refactor: Break up oversized functions in codegen.rs #215

New issue

Closed

opened 2026-01-09 04:43:25 +00:00 by navicore · 1 comment

navicore commented

2026-01-09 04:43:25 +00:00

(Migrated from github.com)

Problem

crates/compiler/src/codegen.rs contains multiple functions exceeding 100 lines, making them difficult to understand, test, and maintain.

Oversized Functions

Function	Lines	Location	Primary Issues
`codegen_program_with_config`	~324	820-1144	Mixed initialization, setup, and codegen
`emit_runtime_declarations`	~245	1298-1543	100+ sequential writeln! calls
`codegen_match_statement`	~117	5269-5386	Deep nesting, multiple responsibilities
`codegen_inline_binary_op`	~111	4079-4190	Fast/slow path mixing

Recommended Splits

`codegen_program_with_config` (324 lines → 4 functions)

fn codegen_program_with_config(&mut self, config: Config) -> Result<String> {
    let context = self.initialize_codegen_context(config)?;
    self.emit_preamble(&context)?;
    self.emit_word_definitions(&context)?;
    self.emit_main_entry(&context)?;
    self.finalize_output(&context)
}

fn initialize_codegen_context(&mut self, config: Config) -> Result<CodegenContext>;
fn emit_preamble(&mut self, ctx: &CodegenContext) -> Result<()>;
fn emit_word_definitions(&mut self, ctx: &CodegenContext) -> Result<()>;
fn emit_main_entry(&mut self, ctx: &CodegenContext) -> Result<()>;

`emit_runtime_declarations` (245 lines → data-driven)

See related issue for extracting to data-driven approach.

`codegen_match_statement` (117 lines → 3 functions)

fn codegen_match_statement(&mut self, stmt: &MatchStmt) -> Result<()> {
    let scrutinee = self.codegen_match_scrutinee(stmt)?;
    self.codegen_match_arms(stmt, scrutinee)?;
    self.codegen_match_epilogue(stmt)
}

`codegen_inline_binary_op` (111 lines → 2 functions)

fn codegen_inline_binary_op(&mut self, op: BinaryOp) -> Result<()> {
    if self.can_use_fast_path() {
        self.codegen_inline_binary_op_virtual(op)
    } else {
        self.codegen_inline_binary_op_memory(op)
    }
}

Guidelines for Function Size

Target: Functions should be < 50 lines
Maximum: No function should exceed 80 lines
Exception: Data definitions (tables, constants) can be longer

Acceptance Criteria

No function exceeds 80 lines
Each function has single clear purpose
Function names describe what they do
All tests pass
Code coverage maintained

Labels

refactor, technical-debt, high-priority

## Problem `crates/compiler/src/codegen.rs` contains multiple functions exceeding 100 lines, making them difficult to understand, test, and maintain. ## Oversized Functions | Function | Lines | Location | Primary Issues | |----------|-------|----------|----------------| | `codegen_program_with_config` | ~324 | 820-1144 | Mixed initialization, setup, and codegen | | `emit_runtime_declarations` | ~245 | 1298-1543 | 100+ sequential writeln! calls | | `codegen_match_statement` | ~117 | 5269-5386 | Deep nesting, multiple responsibilities | | `codegen_inline_binary_op` | ~111 | 4079-4190 | Fast/slow path mixing | ## Recommended Splits ### `codegen_program_with_config` (324 lines → 4 functions) ```rust fn codegen_program_with_config(&mut self, config: Config) -> Result<String> { let context = self.initialize_codegen_context(config)?; self.emit_preamble(&context)?; self.emit_word_definitions(&context)?; self.emit_main_entry(&context)?; self.finalize_output(&context) } fn initialize_codegen_context(&mut self, config: Config) -> Result<CodegenContext>; fn emit_preamble(&mut self, ctx: &CodegenContext) -> Result<()>; fn emit_word_definitions(&mut self, ctx: &CodegenContext) -> Result<()>; fn emit_main_entry(&mut self, ctx: &CodegenContext) -> Result<()>; ``` ### `emit_runtime_declarations` (245 lines → data-driven) See related issue for extracting to data-driven approach. ### `codegen_match_statement` (117 lines → 3 functions) ```rust fn codegen_match_statement(&mut self, stmt: &MatchStmt) -> Result<()> { let scrutinee = self.codegen_match_scrutinee(stmt)?; self.codegen_match_arms(stmt, scrutinee)?; self.codegen_match_epilogue(stmt) } ``` ### `codegen_inline_binary_op` (111 lines → 2 functions) ```rust fn codegen_inline_binary_op(&mut self, op: BinaryOp) -> Result<()> { if self.can_use_fast_path() { self.codegen_inline_binary_op_virtual(op) } else { self.codegen_inline_binary_op_memory(op) } } ``` ## Guidelines for Function Size - **Target**: Functions should be < 50 lines - **Maximum**: No function should exceed 80 lines - **Exception**: Data definitions (tables, constants) can be longer ## Acceptance Criteria - [ ] No function exceeds 80 lines - [ ] Each function has single clear purpose - [ ] Function names describe what they do - [ ] All tests pass - [ ] Code coverage maintained ## Labels refactor, technical-debt, high-priority

navicore commented

2026-01-10 18:32:47 +00:00

(Migrated from github.com)

https://github.com/navicore/patch-seq/pull/225