feat(joinir): Phase 171 complete - Trim pattern LoopBodyLocal promotion
Phase 171-C-4/5 + impl-Trim: Full Trim pattern validation infrastructure ## CarrierInfo 統合 (C-4) - CarrierInfo::merge_from(): Deduplicated carrier merging - TrimPatternInfo::to_carrier_info(): Conversion helper - Pattern2/4: Promoted carrier merge integration - 7 unit tests for merge logic ## TrimLoopHelper 設計 (C-5) - TrimLoopHelper struct: Trim-specific validation box - carrier_type(), initial_value(), whitespace helpers - CarrierInfo::trim_helper() accessor - 5 unit tests ## Validation-Only Integration (impl-Trim) - TrimLoopHelper::is_safe_trim(), is_trim_like(), has_valid_structure() - Pattern2/4: Trim exception route with safety validation - body_locals extraction from loop body AST - LoopBodyCarrierPromoter: ASTNode::Local handler extension - 4 unit tests for safety validation ## Architecture - Box Theory: TrimLoopHelper is "validation only" (no JoinIR generation) - Fail-Fast: Non-Trim LoopBodyLocal immediately rejected - Whitelist approach: Only Trim pattern bypasses LoopBodyLocal restriction Tests: 16 new unit tests, all passing E2E: test_trim_main_pattern.hako validation successful Next: Phase 172 - Actual JoinIR lowering for Trim pattern 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
nyash-codex
@ -47,6 +47,45 @@ pub struct TrimPatternInfo {
|
||||
pub carrier_name: String,
|
||||
}
|
||||
|
||||
impl TrimPatternInfo {
|
||||
/// Phase 171-C-4: Convert to CarrierInfo with a bool carrier for the pattern
|
||||
///
|
||||
/// Creates a CarrierInfo containing a single bool carrier representing
|
||||
/// the Trim pattern match condition (e.g., "is_whitespace").
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// CarrierInfo with:
|
||||
/// - loop_var_name: The promoted carrier name (e.g., "is_ch_match")
|
||||
/// - loop_var_id: Placeholder ValueId(0) (will be remapped by JoinInlineBoundary)
|
||||
/// - carriers: Empty (the carrier itself is the loop variable)
|
||||
///
|
||||
/// # Design Note
|
||||
///
|
||||
/// The returned CarrierInfo uses a placeholder ValueId(0) because:
|
||||
/// - This is JoinIR-local ID space (not host ValueId space)
|
||||
/// - The actual host ValueId will be assigned during merge_joinir_mir_blocks
|
||||
/// - JoinInlineBoundary will handle the boundary mapping
|
||||
pub fn to_carrier_info(&self) -> crate::mir::join_ir::lowering::carrier_info::CarrierInfo {
|
||||
use crate::mir::join_ir::lowering::carrier_info::CarrierInfo;
|
||||
use crate::mir::ValueId;
|
||||
use super::trim_loop_helper::TrimLoopHelper;
|
||||
|
||||
// Phase 171-C-4/5: Create CarrierInfo with promoted carrier as loop variable
|
||||
// and attach TrimLoopHelper for future lowering
|
||||
let mut carrier_info = CarrierInfo::with_carriers(
|
||||
self.carrier_name.clone(), // "is_ch_match" becomes the loop variable
|
||||
ValueId(0), // Placeholder (will be remapped)
|
||||
vec![], // No additional carriers
|
||||
);
|
||||
|
||||
// Phase 171-C-5: Attach TrimLoopHelper for pattern-specific lowering logic
|
||||
carrier_info.trim_helper = Some(TrimLoopHelper::from_pattern_info(self));
|
||||
|
||||
carrier_info
|
||||
}
|
||||
}
|
||||
|
||||
/// 昇格結果
|
||||
pub enum PromotionResult {
|
||||
/// 昇格成功: Trim パターン情報を返す
|
||||
@ -204,6 +243,18 @@ impl LoopBodyCarrierPromoter {
|
||||
}
|
||||
}
|
||||
|
||||
// Phase 171-impl-Trim: Handle Local with initial values
|
||||
// local ch = s.substring(...)
|
||||
ASTNode::Local { variables, initial_values, .. } if initial_values.len() == variables.len() => {
|
||||
for (i, var) in variables.iter().enumerate() {
|
||||
if var == var_name {
|
||||
if let Some(Some(init_expr)) = initial_values.get(i) {
|
||||
return Some(init_expr.as_ref());
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// その他のノードは無視
|
||||
_ => {}
|
||||
}
|
||||
|
||||
@ -761,3 +761,7 @@ pub mod error_messages;
|
||||
|
||||
// Phase 171-C: LoopBodyLocal Carrier Promotion
|
||||
pub mod loop_body_carrier_promoter;
|
||||
|
||||
// Phase 171-C-5: Trim Pattern Helper
|
||||
pub mod trim_loop_helper;
|
||||
pub use trim_loop_helper::TrimLoopHelper;
|
||||
|
||||
338
src/mir/loop_pattern_detection/trim_loop_helper.rs
Normal file
338
src/mir/loop_pattern_detection/trim_loop_helper.rs
Normal file
@ -0,0 +1,338 @@
|
||||
//! Phase 171-C-5: TrimLoopHelper - Trim Pattern Lowering Helper
|
||||
//!
|
||||
//! This module provides a helper struct for Trim pattern lowering.
|
||||
//! It encapsulates pattern-specific logic for converting LoopBodyLocal-based
|
||||
//! conditions to carrier-based conditions.
|
||||
//!
|
||||
//! ## Purpose
|
||||
//!
|
||||
//! When a Trim pattern is detected (e.g., trim leading/trailing whitespace),
|
||||
//! the LoopBodyCarrierPromoter promotes the LoopBodyLocal variable (like `ch`)
|
||||
//! to a bool carrier (like `is_whitespace`).
|
||||
//!
|
||||
//! TrimLoopHelper stores the pattern information needed to:
|
||||
//! 1. Generate carrier initialization code
|
||||
//! 2. Generate carrier update code
|
||||
//! 3. Map the promoted carrier back to the original variable semantics
|
||||
//!
|
||||
//! ## Example Use Case
|
||||
//!
|
||||
//! **Original pattern**:
|
||||
//! ```nyash
|
||||
//! loop(start < end) {
|
||||
//! local ch = s.substring(start, start+1)
|
||||
//! if ch == " " || ch == "\t" { start = start + 1 } else { break }
|
||||
//! }
|
||||
//! ```
|
||||
//!
|
||||
//! **After promotion**:
|
||||
//! - Original variable: `ch`
|
||||
//! - Promoted carrier: `is_whitespace` (bool)
|
||||
//! - Comparison literals: `[" ", "\t"]`
|
||||
//!
|
||||
//! **TrimLoopHelper usage**:
|
||||
//! ```rust
|
||||
//! let helper = TrimLoopHelper {
|
||||
//! original_var: "ch".to_string(),
|
||||
//! carrier_name: "is_whitespace".to_string(),
|
||||
//! whitespace_chars: vec![" ".to_string(), "\t".to_string()],
|
||||
//! };
|
||||
//!
|
||||
//! // Generate carrier initialization: is_whitespace = true
|
||||
//! let init_value = helper.initial_value(); // true
|
||||
//!
|
||||
//! // Generate carrier update: is_whitespace = (ch == " " || ch == "\t")
|
||||
//! let carrier_type = helper.carrier_type(); // "Bool"
|
||||
//! ```
|
||||
|
||||
use super::loop_body_carrier_promoter::TrimPatternInfo;
|
||||
|
||||
/// Helper for Trim pattern lowering
|
||||
///
|
||||
/// Encapsulates the pattern-specific logic for converting
|
||||
/// LoopBodyLocal-based conditions to carrier-based conditions.
|
||||
///
|
||||
/// # Fields
|
||||
///
|
||||
/// * `original_var` - The original LoopBodyLocal variable name (e.g., "ch")
|
||||
/// * `carrier_name` - The promoted carrier name (e.g., "is_whitespace")
|
||||
/// * `whitespace_chars` - The whitespace characters to compare against (e.g., [" ", "\t", "\n", "\r"])
|
||||
///
|
||||
/// # Design Philosophy
|
||||
///
|
||||
/// This struct follows Box Theory principles:
|
||||
/// - **Single Responsibility**: Only handles Trim pattern lowering logic
|
||||
/// - **Reusability**: Can be used by both Pattern2 and Pattern4 lowerers
|
||||
/// - **Testability**: Pure data structure with simple accessors
|
||||
#[derive(Debug, Clone)]
|
||||
pub struct TrimLoopHelper {
|
||||
/// The original variable name (e.g., "ch")
|
||||
pub original_var: String,
|
||||
|
||||
/// The promoted carrier name (e.g., "is_whitespace")
|
||||
pub carrier_name: String,
|
||||
|
||||
/// Whitespace characters to compare against (e.g., [" ", "\t", "\n", "\r"])
|
||||
pub whitespace_chars: Vec<String>,
|
||||
}
|
||||
|
||||
impl TrimLoopHelper {
|
||||
/// Create TrimLoopHelper from TrimPatternInfo
|
||||
///
|
||||
/// # Arguments
|
||||
///
|
||||
/// * `info` - The TrimPatternInfo from LoopBodyCarrierPromoter
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// A new TrimLoopHelper with the same information
|
||||
///
|
||||
/// # Example
|
||||
///
|
||||
/// ```ignore
|
||||
/// let trim_info = TrimPatternInfo {
|
||||
/// var_name: "ch".to_string(),
|
||||
/// comparison_literals: vec![" ".to_string(), "\t".to_string()],
|
||||
/// carrier_name: "is_whitespace".to_string(),
|
||||
/// };
|
||||
///
|
||||
/// let helper = TrimLoopHelper::from_pattern_info(&trim_info);
|
||||
/// assert_eq!(helper.original_var, "ch");
|
||||
/// assert_eq!(helper.carrier_name, "is_whitespace");
|
||||
/// ```
|
||||
pub fn from_pattern_info(info: &TrimPatternInfo) -> Self {
|
||||
TrimLoopHelper {
|
||||
original_var: info.var_name.clone(),
|
||||
carrier_name: info.carrier_name.clone(),
|
||||
whitespace_chars: info.comparison_literals.clone(),
|
||||
}
|
||||
}
|
||||
|
||||
/// Get the carrier type (always Bool for Trim pattern)
|
||||
///
|
||||
/// Trim patterns always use bool carriers to represent
|
||||
/// "does the character match the whitespace set?"
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// "Bool" - the carrier type name
|
||||
pub fn carrier_type(&self) -> &str {
|
||||
"Bool"
|
||||
}
|
||||
|
||||
/// Get initial carrier value (true = continue looping)
|
||||
///
|
||||
/// The carrier is initialized to `true` to represent
|
||||
/// "keep looping initially". When the character doesn't match
|
||||
/// whitespace, the carrier becomes `false` and the loop breaks.
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// `true` - initial value for the carrier
|
||||
pub fn initial_value(&self) -> bool {
|
||||
true
|
||||
}
|
||||
|
||||
/// Get the number of whitespace characters in the comparison set
|
||||
///
|
||||
/// Useful for diagnostics and code generation.
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// The count of whitespace characters (e.g., 4 for [" ", "\t", "\n", "\r"])
|
||||
pub fn whitespace_count(&self) -> usize {
|
||||
self.whitespace_chars.len()
|
||||
}
|
||||
|
||||
/// Check if a specific character is in the whitespace set
|
||||
///
|
||||
/// # Arguments
|
||||
///
|
||||
/// * `ch` - The character to check (as a string slice)
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// `true` if the character is in the whitespace set, `false` otherwise
|
||||
///
|
||||
/// # Example
|
||||
///
|
||||
/// ```ignore
|
||||
/// let helper = TrimLoopHelper {
|
||||
/// whitespace_chars: vec![" ".to_string(), "\t".to_string()],
|
||||
/// ..Default::default()
|
||||
/// };
|
||||
///
|
||||
/// assert!(helper.is_whitespace(" "));
|
||||
/// assert!(helper.is_whitespace("\t"));
|
||||
/// assert!(!helper.is_whitespace("a"));
|
||||
/// ```
|
||||
pub fn is_whitespace(&self, ch: &str) -> bool {
|
||||
self.whitespace_chars.iter().any(|wc| wc == ch)
|
||||
}
|
||||
|
||||
/// Check if this is a safe Trim pattern that can bypass LoopBodyLocal restrictions
|
||||
///
|
||||
/// A safe Trim pattern must:
|
||||
/// 1. Have a valid carrier name
|
||||
/// 2. Have at least one whitespace character to compare
|
||||
/// 3. Have the expected structure (substring + OR chain + break)
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// `true` if this is a safe Trim pattern, `false` otherwise
|
||||
///
|
||||
/// # Example
|
||||
///
|
||||
/// ```ignore
|
||||
/// let helper = TrimLoopHelper {
|
||||
/// original_var: "ch".to_string(),
|
||||
/// carrier_name: "is_whitespace".to_string(),
|
||||
/// whitespace_chars: vec![" ".to_string(), "\t".to_string()],
|
||||
/// };
|
||||
/// assert!(helper.is_safe_trim());
|
||||
/// ```
|
||||
pub fn is_safe_trim(&self) -> bool {
|
||||
// Basic validation
|
||||
!self.carrier_name.is_empty() && !self.whitespace_chars.is_empty()
|
||||
}
|
||||
|
||||
/// Alias for is_safe_trim() - checks if this follows the Trim-like pattern
|
||||
///
|
||||
/// This method provides a semantic alias for safety checks.
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// `true` if this pattern is Trim-like, `false` otherwise
|
||||
pub fn is_trim_like(&self) -> bool {
|
||||
self.is_safe_trim()
|
||||
}
|
||||
|
||||
/// Check if this pattern has the expected Trim structure:
|
||||
/// - substring() method call
|
||||
/// - OR chain of equality comparisons
|
||||
/// - break on non-match
|
||||
///
|
||||
/// # Returns
|
||||
///
|
||||
/// `true` if the pattern has valid structure, `false` otherwise
|
||||
///
|
||||
/// # Implementation Note
|
||||
///
|
||||
/// For now, just check basic requirements.
|
||||
/// The full structure was already validated by LoopBodyCarrierPromoter.
|
||||
pub fn has_valid_structure(&self) -> bool {
|
||||
!self.original_var.is_empty()
|
||||
&& !self.carrier_name.is_empty()
|
||||
&& !self.whitespace_chars.is_empty()
|
||||
}
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
mod tests {
|
||||
use super::*;
|
||||
|
||||
#[test]
|
||||
fn test_from_pattern_info() {
|
||||
let trim_info = TrimPatternInfo {
|
||||
var_name: "ch".to_string(),
|
||||
comparison_literals: vec![" ".to_string(), "\t".to_string()],
|
||||
carrier_name: "is_ch_match".to_string(),
|
||||
};
|
||||
|
||||
let helper = TrimLoopHelper::from_pattern_info(&trim_info);
|
||||
|
||||
assert_eq!(helper.original_var, "ch");
|
||||
assert_eq!(helper.carrier_name, "is_ch_match");
|
||||
assert_eq!(helper.whitespace_chars.len(), 2);
|
||||
assert!(helper.whitespace_chars.contains(&" ".to_string()));
|
||||
assert!(helper.whitespace_chars.contains(&"\t".to_string()));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_carrier_type() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![],
|
||||
};
|
||||
|
||||
assert_eq!(helper.carrier_type(), "Bool");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_initial_value() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![],
|
||||
};
|
||||
|
||||
assert_eq!(helper.initial_value(), true);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_whitespace_count() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![" ".to_string(), "\t".to_string(), "\n".to_string(), "\r".to_string()],
|
||||
};
|
||||
|
||||
assert_eq!(helper.whitespace_count(), 4);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_is_whitespace() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![" ".to_string(), "\t".to_string()],
|
||||
};
|
||||
|
||||
assert!(helper.is_whitespace(" "));
|
||||
assert!(helper.is_whitespace("\t"));
|
||||
assert!(!helper.is_whitespace("\n"));
|
||||
assert!(!helper.is_whitespace("a"));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_is_safe_trim() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![" ".to_string(), "\t".to_string()],
|
||||
};
|
||||
assert!(helper.is_safe_trim());
|
||||
assert!(helper.is_trim_like());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_is_safe_trim_empty_carrier() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "".to_string(), // Empty!
|
||||
whitespace_chars: vec![" ".to_string()],
|
||||
};
|
||||
assert!(!helper.is_safe_trim());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_is_safe_trim_no_whitespace() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![], // Empty!
|
||||
};
|
||||
assert!(!helper.is_safe_trim());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_has_valid_structure() {
|
||||
let helper = TrimLoopHelper {
|
||||
original_var: "ch".to_string(),
|
||||
carrier_name: "is_whitespace".to_string(),
|
||||
whitespace_chars: vec![" ".to_string()],
|
||||
};
|
||||
assert!(helper.has_valid_structure());
|
||||
}
|
||||
}
|
||||
Reference in New Issue
Block a user