mirror of
https://github.com/elicpeter/nyx.git
synced 2026-06-12 19:55:14 +02:00
Add multi-language AST-pattern scanning support
- Introduced `patterns` module with language-specific vulnerability patterns. - Added `query_cache` utility for caching compiled queries. - Expanded `scan.rs` to support scanning multiple languages dynamically. - Updated `Cargo.toml` with additional tree-sitter dependencies. - Added severity filtering to `ScannerConfig` for better configuration.
This commit is contained in:
parent
0831b9fb48
commit
22369cc404
17 changed files with 665 additions and 25 deletions
79
src/patterns/mod.rs
Normal file
79
src/patterns/mod.rs
Normal file
|
|
@ -0,0 +1,79 @@
|
|||
pub mod rust;
|
||||
pub mod typescript;
|
||||
pub mod javascript;
|
||||
pub mod cpp;
|
||||
pub mod c;
|
||||
mod java;
|
||||
mod go;
|
||||
mod php;
|
||||
mod python;
|
||||
|
||||
use std::collections::HashMap;
|
||||
use serde::{Deserialize, Serialize};
|
||||
use once_cell::sync::Lazy;
|
||||
|
||||
/// How bad / noisy a pattern is considered.
|
||||
#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, PartialOrd)]
|
||||
pub enum Severity {
|
||||
Low,
|
||||
Medium,
|
||||
High,
|
||||
}
|
||||
|
||||
/// One AST pattern with a tree-sitter query and meta-data.
|
||||
#[derive(Debug, Clone, Serialize)]
|
||||
pub struct Pattern {
|
||||
/// Unique identifier (snake-case preferred).
|
||||
pub id: &'static str,
|
||||
/// Human-readable explanation.
|
||||
pub description: &'static str,
|
||||
/// tree-sitter query string.
|
||||
pub query: &'static str,
|
||||
/// Rough severity bucket.
|
||||
pub severity: Severity,
|
||||
}
|
||||
|
||||
/// Global, lazily-initialised registry: lang-name → pattern slice
|
||||
static REGISTRY: Lazy<HashMap<&'static str, &'static [Pattern]>> = Lazy::new(|| {
|
||||
let mut m = HashMap::new();
|
||||
|
||||
// ---- Rust ----
|
||||
m.insert("rust", rust::PATTERNS);
|
||||
|
||||
// ---- TypeScript ----
|
||||
m.insert("typescript", typescript::PATTERNS);
|
||||
m.insert("ts", typescript::PATTERNS);
|
||||
m.insert("tsx", typescript::PATTERNS);
|
||||
|
||||
// ---- JavaScript ----
|
||||
m.insert("javascript", javascript::PATTERNS);
|
||||
m.insert("js", javascript::PATTERNS);
|
||||
|
||||
// ---- C & C++ ----
|
||||
m.insert("c", c::PATTERNS);
|
||||
m.insert("cpp", cpp::PATTERNS);
|
||||
m.insert("c++", cpp::PATTERNS);
|
||||
|
||||
// ---- Other languages in the folder ----
|
||||
m.insert("java", java::PATTERNS);
|
||||
m.insert("go", go::PATTERNS);
|
||||
m.insert("php", php::PATTERNS);
|
||||
m.insert("python", python::PATTERNS);
|
||||
m.insert("py", python::PATTERNS);
|
||||
|
||||
tracing::debug!("AST-pattern registry initialised ({} languages)", m.len());
|
||||
|
||||
m
|
||||
});
|
||||
|
||||
/// Return all patterns for the requested language (case-insensitive).
|
||||
///
|
||||
/// Unknown languages yield an **empty** `Vec`.
|
||||
pub fn load(lang: &str) -> Vec<Pattern> {
|
||||
let key = lang.to_ascii_lowercase();
|
||||
REGISTRY
|
||||
.get(key.as_str())
|
||||
.copied() // `&'static [Pattern]` → *copy* the slice pointer
|
||||
.unwrap_or(&[]) // unknown lang ⇒ empty slice
|
||||
.to_vec() // caller owns the `Vec`
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue