feat(normalization): Phase 142 P0 - Statement-level normalization
## Summary Changed normalization unit from "block suffix" to "statement (loop only)" to prevent pattern explosion. ## Changes 1. **PlanBox** (`plan_box.rs`): - Always return `loop_only()` for any `loop(true)`, regardless of what follows - Subsequent statements (return, assignments) handled by normal MIR lowering - ~70 lines reduced, 7 unit tests updated 2. **build_block** (`stmts.rs`): - Removed `break` after consumed=1 from suffix_router - Continue processing subsequent statements normally - Phase 142 P0 comments added 3. **Tests**: - Fixture: `phase142_loop_stmt_only_then_return_length_min.hako` - VM smoke: exit code 3 (s="abc" → s.length() → 3) ## Results - ✅ Unit tests: 10/10 passed - ✅ Phase 142 VM smoke: PASS - ✅ Phase 131 regression: PASS - ✅ Build: Success ## Design - **Pattern Explosion Prevention**: Normalize only the loop (consumed=1) - **Out-of-Scope Policy**: Always Ok(None) for fallback - **Fail-Fast**: Only for "in-scope but broken" cases 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
@ -0,0 +1,24 @@
|
||||
// Phase 142 P0: Loop normalization as single statement
|
||||
//
|
||||
// Purpose: Verify loop(true) normalization returns consumed=1, allowing
|
||||
// subsequent statements (return s.length()) to be processed normally.
|
||||
//
|
||||
// Expected output: 3 (s="abc" → s.length() → 3)
|
||||
//
|
||||
// Structure:
|
||||
// s = "abc" // pre-loop init with string
|
||||
// loop(true) { // condition is Bool literal true
|
||||
// break // break at end
|
||||
// }
|
||||
// return s.length() // subsequent statement processed normally
|
||||
|
||||
static box Main {
|
||||
main() {
|
||||
local s
|
||||
s = "abc"
|
||||
loop(true) {
|
||||
break
|
||||
}
|
||||
return s.length()
|
||||
}
|
||||
}
|
||||
@ -70,76 +70,17 @@ impl NormalizationPlanBox {
|
||||
return Ok(None);
|
||||
}
|
||||
|
||||
// Phase 131: Loop-only pattern (single statement)
|
||||
if remaining.len() == 1 {
|
||||
// Phase 142 P0: Always return loop_only for loop(true), regardless of what follows
|
||||
// Normalization unit is now "statement (loop 1個)" not "block suffix"
|
||||
// Subsequent statements (return, assignments, etc.) handled by normal MIR lowering
|
||||
if debug {
|
||||
trace.routing(
|
||||
"normalization/plan",
|
||||
func_name,
|
||||
"Detected Phase 131 pattern: loop-only",
|
||||
"Detected loop(true) - Phase 142 P0: returning loop_only (consumed=1)",
|
||||
);
|
||||
}
|
||||
return Ok(Some(NormalizationPlan::loop_only()));
|
||||
}
|
||||
|
||||
// Phase 132-135: Loop + (assign*) + return
|
||||
// Count consecutive assignments after the loop
|
||||
let mut post_assign_count = 0;
|
||||
for i in 1..remaining.len() {
|
||||
if matches!(&remaining[i], ASTNode::Assignment { .. }) {
|
||||
post_assign_count += 1;
|
||||
} else {
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
// After assignments (0 or more), we need a return statement
|
||||
let return_index = 1 + post_assign_count;
|
||||
if return_index >= remaining.len() {
|
||||
// No statement after assignments - not a post pattern
|
||||
if debug {
|
||||
trace.routing(
|
||||
"normalization/plan",
|
||||
func_name,
|
||||
&format!(
|
||||
"No statement after {} assignments, not a valid post pattern",
|
||||
post_assign_count
|
||||
),
|
||||
);
|
||||
}
|
||||
// If there's a non-assignment statement after loop but no return, treat as not a pattern
|
||||
return Ok(None);
|
||||
}
|
||||
|
||||
let has_return = matches!(&remaining[return_index], ASTNode::Return { .. });
|
||||
if !has_return {
|
||||
// Statement after loop (and optional assignments) is not return - not a post pattern
|
||||
if debug {
|
||||
trace.routing(
|
||||
"normalization/plan",
|
||||
func_name,
|
||||
&format!(
|
||||
"Statement after loop and {} assignments is not return, not a valid post pattern",
|
||||
post_assign_count
|
||||
),
|
||||
);
|
||||
}
|
||||
return Ok(None);
|
||||
}
|
||||
|
||||
// Valid Phase 132-135 pattern: loop + N assignments (N >= 0) + return
|
||||
if debug {
|
||||
trace.routing(
|
||||
"normalization/plan",
|
||||
func_name,
|
||||
&format!(
|
||||
"Detected Phase 132-135 pattern: loop + {} post assigns + return",
|
||||
post_assign_count
|
||||
),
|
||||
);
|
||||
}
|
||||
|
||||
Ok(Some(NormalizationPlan::loop_with_post(post_assign_count)))
|
||||
Ok(Some(NormalizationPlan::loop_only()))
|
||||
}
|
||||
}
|
||||
|
||||
@ -226,9 +167,10 @@ mod tests {
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_phase132_loop_post_single() {
|
||||
fn test_plan_block_suffix_phase142_loop_with_subsequent_stmts() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Phase 142 P0: loop(true) returns loop_only regardless of subsequent statements
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_assignment("x", 2),
|
||||
@ -241,15 +183,16 @@ mod tests {
|
||||
|
||||
assert!(plan.is_some());
|
||||
let plan = plan.unwrap();
|
||||
assert_eq!(plan.consumed, 3); // loop + 1 assign + return
|
||||
assert_eq!(plan.kind, PlanKind::LoopWithPost { post_assign_count: 1 });
|
||||
assert!(plan.requires_return);
|
||||
assert_eq!(plan.consumed, 1); // Phase 142: only consume loop
|
||||
assert_eq!(plan.kind, PlanKind::LoopOnly);
|
||||
assert!(!plan.requires_return); // Loop itself doesn't require return
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_phase133_loop_post_multi() {
|
||||
fn test_plan_block_suffix_phase142_loop_only_always() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Phase 142 P0: loop(true) always returns loop_only, even with multiple statements after
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_assignment("x", 2),
|
||||
@ -263,32 +206,9 @@ mod tests {
|
||||
|
||||
assert!(plan.is_some());
|
||||
let plan = plan.unwrap();
|
||||
assert_eq!(plan.consumed, 4); // loop + 2 assigns + return
|
||||
assert_eq!(plan.kind, PlanKind::LoopWithPost { post_assign_count: 2 });
|
||||
assert!(plan.requires_return);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_return_boundary() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Pattern with unreachable statement after return
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_assignment("x", 2),
|
||||
make_return("x"),
|
||||
make_assignment("y", 999), // Unreachable
|
||||
];
|
||||
|
||||
let builder = MirBuilder::new();
|
||||
let plan = NormalizationPlanBox::plan_block_suffix(&builder, &remaining, "test", false)
|
||||
.expect("Should not error");
|
||||
|
||||
assert!(plan.is_some());
|
||||
let plan = plan.unwrap();
|
||||
// Should consume only up to return (not the unreachable statement)
|
||||
assert_eq!(plan.consumed, 3); // loop + 1 assign + return
|
||||
assert_eq!(plan.kind, PlanKind::LoopWithPost { post_assign_count: 1 });
|
||||
assert_eq!(plan.consumed, 1); // Phase 142: only consume loop
|
||||
assert_eq!(plan.kind, PlanKind::LoopOnly);
|
||||
assert!(!plan.requires_return);
|
||||
}
|
||||
|
||||
#[test]
|
||||
@ -321,64 +241,25 @@ mod tests {
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_no_match_no_return() {
|
||||
fn test_plan_block_suffix_phase142_loop_with_trailing_stmt() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Phase 142 P0: loop(true) returns loop_only even if no return follows
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_assignment("x", 2),
|
||||
// Missing return
|
||||
];
|
||||
|
||||
let builder = MirBuilder::new();
|
||||
let plan = NormalizationPlanBox::plan_block_suffix(&builder, &remaining, "test", false)
|
||||
.expect("Should not error");
|
||||
|
||||
// No return means not a valid post pattern
|
||||
assert!(plan.is_none());
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_phase135_loop_return_only() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Phase 135: loop + return (0 post assignments)
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_return("x"),
|
||||
// No return - but still returns loop_only
|
||||
];
|
||||
|
||||
let builder = MirBuilder::new();
|
||||
let plan = NormalizationPlanBox::plan_block_suffix(&builder, &remaining, "test", false)
|
||||
.expect("Should not error");
|
||||
|
||||
// Phase 142: loop(true) always matches, regardless of subsequent statements
|
||||
assert!(plan.is_some());
|
||||
let plan = plan.unwrap();
|
||||
assert_eq!(plan.consumed, 2); // loop + return
|
||||
assert_eq!(plan.kind, PlanKind::LoopWithPost { post_assign_count: 0 });
|
||||
assert!(plan.requires_return);
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_plan_block_suffix_return_boundary_with_trailing() {
|
||||
use crate::mir::builder::MirBuilder;
|
||||
|
||||
// Pattern with unreachable statement after return (Phase 135 variation)
|
||||
let remaining = vec![
|
||||
make_loop(),
|
||||
make_return("x"),
|
||||
make_assignment("y", 999), // Unreachable
|
||||
];
|
||||
|
||||
let builder = MirBuilder::new();
|
||||
let plan = NormalizationPlanBox::plan_block_suffix(&builder, &remaining, "test", false)
|
||||
.expect("Should not error");
|
||||
|
||||
assert!(plan.is_some());
|
||||
let plan = plan.unwrap();
|
||||
// Should consume only up to return (not the unreachable statement)
|
||||
assert_eq!(plan.consumed, 2); // loop + return
|
||||
assert_eq!(plan.kind, PlanKind::LoopWithPost { post_assign_count: 0 });
|
||||
assert_eq!(plan.consumed, 1); // Only consume loop
|
||||
assert_eq!(plan.kind, PlanKind::LoopOnly);
|
||||
}
|
||||
|
||||
#[test]
|
||||
|
||||
@ -208,13 +208,14 @@ impl super::MirBuilder {
|
||||
trace.emit_if(
|
||||
"debug",
|
||||
"build_block/suffix_router",
|
||||
&format!("Suffix router consumed {} statements (including return), stopping block build", consumed),
|
||||
&format!("Phase 142 P0: Suffix router consumed {} statement(s), continuing to process subsequent statements", consumed),
|
||||
debug,
|
||||
);
|
||||
// Suffix router consumes a return statement and emits a host-level Return
|
||||
// after the JoinIR merge. Break to avoid double-processing.
|
||||
// Phase 142 P0: Normalization unit is now "statement (loop 1個)"
|
||||
// Loop normalization returns consumed=1, and subsequent statements
|
||||
// (return, assignments, etc.) are handled by normal MIR lowering
|
||||
idx += consumed;
|
||||
break;
|
||||
// No break - continue processing subsequent statements
|
||||
}
|
||||
None => {
|
||||
// No match, proceed with normal statement build
|
||||
|
||||
@ -0,0 +1,56 @@
|
||||
#!/bin/bash
|
||||
# Phase 142 P0: Loop normalization as single statement (VM)
|
||||
#
|
||||
# Verifies that loop(true) normalization returns consumed=1, allowing
|
||||
# subsequent statements (return s.length()) to be processed normally.
|
||||
# Expected: exit code 3 (s="abc" → s.length() → 3)
|
||||
#
|
||||
# Dev-only: NYASH_JOINIR_DEV=1 HAKO_JOINIR_STRICT=1
|
||||
|
||||
source "$(dirname "$0")/../../../lib/test_runner.sh"
|
||||
export SMOKES_USE_PYVM=0
|
||||
require_env || exit 2
|
||||
|
||||
# Phase 142 is a dev-only Normalized shadow loop case
|
||||
require_joinir_dev
|
||||
|
||||
PASS_COUNT=0
|
||||
FAIL_COUNT=0
|
||||
RUN_TIMEOUT_SECS=${RUN_TIMEOUT_SECS:-10}
|
||||
|
||||
echo "[INFO] Phase 142 P0: Loop normalization as single statement (VM)"
|
||||
|
||||
# Test 1: phase142_loop_stmt_only_then_return_length_min.hako
|
||||
echo "[INFO] Test 1: phase142_loop_stmt_only_then_return_length_min.hako"
|
||||
INPUT="$NYASH_ROOT/apps/tests/phase142_loop_stmt_only_then_return_length_min.hako"
|
||||
|
||||
set +e
|
||||
OUTPUT=$(timeout "$RUN_TIMEOUT_SECS" env \
|
||||
NYASH_DISABLE_PLUGINS=1 \
|
||||
"$NYASH_BIN" --backend vm "$INPUT" 2>&1)
|
||||
EXIT_CODE=$?
|
||||
set -e
|
||||
|
||||
if [ "$EXIT_CODE" -eq 124 ]; then
|
||||
echo "[FAIL] hakorune timed out (>${RUN_TIMEOUT_SECS}s)"
|
||||
FAIL_COUNT=$((FAIL_COUNT + 1))
|
||||
elif [ "$EXIT_CODE" -eq 3 ]; then
|
||||
# Phase 142: expected output is exit code 3 (s.length() where s="abc")
|
||||
echo "[PASS] exit code verified: 3"
|
||||
PASS_COUNT=$((PASS_COUNT + 1))
|
||||
else
|
||||
echo "[FAIL] hakorune failed with exit code $EXIT_CODE (expected 3)"
|
||||
echo "[INFO] output (tail):"
|
||||
echo "$OUTPUT" | tail -n 50 || true
|
||||
FAIL_COUNT=$((FAIL_COUNT + 1))
|
||||
fi
|
||||
|
||||
echo "[INFO] PASS: $PASS_COUNT, FAIL: $FAIL_COUNT"
|
||||
|
||||
if [ "$FAIL_COUNT" -eq 0 ]; then
|
||||
test_pass "phase142_loop_stmt_only_then_return_length_min_vm: All tests passed"
|
||||
exit 0
|
||||
else
|
||||
test_fail "phase142_loop_stmt_only_then_return_length_min_vm: $FAIL_COUNT test(s) failed"
|
||||
exit 1
|
||||
fi
|
||||
Reference in New Issue
Block a user