, powered by Cranelift.
- Rich Operator Support: Full support for arithmetic (
+,-,*,/), comparison (==,!=,<,<=,>,>=), and logical (&&,||) operators. - Efficient & Safe:
- Short-circuiting for
if(...),&&, and||operators avoids unnecessary computation. - AST-level optimizations like constant folding are performed automatically.
- All execution is sandboxed within the JIT engine.
- Short-circuiting for
- Extensible: Register your own custom Rust functions to be called from within an expression.
- Flexible: Use custom resolvers to map variables to your application’s specific data structures.
§Quick Start
Use the engine to compile and evaluate an expression:
use tabulon::Tabula;
fn main() -> Result<(), Box<dyn std::error::Error>> {
// Create a new engine instance
let mut engine = Tabula::new();
// Compile an expression
let expr = engine.compile("power > 9000 && is_angry * 10")?;
// `ordered_vars` shows the order variables must be supplied in.
// Here it would be: ["power", "is_angry"]
assert_eq!(expr.vars(), &["power", "is_angry"]);
// Prepare variables for evaluation
let power_level = 9001.0;
let is_angry_val = 1.0; // Use 1.0 for true, 0.0 for false
// `eval` takes a slice of pointers to your f64 variables
let result = expr.eval(&[&power_level, &is_angry_val])?;
// The expression "(9001 > 9000) && (1.0 * 10)" is true, so the result is 1.0.
assert_eq!(result, 1.0);
println!("It's over 9000! Result: {}", result);
Ok(())
}§Custom Functions
You can easily register your own Rust functions to be used in expressions.
use tabulon::{Tabula, function, register_functions};
// Use the `#[function]` attribute to make a function discoverable.
#[function]
pub fn custom_max(a: f64, b: f64) -> f64 {
if a > b { a } else { b }
}
fn main() -> Result<(), Box<dyn std::error::Error>> {
let mut engine = Tabula::new();
// Register the function with the engine instance.
register_functions!(engine, custom_max)?;
let expr = engine.compile("custom_max(player_score, high_score) * 1.1")?;
let player_score = 120.0;
let high_score = 150.0;
let result = expr.eval(&[&player_score, &high_score])?;
assert_eq!(result, 165.0);
println!("Potential new high score: {}", result);
Ok(())
}§Built-in Functions and Operators
tabulon supports a rich set of built-in operators and functions.
§Operators
| Category | Operators | Description |
|---|---|---|
| Arithmetic | +, -, *, / | Addition, Subtraction, Multiplication, Division. |
| Comparison | ==, !=, >, >=, <, <= | Equal, Not Equal, Greater Than, Greater/Equal, Less Than, Less/Equal. |
| Logical | && (and), || (or) | Evaluates logical AND and OR. Both operators are short-circuiting. |
| Unary | - | Negates a value (e.g., -x). |
§Built-in Functions & Constructs
Boolean logic in tabulon follows the convention where 1.0 is true and 0.0 is false.
| Name | Signature | Description |
|---|---|---|
| if | if(condition, then_expr, else_expr) | If condition evaluates to true (1.0), returns then_expr, otherwise returns else_expr. This is short-circuiting. |
| ifs | ifs(cond1, then1, cond2, then2, ..., else_val) | Evaluates multiple conditions in order. Returns the value corresponding to the first true condition. If all are false, returns else_val. |
| min | min(a, b) | Returns the smaller of the two numbers a and b. |
| max | max(a, b) | Returns the larger of the two numbers a and b. |
§Examples
use tabulon::Tabula;
let mut engine = Tabula::new();
// Using the `if` function
let expr_if = engine.compile("if(health > 50, 100, 20)").unwrap();
let health = 75.0;
assert_eq!(expr_if.eval(&[&health]).unwrap(), 100.0);
// Using `ifs` for multiple conditions
let expr_ifs = engine.compile("ifs(score > 90, 1, score > 50, 0.5, 0)").unwrap();
let score = 70.0;
assert_eq!(expr_ifs.eval(&[&score]).unwrap(), 0.5);
// Using `min` and `max`
let expr_minmax = engine.compile("min(max(a, b), 100)").unwrap();
let a = 50.0;
let b = 120.0;
assert_eq!(expr_minmax.eval(&[&a, &b]).unwrap(), 100.0);§Performance
tabulon is designed for speed. By compiling expressions down to a few simple machine instructions, it can evaluate them orders of magnitude faster than a tree-walking interpreter. Benchmarks are included in the repository (cargo bench).
§License
This project is licensed under the MIT License.
Re-exports§
pub use inventory;
Macros§
Structs§
- Compiled
Expr - A compiled, executable expression that owns its variable map.
- Compiled
Expr Ref - A compiled, executable expression that is evaluated via references or pointers.
- FnMeta
- Identity
Resolver - The default
VarResolverthat treats variable names as their own keys. - Tabula
- The main JIT compilation and evaluation engine.
Enums§
- JitError
- The primary error type for the
tabuloncrate. - Registered
Fn - VarResolve
Error - An error type returned by a
VarResolverwhen it fails to resolve an identifier.
Traits§
- VarResolver
- A trait for resolving variable names from an expression string into a custom key type
K.