feat(phase21.5/22.1): MirBuilder JsonFrag refactor + FileBox ring-1 + registry tests

Phase 21.5 (AOT/LLVM Optimization Prep)
- FileBox ring-1 (core-ro) provider: priority=-100, always available, no panic path
  - src/runner/modes/common_util/provider_registry.rs: CoreRoFileProviderFactory
  - Auto-registers at startup, eliminates fallback panic structurally
- StringBox fast path prototypes (length/size optimization)
- Performance benchmarks (C/Python/Hako comparison baseline)

Phase 22.1 (JsonFrag Unification)
- JsonFrag.last_index_of_from() for backward search (VM fallback)
- Replace hand-written lastIndexOf in lower_loop_sum_bc_box.hako
- SentinelExtractorBox for Break/Continue pattern extraction

MirBuilder Refactor (Box → JsonFrag Migration)
- 20+ lower_*_box.hako: Box-heavy → JsonFrag text assembly
- MirBuilderMinBox: lightweight using set for dev env
- Registry-only fast path with [registry:*] tag observation
- pattern_util_box.hako: enhanced pattern matching

Dev Environment & Testing
- Dev toggles: SMOKES_DEV_PREINCLUDE=1 (point-enable), HAKO_MIR_BUILDER_SKIP_LOOPS=1
- phase2160: registry opt-in tests (array/map get/set/push/len) - content verification
- phase2034: rc-dependent → token grep (grep -F based validation)
- run_quick.sh: fast smoke testing harness
- ENV documentation: docs/ENV_VARS.md

Test Results
 quick phase2034: ALL GREEN (MirBuilder internal patterns)
 registry phase2160: ALL GREEN (array/map get/set/push/len)
 rc-dependent tests → content token verification complete
 PREINCLUDE policy: default OFF, point-enable only where needed

Technical Notes
- No INCLUDE by default (maintain minimalism)
- FAIL_FAST=0 in Bring-up contexts only (explicit dev toggles)
- Tag-based route observation ([mirbuilder/min:*], [registry:*])
- MIR structure validation (not just rc parity)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
nyash-codex
2025-11-10 19:42:42 +09:00
parent fc5706e3f2
commit 6055d53eff
135 changed files with 3983 additions and 1150 deletions

View File

@ -13,6 +13,31 @@ impl MirInterpreter {
if let Err(e) = crate::runtime::provider_lock::guard_before_new_box(box_type) {
return Err(self.err_invalid(e));
}
// Fast path (bench/profile-only): new StringBox("const") without registry roundtrip
if std::env::var("NYASH_VM_FAST").ok().as_deref() == Some("1") && box_type == "StringBox" {
if args.len() == 1 {
let v0 = self.reg_load(args[0])?;
let s_opt: Option<String> = match v0.clone() {
VMValue::String(s) => Some(s),
VMValue::BoxRef(b) => {
if let Some(sb) = b.as_any().downcast_ref::<crate::boxes::basic::StringBox>() {
Some(sb.value.clone())
} else {
None
}
}
_ => None,
};
if let Some(s) = s_opt {
let boxed: Box<dyn crate::box_trait::NyashBox> = Box::new(crate::boxes::basic::StringBox::new(s));
let created_vm = VMValue::from_nyash_box(boxed);
self.regs.insert(dst, created_vm);
if Self::box_trace_enabled() { self.box_trace_emit_new(box_type, args.len()); }
return Ok(());
}
}
}
let converted = self.load_args_as_boxes(args)?;
let reg = crate::runtime::unified_registry::get_global_unified_registry();
let created = reg

View File

@ -13,6 +13,17 @@ pub(super) fn try_handle_string_box(
eprintln!("[vm-trace] try_handle_string_box(method={})", method);
}
let recv = this.reg_load(box_val)?;
// Ultra-fast path: raw VM string receiver for length/size (no boxing at all)
if (method == "length" || method == "size")
&& std::env::var("NYASH_VM_FAST").ok().as_deref() == Some("1")
{
if let VMValue::String(ref raw) = recv {
let use_cp = std::env::var("NYASH_STR_CP").ok().as_deref() == Some("1");
let n = if use_cp { raw.chars().count() as i64 } else { raw.len() as i64 };
this.write_result(dst, VMValue::Integer(n));
return Ok(true);
}
}
// Handle ONLY when the receiver is actually a string.
// Do NOT coerce arbitrary boxes to StringBox (e.g., ArrayBox.length()).
let sb_norm_opt: Option<crate::box_trait::StringBox> = match recv.clone() {
@ -30,6 +41,13 @@ pub(super) fn try_handle_string_box(
// Only handle known string methods here (receiver is confirmed string)
match method {
"length" | "size" => {
// Bench/profile fast path: return VMValue::Integer directly (avoid boxing overhead)
if std::env::var("NYASH_VM_FAST").ok().as_deref() == Some("1") {
let use_cp = std::env::var("NYASH_STR_CP").ok().as_deref() == Some("1");
let n = if use_cp { sb_norm.value.chars().count() as i64 } else { sb_norm.value.len() as i64 };
this.write_result(dst, VMValue::Integer(n));
return Ok(true);
}
let ret = sb_norm.length();
this.write_result(dst, VMValue::from_nyash_box(ret));
return Ok(true);