classfile: intern constant-pool entries on write (fix LVTT/LVT ClassFormatError)

ristretto's ConstantPool never deduplicates: every add_* pushes a new entry.
HotSpot matches each LocalVariableTypeTable entry to a LocalVariableTable
entry by raw name_index, so when a name (e.g. a local's) was interned twice
the two tables got different indices and the class failed to load:
'ClassFormatError: LVTT entry ... does not match any LVT entry'.

The writer now memoizes every pool builder by its semantic key and rebuilds
composite constants (Class/NameAndType/Fieldref/Methodref/MethodHandle/
Dynamic) from our own utf8/class primitives so ristretto never mints a
shadow Utf8. This also shrinks the pool (no more ldc -> ldc_w widening that
changed code length on a pure round-trip).

Split pool.rs into a pool/ module (refs in pool/refs.rs) to respect the
file-size gate. Regression test: tests/pool asserts the written pool has no
duplicate CONSTANT_Utf8. Verified: pure parse->write of the neoforge jar
(92 classes) now loads under java -Xverify:all with 0 errors.
This commit is contained in:
Loki 2026-10-08 22:55:49 +02:00
parent a2e4c69210
commit f08bf60635
10 changed files with 509 additions and 206 deletions

View file

@ -0,0 +1,47 @@
//! Инвариант писателя: записанный пул не содержит дублей `CONSTANT_Utf8`.
//!
//! Backend не дедуплицирует (`add_*` всегда толкает новый элемент), а HotSpot
//! сопоставляет запись `LocalVariableTypeTable` с `LocalVariableTable` по
//! сырому `name_index`. Дублирующий `Utf8` для одного имени даёт разные индексы
//! и рушит класс: `ClassFormatError: LVTT entry ... does not match any LVT
//! entry`. Писатель обязан интернировать строки — поэтому в записанном пуле
//! каждый `CONSTANT_Utf8` уникален.
#![allow(clippy::expect_used, clippy::unwrap_used)]
use std::collections::HashSet;
use std::path::PathBuf;
use noxis_classfile::format;
use ristretto_classfile::{ClassFile as RtClass, Constant};
#[test]
fn written_pool_has_no_duplicate_utf8() {
let dir = PathBuf::from(env!("CARGO_MANIFEST_DIR")).join("tests/data");
let mut names: Vec<PathBuf> = std::fs::read_dir(&dir)
.expect("tests/data")
.filter_map(|entry| entry.ok().map(|entry| entry.path()))
.filter(|path| path.extension().is_some_and(|ext| ext == "class"))
.collect();
names.sort();
assert!(!names.is_empty(), "нет фикстур в tests/data");
for path in &names {
let parsed = format::parse(&std::fs::read(path).expect("read")).expect("parse");
let written = format::write(&parsed).expect("write");
let raw = RtClass::from_bytes(&written).expect("written parses");
raw.verify().expect("written verifies");
let mut seen = HashSet::new();
for index in 1..=raw.constant_pool.len() {
if let Some(Constant::Utf8(value)) = raw.constant_pool.get(index as u16) {
let bytes = value.as_bytes().to_vec();
assert!(
seen.insert(bytes),
"дубль CONSTANT_Utf8 {:?} в {}",
String::from_utf8_lossy(value.as_bytes()),
parsed.this_class
);
}
}
}
}