Supported Languages

Omen supports 13 programming languages through tree-sitter grammars. Language detection is automatic based on file extension. All parsing is syntax-aware -- analyzers operate on AST nodes, not text patterns.

Languages and File Extensions

Language	Extensions
Go	.go
Rust	.rs
Python	.py, .pyi
TypeScript	.ts
JavaScript	.js, .mjs, .cjs
TSX	.tsx
JSX	.jsx
Java	.java
C	.c, .h
C++	.cpp, .cc, .cxx, .hpp
C#	.cs
Ruby	.rb
PHP	.php
Bash	.sh, .bash

Files with unrecognized extensions are silently skipped. There is no heuristic-based detection.

Filtering by Language

Use the --language flag to restrict analysis to a single language:

omen complexity --language rust
omen clones --language typescript
omen satd --language python

Analyzer Coverage by Language

Not all analyzers apply equally to all languages. Some analyzers are universal (they work on any language with a tree-sitter grammar), while others have language-specific behavior or are limited to a subset of languages.

Coverage Matrix

Analyzer	Go	Rust	Python	TS	JS	TSX/JSX	Java	C	C++	C#	Ruby	PHP	Bash
Complexity	x	x	x	x	x	x	x	x	x	x	x	x	x
SATD	x	x	x	x	x	x	x	x	x	x	x	x	x
Dead Code	x	x	x	x	x	x	x	x	x	x	x	x	x
Churn	x	x	x	x	x	x	x	x	x	x	x	x	x
Code Clones	x	x	x	x	x	x	x	x	x	x	x	x	x
TDG	x	x	x	x	x	x	x	x	x	x	x	x	x
Hotspots	x	x	x	x	x	x	x	x	x	x	x	x	x
Repo Map	x	x	x	x	x	x	x	x	x	x	x	x	x
Smells	x	x	x	x	x	x	x	x	x	x	x	x	x
Defect Prediction	x	x	x	x	x	x	x	x	x	x	x	x	x
Change Risk	x	x	x	x	x	x	x	x	x	x	x	x	x
Ownership	x	x	x	x	x	x	x	x	x	x	x	x	x
Temporal Coupling	x	x	x	x	x	x	x	x	x	x	x	x	x
Diff Analysis	x	x	x	x	x	x	x	x	x	x	x	x	x
Semantic Search	x	x	x	x	x	x	x	x	x	x	x	x	x
CK Metrics	-	-	x	-	-	-	x	-	x	x	x	-	-
Dependency Graph	x	x	x	x	-	-	x	-	-	-	x	-	-
Mutation Testing	x	x	-	x	-	-	-	-	-	-	x	-	-
Feature Flags	-	-	x	x	x	x	-	-	-	-	x	-	-

Legend: x = supported, - = not applicable or not yet implemented.

Universal Analyzers

The following analyzers work identically across all 13 languages. They rely on general tree-sitter node types (functions, classes, comments, blocks) or on Git history rather than language-specific constructs:

• Complexity -- cyclomatic and cognitive complexity, nesting depth
• SATD -- comment scanning for technical debt markers
• Dead Code -- unreachable functions and unused exports
• Churn -- file change frequency from Git history
• Code Clones -- token-based duplicate detection
• TDG -- technical debt gradient (composite of multiple signals)
• Hotspots -- intersection of complexity and churn
• Repository Map -- structural map of modules and symbols
• Smells -- architectural smell detection
• Defect Prediction -- risk scoring based on complexity, churn, and ownership
• Change Risk -- risk analysis of recent modifications
• Ownership -- contributor distribution from Git blame
• Temporal Coupling -- co-change detection from Git history
• Diff Analysis -- structural analysis of uncommitted changes
• Semantic Search -- embedding-based code discovery

CK Metrics (Cohesion)

CK metrics (Chidamber-Kemerer) are object-oriented metrics that measure class-level properties: weighted methods per class (WMC), coupling between objects (CBO), response for a class (RFC), lack of cohesion (LCOM4), depth of inheritance tree (DIT), and number of children (NOC).

These metrics are meaningful only for languages with class-based OO semantics:

• Python -- classes defined with class
• Java -- classes, interfaces, abstract classes
• C++ -- classes and structs with methods
• C# -- classes, structs, interfaces
• Ruby -- classes and modules

Languages without class constructs (Go, C, Bash) or where classes are uncommon in idiomatic usage (Rust traits/structs without inheritance, PHP historically procedural) are not analyzed for CK metrics.

Dependency Graph

The dependency graph analyzer parses import and module statements to build a directed graph of file dependencies. This requires language-specific understanding of the import system:

• Go -- import statements
• Python -- import and from ... import statements
• TypeScript -- import and require statements
• Rust -- use and mod statements
• Java -- import statements
• Ruby -- require and require_relative statements

Languages where the import mechanism is not yet implemented (C/C++ #include, C# using, PHP use/require, JavaScript without TypeScript, Bash source) are excluded from dependency graph analysis.

Mutation Testing

Mutation testing injects controlled modifications into source code and runs the project's test suite to evaluate test effectiveness. The mutations are language-specific, targeting operators and constructs that are syntactically and semantically meaningful in each language:

Language	Operators	Description
Rust	3	Boundary mutations (< to <=), negation removal, return value substitution
Go	2	Boundary mutations, conditional negation
TypeScript	2	Boundary mutations, boolean literal flipping
Python	2	Comparison operator swaps, boolean negation
Ruby	2	Comparison operator swaps, conditional negation

Mutation testing requires a working test suite and a test command. It is not currently supported for C, C++, C#, Java, PHP, JavaScript (without TypeScript), JSX, TSX, or Bash.

Feature Flag Detection

Feature flag detection identifies usage of feature flag SDKs and custom flag patterns in code. Detection is provider-specific and currently supports:

Language	Providers
JavaScript/TypeScript/TSX/JSX	LaunchDarkly (ldclient, @launchdarkly/node-server-sdk), Split (@splitsoftware/splitio), Unleash (unleash-client)
Python	Unleash (UnleashClient)
Ruby	Flipper (Flipper.enabled?), environment variable flags (ENV["FEATURE_*"], ENV.fetch("FEATURE_*"))

Additional providers can be defined through custom tree-sitter queries in the configuration file. See Configuration for details on the [feature_flags.custom_providers] section.

Adding Language Support

Omen's language support is determined by its compiled tree-sitter grammars. Adding a new language requires adding the corresponding tree-sitter grammar crate as a dependency, implementing the extension-to-grammar mapping, and ensuring the universal analyzers handle the new grammar's node types correctly. Language-specific analyzers (dependency graph, mutation testing, feature flags) require additional per-language implementation.