mirror of
https://github.com/hpd840321/starRiverProperty.git
synced 2026-06-09 16:30:29 +08:00
fix: relocate cwos-portal decompiled output to correct path; remove nested directory
Former-commit-id: dc30d42a8c55ed8b2382a41dc2434233fbed9930
This commit is contained in:
反编译工作区
+196
@@ -0,0 +1,196 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
生成“贴合源码”的反编译代码树：
|
||||
|
||||
- 以 V1 CFR 反编译目录为骨架；
|
||||
- 若同路径在 V2 Maven 源码存在，则用 V2 源码覆盖该文件；
|
||||
- 若 V2 不存在，则保留 V1 CFR 文件；
|
||||
- 同时补充 V2 独有文件（V1 中无）。
|
||||
|
||||
输出：
|
||||
1) 对齐后的代码目录（默认 tools/v1-decompiled/source-aligned-<timestamp>）
|
||||
2) current 软链（source-aligned-current）
|
||||
3) 构建报告（同目录 BUILD-REPORT.md / BUILD-REPORT.json）
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import json
|
||||
import os
|
||||
import shutil
|
||||
from datetime import datetime, timezone
|
||||
from pathlib import Path
|
||||
|
||||
SCR = Path(__file__).resolve()
|
||||
MV_ROOT = Path(os.environ.get("MV_ROOT", SCR.parents[3])).resolve()
|
||||
DEFAULT_V1 = (MV_ROOT / "tools" / "v1-decompiled" / "cfr-from-cw-lib-current").resolve()
|
||||
|
||||
PAIR = [
|
||||
("cw-elevator-application-common", "cw-elevator-application-common-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-data", "cw-elevator-application-data-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-service", "cw-elevator-application-service-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-web", "cw-elevator-application-web-1.0-SNAPSHOT"),
|
||||
]
|
||||
|
||||
|
||||
def copy_file(src: Path, dst: Path) -> None:
|
||||
dst.parent.mkdir(parents=True, exist_ok=True)
|
||||
shutil.copy2(src, dst)
|
||||
|
||||
|
||||
def rel_java_map(root: Path) -> dict[str, Path]:
|
||||
return {str(p.relative_to(root)).replace("\\", "/"): p for p in root.rglob("*.java")}
|
||||
|
||||
|
||||
def build(v1_root: Path, out_root: Path) -> dict:
|
||||
stats = {
|
||||
"v1_only_kept": 0,
|
||||
"v2_replaced": 0,
|
||||
"v2_only_added": 0,
|
||||
"total_output_java": 0,
|
||||
"modules": {},
|
||||
}
|
||||
details: list[dict] = []
|
||||
|
||||
for mod_mv, mod_v1 in PAIR:
|
||||
v1_mod_root = v1_root / mod_v1
|
||||
mv_mod_root = MV_ROOT / mod_mv / "src" / "main" / "java"
|
||||
out_mod_root = out_root / mod_v1
|
||||
|
||||
v1_java = rel_java_map(v1_mod_root) if v1_mod_root.is_dir() else {}
|
||||
mv_java = rel_java_map(mv_mod_root) if mv_mod_root.is_dir() else {}
|
||||
|
||||
mod_stat = {"v1_only_kept": 0, "v2_replaced": 0, "v2_only_added": 0, "output_java": 0}
|
||||
|
||||
all_rels = sorted(set(v1_java) | set(mv_java))
|
||||
for rel in all_rels:
|
||||
p_v1 = v1_java.get(rel)
|
||||
p_v2 = mv_java.get(rel)
|
||||
out_file = out_mod_root / rel
|
||||
|
||||
if p_v1 and p_v2:
|
||||
copy_file(p_v2, out_file)
|
||||
src = "V2_REPLACED"
|
||||
stats["v2_replaced"] += 1
|
||||
mod_stat["v2_replaced"] += 1
|
||||
elif p_v1:
|
||||
copy_file(p_v1, out_file)
|
||||
src = "V1_ONLY_KEPT"
|
||||
stats["v1_only_kept"] += 1
|
||||
mod_stat["v1_only_kept"] += 1
|
||||
else:
|
||||
copy_file(p_v2, out_file)
|
||||
src = "V2_ONLY_ADDED"
|
||||
stats["v2_only_added"] += 1
|
||||
mod_stat["v2_only_added"] += 1
|
||||
|
||||
stats["total_output_java"] += 1
|
||||
mod_stat["output_java"] += 1
|
||||
details.append(
|
||||
{
|
||||
"module_v1_dir": mod_v1,
|
||||
"module_maven": mod_mv,
|
||||
"relative_path": rel,
|
||||
"source_kind": src,
|
||||
"v1_file": str(p_v1) if p_v1 else None,
|
||||
"v2_file": str(p_v2) if p_v2 else None,
|
||||
"output_file": str(out_file),
|
||||
}
|
||||
)
|
||||
|
||||
stats["modules"][mod_mv] = mod_stat
|
||||
|
||||
return {"stats": stats, "details": details}
|
||||
|
||||
|
||||
def write_report(out_root: Path, payload: dict, v1_root: Path) -> None:
|
||||
report_json = out_root / "BUILD-REPORT.json"
|
||||
report_md = out_root / "BUILD-REPORT.md"
|
||||
report_json.write_text(json.dumps(payload, ensure_ascii=False, indent=2), encoding="utf-8")
|
||||
|
||||
s = payload["stats"]
|
||||
lines = [
|
||||
"# Source-Aligned Decompiled Build Report",
|
||||
"",
|
||||
f"- UTC: `{datetime.now(timezone.utc).strftime('%Y-%m-%dT%H:%M:%SZ')}`",
|
||||
f"- V1 base: `{v1_root}`",
|
||||
f"- Maven root: `{MV_ROOT}`",
|
||||
"",
|
||||
"## Summary",
|
||||
"",
|
||||
"| metric | count |",
|
||||
"|---|---:|",
|
||||
f"| v2_replaced | {s['v2_replaced']} |",
|
||||
f"| v1_only_kept | {s['v1_only_kept']} |",
|
||||
f"| v2_only_added | {s['v2_only_added']} |",
|
||||
f"| total_output_java | {s['total_output_java']} |",
|
||||
"",
|
||||
"## Module stats",
|
||||
"",
|
||||
"| module | v2_replaced | v1_only_kept | v2_only_added | output_java |",
|
||||
"|---|---:|---:|---:|---:|",
|
||||
]
|
||||
for m, ms in s["modules"].items():
|
||||
lines.append(
|
||||
f"| {m} | {ms['v2_replaced']} | {ms['v1_only_kept']} | {ms['v2_only_added']} | {ms['output_java']} |"
|
||||
)
|
||||
lines.append("")
|
||||
lines.append("## Notes")
|
||||
lines.append("")
|
||||
lines.append("- 该目录可视作“源码优先的反编译校正版”：同路径优先采用 Maven 源码。")
|
||||
lines.append("- 若你要严格保留历史 jar 语义，请结合 `details` 过滤 `V2_REPLACED`/`V2_ONLY_ADDED`。")
|
||||
report_md.write_text("\n".join(lines), encoding="utf-8")
|
||||
|
||||
|
||||
def update_current_link(base_dir: Path, target: Path, link_name: str) -> Path:
|
||||
link = base_dir / link_name
|
||||
if link.exists() or link.is_symlink():
|
||||
link.unlink()
|
||||
link.symlink_to(target.name)
|
||||
return link
|
||||
|
||||
|
||||
def main() -> int:
|
||||
ap = argparse.ArgumentParser()
|
||||
ap.add_argument("--v1-root", default=str(DEFAULT_V1), help="V1 CFR 根目录")
|
||||
ap.add_argument("--out-dir", default="", help="输出目录；默认自动生成时间戳目录")
|
||||
ap.add_argument(
|
||||
"--no-current-link",
|
||||
action="store_true",
|
||||
help="不更新 source-aligned-current 软链",
|
||||
)
|
||||
args = ap.parse_args()
|
||||
|
||||
v1_root = Path(args.v1_root).resolve()
|
||||
if not v1_root.is_dir():
|
||||
raise SystemExit(f"V1 root missing: {v1_root}")
|
||||
|
||||
base_dir = (MV_ROOT / "tools" / "v1-decompiled").resolve()
|
||||
base_dir.mkdir(parents=True, exist_ok=True)
|
||||
if args.out_dir:
|
||||
out_root = Path(args.out_dir).resolve()
|
||||
else:
|
||||
ts = datetime.now().strftime("%Y%m%d-%H%M%S")
|
||||
out_root = base_dir / f"source-aligned-{ts}"
|
||||
out_root.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
payload = build(v1_root=v1_root, out_root=out_root)
|
||||
write_report(out_root=out_root, payload=payload, v1_root=v1_root)
|
||||
|
||||
if not args.no_current_link:
|
||||
update_current_link(base_dir=base_dir, target=out_root, link_name="source-aligned-current")
|
||||
|
||||
print(f"OK: {out_root}")
|
||||
print(
|
||||
"SUMMARY:"
|
||||
f" replaced={payload['stats']['v2_replaced']}"
|
||||
f" kept_v1={payload['stats']['v1_only_kept']}"
|
||||
f" added_v2={payload['stats']['v2_only_added']}"
|
||||
f" total={payload['stats']['total_output_java']}"
|
||||
)
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
+118
@@ -0,0 +1,118 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
javap -c 文本在「行级归一」之后的二次归一：折叠布局等价序列、依赖升级导致的等价 API 形态。
|
||||
供 bytecode_parity_rootcause / export_impl_diff_evidence 共用。
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import re
|
||||
|
||||
|
||||
def _collapse_double_goto(result: str) -> str:
|
||||
while True:
|
||||
n2 = re.sub(r"(?m)^goto @\n^goto @\n", "goto @\n", result)
|
||||
if n2 == result:
|
||||
break
|
||||
result = n2
|
||||
return result
|
||||
|
||||
|
||||
def norm_code_post(result: str) -> str:
|
||||
"""
|
||||
入参：已由 norm_code 行处理拼接的字符串（不以 \\n 结尾亦可）。
|
||||
"""
|
||||
# --- 1) 条件 + 紧跟 goto @（蹦床）：折叠为反条件单指令，对齐 javac 两种发射形态 ---
|
||||
branch_pairs = [
|
||||
(r"(?m)^ifnonnull @\n^goto @\n", "ifnull @\n"),
|
||||
(r"(?m)^if_icmpne @\n^goto @\n", "if_icmpeq @\n"),
|
||||
(r"(?m)^ifne @\n^goto @\n", "ifeq @\n"),
|
||||
(r"(?m)^if_icmpeq @\n^goto @\n", "if_icmpne @\n"),
|
||||
]
|
||||
for pat, repl in branch_pairs:
|
||||
result = re.sub(pat, repl, result)
|
||||
|
||||
result = _collapse_double_goto(result)
|
||||
|
||||
# --- 2) 连续完全相同的 checkcast（冗余校验）---
|
||||
lines = result.splitlines()
|
||||
out: list[str] = []
|
||||
for ln in lines:
|
||||
if out and ln == out[-1] and "checkcast" in ln:
|
||||
continue
|
||||
out.append(ln)
|
||||
result = "\n".join(out)
|
||||
|
||||
# --- 3) String.join 前对 null 入参的 checkcast（仅一版 javac 发射）---
|
||||
result = re.sub(
|
||||
r"(?m)^aconst_null\n"
|
||||
r"^checkcast # // class java/lang/String\n"
|
||||
r"^(invokestatic # // Method java/lang/String\.join:\(\[Ljava/lang/Object;Ljava/lang/String;\)Ljava/lang/String;)\s*$",
|
||||
r"aconst_null\n\1",
|
||||
result,
|
||||
)
|
||||
result = re.sub(
|
||||
r"(?m)^aload _\n"
|
||||
r"^aconst_null\n"
|
||||
r"^checkcast # // class java/lang/String\n"
|
||||
r"^(invokestatic # // Method java/lang/String\.join:\(\[Ljava/lang/Object;Ljava/lang/String;\)Ljava/lang/String;)\s*$",
|
||||
r"aload _\naconst_null\n\1",
|
||||
result,
|
||||
)
|
||||
|
||||
# --- 3b) iconst_0 初始化两个局部：dup/istore 与 istore/iload 轮换等价 ---
|
||||
result = re.sub(
|
||||
r"(?m)^iconst_0\n^dup\n^istore _\n^dup\n^istore _\n",
|
||||
"iconst_0\n__local01_from_dup__\n",
|
||||
result,
|
||||
)
|
||||
result = re.sub(
|
||||
r"(?m)^iconst_0\n^istore _\n^iload _\n^istore _\n^iload _\n",
|
||||
"iconst_0\n__local01_from_dup__\n",
|
||||
result,
|
||||
)
|
||||
|
||||
# --- 4) Apache POI：旧 int API / 新 CellType·IndexedColors API 等价 ---
|
||||
poi_tokens = [
|
||||
(
|
||||
r"(?m)^iconst_1\n^invokeinterface # // InterfaceMethod org/apache/poi/ss/usermodel/Cell\.setCellType:\(I\)V\s*$",
|
||||
"__poi_Cell_setCellType_STRING__",
|
||||
),
|
||||
(
|
||||
r"(?m)^getstatic # // Field org/apache/poi/ss/usermodel/CellType\.STRING:.+\n"
|
||||
r"^invokeinterface # // InterfaceMethod org/apache/poi/ss/usermodel/Cell\.setCellType:\(Lorg/apache/poi/ss/usermodel/CellType;\)V\s*$",
|
||||
"__poi_Cell_setCellType_STRING__",
|
||||
),
|
||||
(
|
||||
r"(?m)^iconst_1\n^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFCell\.setCellType:\(I\)V\s*$",
|
||||
"__poi_HSSFCell_setCellType_STRING__",
|
||||
),
|
||||
(
|
||||
r"(?m)^getstatic # // Field org/apache/poi/ss/usermodel/CellType\.STRING:.+\n"
|
||||
r"^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFCell\.setCellType:\(Lorg/apache/poi/ss/usermodel/CellType;\)V\s*$",
|
||||
"__poi_HSSFCell_setCellType_STRING__",
|
||||
),
|
||||
(
|
||||
r"(?m)^sipush 700\n^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFFont\.setBoldweight:\(S\)V\s*$",
|
||||
"__poi_HSSFFont_bold__",
|
||||
),
|
||||
(
|
||||
r"(?m)^iconst_1\n^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFFont\.setBold:\(Z\)V\s*$",
|
||||
"__poi_HSSFFont_bold__",
|
||||
),
|
||||
(
|
||||
r"(?m)^sipush 32767\n^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFFont\.setColor:\(S\)V\s*$",
|
||||
"__poi_HSSFFont_setColor_RED__",
|
||||
),
|
||||
(
|
||||
r"(?m)^getstatic # // Field org/apache/poi/ss/usermodel/IndexedColors\.RED:.+\n"
|
||||
r"^invokevirtual # // Method org/apache/poi/ss/usermodel/IndexedColors\.getIndex:\(\)S\n"
|
||||
r"^invokevirtual # // Method org/apache/poi/hssf/usermodel/HSSFFont\.setColor:\(S\)V\s*$",
|
||||
"__poi_HSSFFont_setColor_RED__",
|
||||
),
|
||||
]
|
||||
for pat, tok in poi_tokens:
|
||||
result = re.sub(pat, tok, result)
|
||||
|
||||
result = _collapse_double_goto(result)
|
||||
return result
|
||||
+331
@@ -0,0 +1,331 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
V1 原始 jar vs V2 编译 jar 的字节码多维对比（根因归类）
|
||||
|
||||
维度：
|
||||
1) 类存在性：ONLY_V1 / ONLY_V2
|
||||
2) API 签名：javap -p -s 归一化后 hash
|
||||
3) 方法实现：javap -c -p 归一化后 hash
|
||||
|
||||
说明：
|
||||
- 仅比较 cn/cloudwalk/elevator/** 下 class
|
||||
- 忽略常量池索引号、栈帧偏移等不稳定文本
|
||||
- 行级归一见 `norm_code`，二次归一见同目录 `bytecode_norm_post.norm_code_post`
|
||||
- 输出 Markdown + JSON，便于复盘与持续回归
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import hashlib
|
||||
import json
|
||||
import re
|
||||
import subprocess
|
||||
from collections import Counter, defaultdict
|
||||
from dataclasses import dataclass
|
||||
from datetime import datetime, timezone
|
||||
from pathlib import Path
|
||||
|
||||
from bytecode_norm_post import norm_code_post
|
||||
|
||||
REPO_ROOT = Path(__file__).resolve().parents[3]
|
||||
V1_ROOT = REPO_ROOT.parent / "cw-elevator-application-V1.0.0.20211103" / "cw_lib"
|
||||
|
||||
PAIR = [
|
||||
("cw-elevator-application-common", "cw-elevator-application-common"),
|
||||
("cw-elevator-application-data", "cw-elevator-application-data"),
|
||||
("cw-elevator-application-service", "cw-elevator-application-service"),
|
||||
("cw-elevator-application-web", "cw-elevator-application-web"),
|
||||
]
|
||||
|
||||
|
||||
@dataclass
|
||||
class ClassCompare:
|
||||
module: str
|
||||
class_name: str
|
||||
status: str
|
||||
api_equal: bool | None = None
|
||||
code_equal: bool | None = None
|
||||
api_hash_v1: str | None = None
|
||||
api_hash_v2: str | None = None
|
||||
code_hash_v1: str | None = None
|
||||
code_hash_v2: str | None = None
|
||||
|
||||
|
||||
def shasum(text: str) -> str:
|
||||
return hashlib.sha256(text.encode("utf-8")).hexdigest()
|
||||
|
||||
|
||||
def run(cmd: list[str]) -> str:
|
||||
return subprocess.check_output(cmd, text=True, stderr=subprocess.DEVNULL)
|
||||
|
||||
|
||||
def jar_classes(jar_file: Path) -> set[str]:
|
||||
lines = run(["jar", "tf", str(jar_file)]).splitlines()
|
||||
out: set[str] = set()
|
||||
for ln in lines:
|
||||
if not ln.endswith(".class"):
|
||||
continue
|
||||
if "$" in ln:
|
||||
# 内部类单独比较噪声较大，先跳过；主类稳定后可再扩展
|
||||
continue
|
||||
if not ln.startswith("cn/cloudwalk/elevator/"):
|
||||
continue
|
||||
out.add(ln[:-6].replace("/", "."))
|
||||
return out
|
||||
|
||||
|
||||
def norm_api(text: str) -> str:
|
||||
"""
|
||||
取 javap -p -s 中成员声明 + descriptor，将「成员行 + 下一行 descriptor:」合并为一条后排序，
|
||||
避免 Lombok/编译器仅改变类内声明顺序时误判为 API 变更。
|
||||
"""
|
||||
kept = []
|
||||
for ln in text.splitlines():
|
||||
s = ln.strip()
|
||||
if not s:
|
||||
continue
|
||||
if s.startswith("Compiled from"):
|
||||
continue
|
||||
if s.startswith("Classfile "):
|
||||
continue
|
||||
if s.startswith("flags:"):
|
||||
continue
|
||||
if s.startswith("descriptor:") or s.endswith(";") or s.startswith("public ") or s.startswith("protected ") or s.startswith("private "):
|
||||
kept.append(s)
|
||||
merged: list[str] = []
|
||||
i = 0
|
||||
while i < len(kept):
|
||||
if i + 1 < len(kept) and kept[i + 1].startswith("descriptor:"):
|
||||
merged.append(kept[i] + "\t" + kept[i + 1])
|
||||
i += 2
|
||||
else:
|
||||
merged.append(kept[i])
|
||||
i += 1
|
||||
merged.sort()
|
||||
return "\n".join(merged)
|
||||
|
||||
|
||||
def norm_code(text: str) -> str:
|
||||
out = []
|
||||
skip_exc = False
|
||||
for ln in text.splitlines():
|
||||
s = ln.rstrip()
|
||||
if not s:
|
||||
continue
|
||||
if s.startswith("Compiled from"):
|
||||
continue
|
||||
if s.startswith("Classfile "):
|
||||
continue
|
||||
# Exception table 仅偏移不同，与语义无关；整块跳过以免误判 IMPL_DIFF
|
||||
if "Exception table:" in s:
|
||||
skip_exc = True
|
||||
continue
|
||||
if skip_exc:
|
||||
if re.match(r"^\s{2}public\s+", s) or re.match(r"^\s{2}private\s+", s) or re.match(r"^\s{2}protected\s+", s):
|
||||
skip_exc = False
|
||||
else:
|
||||
continue
|
||||
# 去掉偏移量与常量池索引，保留操作码与调用目标
|
||||
s = re.sub(r"^\s*\d+:\s*", "", s)
|
||||
s = re.sub(r"#\d+", "#", s)
|
||||
# javap 的常量池注释对齐空格数量不稳定，压缩为空格规范形态
|
||||
s = re.sub(r"#\s+//", "# //", s)
|
||||
# 屏蔽 invokeinterface 参数位计数噪声
|
||||
s = re.sub(r",\s+\d+\s+//", " //", s)
|
||||
# 跳转目标偏移对语义无影响，归一化
|
||||
s = re.sub(
|
||||
r"\b(if(?:eq|ne|lt|gt|le|ge|null|nonnull)|if_icmp(?:eq|ne|lt|gt|le|ge)|if_acmp(?:eq|ne)|goto|jsr)\s+\d+\b",
|
||||
r"\1 @",
|
||||
s,
|
||||
)
|
||||
# 局部变量槽位号对语义无影响（编译器可重排）
|
||||
s = re.sub(
|
||||
r"\b([adfil]load|[adfil]store|aload|astore|iload|istore|lload|lstore|fload|fstore|dload|dstore)\s+\d+\b",
|
||||
r"\1 _",
|
||||
s,
|
||||
)
|
||||
# 归一化单字节槽位指令（如 aload_0 / istore_3）
|
||||
s = re.sub(r"\b([adfil]load|[adfil]store|aload|astore|iload|istore|lload|lstore|fload|fstore|dload|dstore)_[0-3]\b", r"\1 _", s)
|
||||
# javap 输出为 iinc slot,const（含逗号）；旧正则漏匹配会导致槽位噪声残留
|
||||
s = re.sub(r"\biinc\s+\d+\s*,\s*(-?\d+)\b", r"iinc _ \1", s)
|
||||
# ArrayList 通过接口类型调用与具体类调用语义一致，归一为 List 接口描述
|
||||
s = re.sub(
|
||||
r"invokevirtual\s+#\s+//\s*Method java/util/ArrayList\.(add|size)\b[^\n]*",
|
||||
r"invokeinterface # // InterfaceMethod java/util/List.\1:(UNKNOWN)",
|
||||
s,
|
||||
)
|
||||
s = re.sub(
|
||||
r"invokeinterface\s+#\s+//\s*InterfaceMethod java/util/List\.(add|size)\b[^\n]*",
|
||||
r"invokeinterface # // InterfaceMethod java/util/List.\1:(UNKNOWN)",
|
||||
s,
|
||||
)
|
||||
# tableswitch/lookupswitch 去掉行首偏移后，残余纯数字行与 default:<off> 仅反映布局，语义无差别
|
||||
st = s.strip()
|
||||
if re.fullmatch(r"\d+", st):
|
||||
s = "<switch_target>"
|
||||
elif re.fullmatch(r"default:\s*\d+", st):
|
||||
s = "default: <switch_target>"
|
||||
out.append(s.strip())
|
||||
return norm_code_post("\n".join(out))
|
||||
|
||||
|
||||
def javap_api(jar_file: Path, class_name: str) -> str:
|
||||
return run(["javap", "-classpath", str(jar_file), "-p", "-s", class_name])
|
||||
|
||||
|
||||
def javap_code(jar_file: Path, class_name: str) -> str:
|
||||
return run(["javap", "-classpath", str(jar_file), "-p", "-c", class_name])
|
||||
|
||||
|
||||
def compare_module(module: str, v1_jar: Path, v2_jar: Path) -> list[ClassCompare]:
|
||||
v1_classes = jar_classes(v1_jar)
|
||||
v2_classes = jar_classes(v2_jar)
|
||||
records: list[ClassCompare] = []
|
||||
for cls in sorted(v1_classes | v2_classes):
|
||||
if cls not in v2_classes:
|
||||
records.append(ClassCompare(module=module, class_name=cls, status="ONLY_V1"))
|
||||
continue
|
||||
if cls not in v1_classes:
|
||||
records.append(ClassCompare(module=module, class_name=cls, status="ONLY_V2"))
|
||||
continue
|
||||
|
||||
api1 = norm_api(javap_api(v1_jar, cls))
|
||||
api2 = norm_api(javap_api(v2_jar, cls))
|
||||
code1 = norm_code(javap_code(v1_jar, cls))
|
||||
code2 = norm_code(javap_code(v2_jar, cls))
|
||||
api_h1, api_h2 = shasum(api1), shasum(api2)
|
||||
code_h1, code_h2 = shasum(code1), shasum(code2)
|
||||
api_eq = api_h1 == api_h2
|
||||
code_eq = code_h1 == code_h2
|
||||
if api_eq and code_eq:
|
||||
st = "MATCH"
|
||||
elif api_eq and not code_eq:
|
||||
st = "IMPL_DIFF"
|
||||
else:
|
||||
st = "API_DIFF"
|
||||
records.append(
|
||||
ClassCompare(
|
||||
module=module,
|
||||
class_name=cls,
|
||||
status=st,
|
||||
api_equal=api_eq,
|
||||
code_equal=code_eq,
|
||||
api_hash_v1=api_h1,
|
||||
api_hash_v2=api_h2,
|
||||
code_hash_v1=code_h1,
|
||||
code_hash_v2=code_h2,
|
||||
)
|
||||
)
|
||||
return records
|
||||
|
||||
|
||||
def main() -> int:
|
||||
ap = argparse.ArgumentParser()
|
||||
ap.add_argument(
|
||||
"--out-md",
|
||||
default=str(REPO_ROOT / "tools/elevator_api_parity/report/V1-V2-BYTECODE-ROOTCAUSE.md"),
|
||||
)
|
||||
ap.add_argument(
|
||||
"--out-json",
|
||||
default=str(REPO_ROOT / "tools/elevator_api_parity/report/V1-V2-BYTECODE-ROOTCAUSE.json"),
|
||||
)
|
||||
args = ap.parse_args()
|
||||
|
||||
all_records: list[ClassCompare] = []
|
||||
for mod_short, mod_art in PAIR:
|
||||
v1_jar = V1_ROOT / f"{mod_art}-1.0-SNAPSHOT.jar"
|
||||
v2_jar = REPO_ROOT / mod_short / "target" / f"{mod_art}-2.0-SNAPSHOT.jar"
|
||||
if not v1_jar.is_file() or not v2_jar.is_file():
|
||||
raise SystemExit(f"missing jar for module {mod_short}: v1={v1_jar} v2={v2_jar}")
|
||||
all_records.extend(compare_module(mod_short, v1_jar=v1_jar, v2_jar=v2_jar))
|
||||
|
||||
counter = Counter(r.status for r in all_records)
|
||||
mod_counter: dict[str, Counter] = defaultdict(Counter)
|
||||
for r in all_records:
|
||||
mod_counter[r.module][r.status] += 1
|
||||
|
||||
payload = {
|
||||
"generated_at_utc": datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ"),
|
||||
"summary": dict(counter),
|
||||
"module_summary": {k: dict(v) for k, v in mod_counter.items()},
|
||||
"records": [r.__dict__ for r in all_records],
|
||||
}
|
||||
out_json = Path(args.out_json)
|
||||
out_json.parent.mkdir(parents=True, exist_ok=True)
|
||||
out_json.write_text(json.dumps(payload, ensure_ascii=False, indent=2), encoding="utf-8")
|
||||
|
||||
md = []
|
||||
md.append("# V1 vs V2 字节码根因分析")
|
||||
md.append("")
|
||||
md.append(f"- 生成时间（UTC）：`{payload['generated_at_utc']}`")
|
||||
md.append("- 对比范围：四模块 `cn/cloudwalk/elevator/**` 主类（不含 `$` 内部类）")
|
||||
md.append("- 维度：类存在性 / API签名 / 方法实现")
|
||||
md.append("")
|
||||
md.append("## 1. 总览")
|
||||
md.append("")
|
||||
md.append("| 维度 | 数量 |")
|
||||
md.append("|---|---:|")
|
||||
for k in ["MATCH", "IMPL_DIFF", "API_DIFF", "ONLY_V1", "ONLY_V2"]:
|
||||
md.append(f"| {k} | {counter.get(k, 0)} |")
|
||||
md.append("")
|
||||
md.append("## 2. 各模块分布")
|
||||
md.append("")
|
||||
md.append("| 模块 | MATCH | IMPL_DIFF | API_DIFF | ONLY_V1 | ONLY_V2 |")
|
||||
md.append("|---|---:|---:|---:|---:|---:|")
|
||||
for m in [p[0] for p in PAIR]:
|
||||
c = mod_counter[m]
|
||||
md.append(
|
||||
f"| {m} | {c.get('MATCH',0)} | {c.get('IMPL_DIFF',0)} | {c.get('API_DIFF',0)} | {c.get('ONLY_V1',0)} | {c.get('ONLY_V2',0)} |"
|
||||
)
|
||||
md.append("")
|
||||
|
||||
def section(title: str, status: str, limit: int = 120) -> None:
|
||||
md.append(f"## {title}")
|
||||
md.append("")
|
||||
rs = [r for r in all_records if r.status == status]
|
||||
if not rs:
|
||||
md.append("- 无")
|
||||
md.append("")
|
||||
return
|
||||
md.append("| 模块 | 类 |")
|
||||
md.append("|---|---|")
|
||||
for r in rs[:limit]:
|
||||
md.append(f"| {r.module} | `{r.class_name}` |")
|
||||
if len(rs) > limit:
|
||||
md.append(f"| ... | 其余 {len(rs)-limit} 个见 JSON |")
|
||||
md.append("")
|
||||
|
||||
section("3. API 相同但实现不同（优先查 JD-GUI 还原误差/人工改写）", "IMPL_DIFF")
|
||||
section("4. API 不同（优先查源码结构漂移）", "API_DIFF")
|
||||
section("5. 仅 V2 存在（新增类）", "ONLY_V2")
|
||||
section("6. 仅 V1 存在（缺失类）", "ONLY_V1")
|
||||
|
||||
md.append("## 7. 根因判读指引")
|
||||
md.append("")
|
||||
md.append("- `IMPL_DIFF`：若对应源码与 CFR 文本高度相似但字节码仍不同，多为 JD-GUI 语句重排/条件翻译偏差或后续手工改写。")
|
||||
md.append("- `API_DIFF`：优先判定为“源码结构变化”，例如方法签名、字段类型、泛型边界变化。")
|
||||
md.append("- `ONLY_V2/ONLY_V1`：属于文件集合漂移（新增/漏还原/清理）。")
|
||||
md.append("")
|
||||
md.append("## 8. API_DIFF 判读")
|
||||
md.append("")
|
||||
md.append("- **成员顺序**：同签名不同顺序（常见于 Lombok getter/setter 顺序）已由 `norm_api` 合并 descriptor 后排序消除伪差异。")
|
||||
md.append("- **默认可视构造**：仅含 static 的类若未写显式私有空参构造，javac 可能生成 **public** 默认构造，与旧 JAR 中 **private** 不同；可在源码中补充与 V1 一致。")
|
||||
md.append("- **真实契约差异**：仍报 API_DIFF 时多为 V2 **新增/改签名** 的方法或字段，需对照业务是否保留（并非脚本噪声）。")
|
||||
md.append("")
|
||||
md.append("## 9. 实现归一化（摘要）")
|
||||
md.append("")
|
||||
md.append("- **行级**：指令偏移、`#` 常量池下标、`if_icmp*`/`goto` 目标、`xload`/`xstore` 槽位、`iinc`、"
|
||||
"`ArrayList`→`List` 的 add/size、`tableswitch` 纯数字行等。")
|
||||
md.append("- **二次（`bytecode_norm_post`）**：`ifxx`+`goto` 蹦床→反条件单条；连续 `goto @` 折叠；"
|
||||
"连续重复 `checkcast` 去重；`String.join` 前冗余 `checkcast java/lang/String`；"
|
||||
"`iconst_0`+dup 与 istore/iload 轮换初始化局部；Apache POI `setCellType`/`setBold`/`setColor` 新旧 API 占位符。")
|
||||
md.append("")
|
||||
Path(args.out_md).write_text("\n".join(md), encoding="utf-8")
|
||||
print(f"WROTE: {args.out_md}")
|
||||
print(f"WROTE: {args.out_json}")
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
+83
@@ -0,0 +1,83 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
from __future__ import annotations
|
||||
|
||||
import difflib
|
||||
import json
|
||||
import subprocess
|
||||
from datetime import datetime
|
||||
from pathlib import Path
|
||||
|
||||
from bytecode_parity_rootcause import norm_code
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[3]
|
||||
REPORT = ROOT / "tools/elevator_api_parity/report/V1-V2-BYTECODE-ROOTCAUSE.json"
|
||||
OUT = ROOT / "tools/elevator_api_parity/report/impl-diff-evidence-20260428"
|
||||
V1 = ROOT.parent / "cw-elevator-application-V1.0.0.20211103/cw_lib"
|
||||
|
||||
PAIR = {
|
||||
"cw-elevator-application-common": "cw-elevator-application-common",
|
||||
"cw-elevator-application-data": "cw-elevator-application-data",
|
||||
"cw-elevator-application-service": "cw-elevator-application-service",
|
||||
"cw-elevator-application-web": "cw-elevator-application-web",
|
||||
}
|
||||
|
||||
|
||||
def run(cmd: list[str]) -> str:
|
||||
return subprocess.check_output(cmd, text=True, stderr=subprocess.DEVNULL)
|
||||
|
||||
|
||||
def main() -> int:
|
||||
OUT.mkdir(parents=True, exist_ok=True)
|
||||
data = json.loads(REPORT.read_text(encoding="utf-8"))
|
||||
records = [r for r in data["records"] if r["status"] == "IMPL_DIFF"]
|
||||
index = []
|
||||
for r in records:
|
||||
module = r["module"]
|
||||
art = PAIR[module]
|
||||
cls = r["class_name"]
|
||||
v1_jar = V1 / f"{art}-1.0-SNAPSHOT.jar"
|
||||
v2_jar = ROOT / module / "target" / f"{art}-2.0-SNAPSHOT.jar"
|
||||
raw1 = run(["javap", "-classpath", str(v1_jar), "-p", "-c", cls])
|
||||
raw2 = run(["javap", "-classpath", str(v2_jar), "-p", "-c", cls])
|
||||
n1 = norm_code(raw1)
|
||||
n2 = norm_code(raw2)
|
||||
cls_file = cls.replace(".", "_")
|
||||
p1 = OUT / f"{cls_file}.v1.javap.txt"
|
||||
p2 = OUT / f"{cls_file}.v2.javap.txt"
|
||||
pd = OUT / f"{cls_file}.diff.txt"
|
||||
p1.write_text(n1, encoding="utf-8")
|
||||
p2.write_text(n2, encoding="utf-8")
|
||||
diff = "".join(
|
||||
difflib.unified_diff(
|
||||
n1.splitlines(keepends=True),
|
||||
n2.splitlines(keepends=True),
|
||||
fromfile=f"{cls}.v1",
|
||||
tofile=f"{cls}.v2",
|
||||
n=3,
|
||||
)
|
||||
)
|
||||
pd.write_text(diff, encoding="utf-8")
|
||||
index.append(
|
||||
{
|
||||
"module": module,
|
||||
"class_name": cls,
|
||||
"v1_file": str(p1),
|
||||
"v2_file": str(p2),
|
||||
"diff_file": str(pd),
|
||||
"diff_lines": len(diff.splitlines()),
|
||||
}
|
||||
)
|
||||
idx = {
|
||||
"generated_at": datetime.utcnow().isoformat() + "Z",
|
||||
"count": len(index),
|
||||
"items": index,
|
||||
}
|
||||
(OUT / "INDEX.json").write_text(json.dumps(idx, ensure_ascii=False, indent=2), encoding="utf-8")
|
||||
print(str(OUT))
|
||||
print(f"IMPL_DIFF files: {len(index)}")
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
+187
@@ -0,0 +1,187 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
将 Maven 模块 `src/main/java/cn/cloudwalk/elevator` 与 V1 CFR 反编译树（decompile_v1_elevator_jars.sh 输出）按类文件路径配对，写相似度报告。
|
||||
|
||||
环境变量:
|
||||
V1_DECOMP_ROOT 默认: <maven根>/tools/v1-decompiled/cfr-from-cw-lib-current
|
||||
MV_ROOT 默认: 本脚本上溯 3 级到 maven-cw-elevator-application
|
||||
OUT_MD 默认: tools/elevator_api_parity/report/SOURCE-PARITY-BY-CLASS-decomp.md
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import difflib
|
||||
import os
|
||||
import re
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
SCR = Path(__file__).resolve()
|
||||
# scripts -> elevator_api_parity -> tools -> maven-cw-elevator-application
|
||||
MV_ROOT = Path(os.environ.get("MV_ROOT", SCR.parents[3]))
|
||||
_DEFAULT_V1 = MV_ROOT / "tools" / "v1-decompiled" / "cfr-from-cw-lib-current"
|
||||
|
||||
# Maven 模块名 -> V1 反编译目录名（脚本输出的子目录前缀）
|
||||
_PAIR = [
|
||||
("cw-elevator-application-common", "cw-elevator-application-common-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-data", "cw-elevator-application-data-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-service", "cw-elevator-application-service-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-web", "cw-elevator-application-web-1.0-SNAPSHOT"),
|
||||
]
|
||||
|
||||
|
||||
def strip_java_comments(text: str) -> str:
|
||||
text = re.sub(r"/\*.*?\*/", "", text, flags=re.DOTALL)
|
||||
text = re.sub(r"//[^\n]*", "", text)
|
||||
lines = [ln.strip() for ln in text.splitlines() if ln.strip()]
|
||||
return "\n".join(lines)
|
||||
|
||||
|
||||
def similarity(a: str, b: str) -> float:
|
||||
ca, cb = strip_java_comments(a), strip_java_comments(b)
|
||||
return difflib.SequenceMatcher(a=ca, b=cb).ratio()
|
||||
|
||||
|
||||
def tag_only_mv(rel: str) -> str:
|
||||
if "TenantVisitorFloor" in rel:
|
||||
return "租户访客楼层策略(V2)"
|
||||
if rel.endswith("package-info.java"):
|
||||
return "package-info"
|
||||
if "CloudwalkSessionHolderConfiguration" in rel:
|
||||
return "Session 扩展"
|
||||
if "ElevatorRemoteIo" in rel:
|
||||
return "远程 IO 池"
|
||||
if "DavinciStorageBeansConfiguration" in rel or "OpenFeignFileStorageManager" in rel:
|
||||
return "Davinci 存储"
|
||||
if "ZoneTreeCollectors" in rel:
|
||||
return "工具类"
|
||||
return ""
|
||||
|
||||
|
||||
def main() -> int:
|
||||
v1_root_parent = Path(os.environ.get("V1_DECOMP_ROOT", _DEFAULT_V1)).resolve()
|
||||
if not v1_root_parent.is_dir():
|
||||
print(f"ERROR: V1 decompile dir missing: {v1_root_parent}", file=sys.stderr)
|
||||
print("Run: maven-cw-elevator-application/scripts/decompile_v1_elevator_jars.sh", file=sys.stderr)
|
||||
return 1
|
||||
|
||||
out_md = Path(
|
||||
os.environ.get(
|
||||
"OUT_MD",
|
||||
MV_ROOT / "tools/elevator_api_parity/report/SOURCE-PARITY-BY-CLASS-decomp.md",
|
||||
)
|
||||
)
|
||||
|
||||
lines: list[str] = []
|
||||
lines.append("# Maven 源码 vs V1 自动反编译（cw_lib jar / CFR）逐类比对")
|
||||
lines.append("")
|
||||
lines.append(f"- **V1 根**: `{v1_root_parent}`（`decompile_v1_elevator_jars.sh` 生成）")
|
||||
lines.append(f"- **V2 根**: `{MV_ROOT}/{{module}}/src/main/java/cn/cloudwalk/elevator`")
|
||||
lines.append("- **相似度**: 去注释后 `SequenceMatcher.ratio`（CFR 噪声会导致分数偏低）")
|
||||
lines.append("")
|
||||
|
||||
all_rows: list[tuple[str, str, float]] = []
|
||||
tot = {"paired": 0, "ge98": 0, "85": 0, "lt": 0, "o1": 0, "o2": 0}
|
||||
summary_rows: list[str] = []
|
||||
|
||||
for mod_mvn, mod_v1dir in _PAIR:
|
||||
v1_elev = v1_root_parent / mod_v1dir / "cn/cloudwalk/elevator"
|
||||
mv_elev = MV_ROOT / mod_mvn / "src/main/java/cn/cloudwalk/elevator"
|
||||
if not v1_elev.is_dir():
|
||||
summary_rows.append(f"| {mod_mvn} | — | — | **V1 侧目录不存在**: `{v1_elev}` |")
|
||||
continue
|
||||
if not mv_elev.is_dir():
|
||||
summary_rows.append(f"| {mod_mvn} | — | — | **Maven 侧目录不存在** |")
|
||||
continue
|
||||
|
||||
v1_files = {str(p.relative_to(v1_elev)).replace("\\", "/"): p for p in v1_elev.rglob("*.java")}
|
||||
mv_files = {str(p.relative_to(mv_elev)).replace("\\", "/"): p for p in mv_elev.rglob("*.java")}
|
||||
cnt = {"paired": 0, "ge98": 0, "85_98": 0, "lt85": 0, "only_v1": 0, "only_mv": 0}
|
||||
|
||||
for rel in sorted(set(v1_files) | set(mv_files)):
|
||||
p1, pm = v1_files.get(rel), mv_files.get(rel)
|
||||
if p1 and pm:
|
||||
cnt["paired"] += 1
|
||||
try:
|
||||
s = similarity(
|
||||
p1.read_text(encoding="utf-8", errors="replace"),
|
||||
pm.read_text(encoding="utf-8", errors="replace"),
|
||||
)
|
||||
except OSError:
|
||||
s = 0.0
|
||||
if s >= 0.98:
|
||||
cnt["ge98"] += 1
|
||||
elif s >= 0.85:
|
||||
cnt["85_98"] += 1
|
||||
else:
|
||||
cnt["lt85"] += 1
|
||||
all_rows.append((f"{mod_mvn}/{rel}", s))
|
||||
elif p1:
|
||||
cnt["only_v1"] += 1
|
||||
else:
|
||||
cnt["only_mv"] += 1
|
||||
|
||||
tot["paired"] += cnt["paired"]
|
||||
tot["ge98"] += cnt["ge98"]
|
||||
tot["85"] += cnt["85_98"]
|
||||
tot["lt"] += cnt["lt85"]
|
||||
tot["o1"] += cnt["only_v1"]
|
||||
tot["o2"] += cnt["only_mv"]
|
||||
|
||||
summary_rows.append(
|
||||
f"| {mod_mvn} | {cnt['paired']} | {cnt['ge98']} | {cnt['85_98']} | {cnt['lt85']} | "
|
||||
f"{cnt['only_v1']} | {cnt['only_mv']} |"
|
||||
)
|
||||
|
||||
lines.append("## 1. 汇总（四模块来自 cw_lib jar 反编译）")
|
||||
lines.append("")
|
||||
lines.append(
|
||||
"| 模块 | 配对 | ≥0.98 | 0.85–0.98 | <0.85 | 仅 V1 | 仅 Maven |"
|
||||
)
|
||||
lines.append("|------|------|-------|-----------|-------|-------|------------|")
|
||||
lines.extend(summary_rows)
|
||||
lines.append(
|
||||
f"| **合计** | **{tot['paired']}** | **{tot['ge98']}** | **{tot['85']}** | **{tot['lt']}** | **{tot['o1']}** | **{tot['o2']}** |"
|
||||
)
|
||||
lines.append("")
|
||||
|
||||
# 仅 Maven
|
||||
lines.append("## 2. 仅 Maven（相对 `cn/cloudwalk/elevator`）")
|
||||
lines.append("")
|
||||
for mod_mvn, mod_v1dir in _PAIR:
|
||||
v1_elev = v1_root_parent / mod_v1dir / "cn/cloudwalk/elevator"
|
||||
mv_elev = MV_ROOT / mod_mvn / "src/main/java/cn/cloudwalk/elevator"
|
||||
if not mv_elev.is_dir() or not v1_elev.is_dir():
|
||||
continue
|
||||
v1_set = {str(p.relative_to(v1_elev)).replace("\\", "/") for p in v1_elev.rglob("*.java")}
|
||||
extras = []
|
||||
for p in sorted(mv_elev.rglob("*.java")):
|
||||
rel = str(p.relative_to(mv_elev)).replace("\\", "/")
|
||||
if rel not in v1_set:
|
||||
tg = tag_only_mv(rel)
|
||||
extras.append(f"- `{rel}`" + (f" — *{tg}*" if tg else ""))
|
||||
if extras:
|
||||
lines.append(f"### {mod_mvn}")
|
||||
lines.extend(extras)
|
||||
lines.append("")
|
||||
|
||||
worst = sorted(all_rows, key=lambda x: x[1])[:35]
|
||||
best = sorted(all_rows, key=lambda x: -x[1])[:15]
|
||||
lines.append("## 3. 样本：最高 / 最低相似度路径")
|
||||
lines.append("")
|
||||
lines.append("### 最高")
|
||||
for path, s in best:
|
||||
lines.append(f"- `{path}` **{s:.4f}**")
|
||||
lines.append("")
|
||||
lines.append("### 最低（优先 IDE diff）")
|
||||
for path, s in worst:
|
||||
lines.append(f"- `{path}` **{s:.4f}**")
|
||||
|
||||
out_md.parent.mkdir(parents=True, exist_ok=True)
|
||||
out_md.write_text("\n".join(lines) + "\n", encoding="utf-8")
|
||||
print(out_md)
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
Executable
+266
@@ -0,0 +1,266 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
V1（CFR 反编译 cw_lib） vs V2（Maven 源码）全量类级比对，输出：
|
||||
- Markdown 报告（汇总 + 分桶完整清单 + 问题梳理）
|
||||
- 可选 JSON（--json）
|
||||
|
||||
环境变量同 source_parity_scan.py：V1_DECOMP_ROOT、MV_ROOT
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import difflib
|
||||
import json
|
||||
import os
|
||||
import re
|
||||
import sys
|
||||
from datetime import datetime, timezone
|
||||
from pathlib import Path
|
||||
|
||||
SCR = Path(__file__).resolve()
|
||||
MV_ROOT = Path(os.environ.get("MV_ROOT", SCR.parents[3]))
|
||||
_DEFAULT_V1 = MV_ROOT / "tools" / "v1-decompiled" / "cfr-from-cw-lib-current"
|
||||
|
||||
_PAIR = [
|
||||
("cw-elevator-application-common", "cw-elevator-application-common-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-data", "cw-elevator-application-data-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-service", "cw-elevator-application-service-1.0-SNAPSHOT"),
|
||||
("cw-elevator-application-web", "cw-elevator-application-web-1.0-SNAPSHOT"),
|
||||
]
|
||||
|
||||
|
||||
def strip_java_comments(text: str) -> str:
|
||||
text = re.sub(r"/\*.*?\*/", "", text, flags=re.DOTALL)
|
||||
text = re.sub(r"//[^\n]*", "", text)
|
||||
lines = [ln.strip() for ln in text.splitlines() if ln.strip()]
|
||||
return "\n".join(lines)
|
||||
|
||||
|
||||
def similarity(a: str, b: str) -> float:
|
||||
ca, cb = strip_java_comments(a), strip_java_comments(b)
|
||||
return difflib.SequenceMatcher(a=ca, b=cb).ratio()
|
||||
|
||||
|
||||
def tag_only_mv(rel: str) -> str:
|
||||
if "TenantVisitorFloor" in rel:
|
||||
return "V2扩展-租户访客楼层"
|
||||
if rel.endswith("package-info.java"):
|
||||
return "工程化-package-info"
|
||||
if "CloudwalkSessionHolderConfiguration" in rel:
|
||||
return "V2扩展-Session"
|
||||
if "ElevatorRemoteIo" in rel:
|
||||
return "V2扩展-远程IO线程池"
|
||||
if "DavinciStorageBeansConfiguration" in rel or "OpenFeignFileStorageManager" in rel:
|
||||
return "V2扩展-Davinci存储"
|
||||
if "ZoneTreeCollectors" in rel:
|
||||
return "V2独有-工具类"
|
||||
return "V2独有-待分类"
|
||||
|
||||
|
||||
def review_priority(rel: str) -> str:
|
||||
r = rel.replace("\\", "/")
|
||||
if "/controller/" in r or r.endswith("Controller.java"):
|
||||
return "高-Controller"
|
||||
if "FeignClient" in r or "/client/" in r:
|
||||
return "高-Feign/Client"
|
||||
if "ServiceImpl" in r or "/impl/" in r:
|
||||
return "高-Service/DaoImpl"
|
||||
if "/mapper/" in r or "Mapper.java" in r:
|
||||
return "中-Mapper接口"
|
||||
if "/dto/" in r or "/param/" in r or "/form/" in r or "/result/" in r:
|
||||
return "中-DTO表单"
|
||||
if "/em/" in r or "Enum.java" in r:
|
||||
return "低-枚举(CFR噪声常见)"
|
||||
if "/util/" in r:
|
||||
return "低-工具类"
|
||||
return "中-其他"
|
||||
|
||||
|
||||
def bucket(score: float) -> str:
|
||||
if score >= 0.98:
|
||||
return "A-高度一致"
|
||||
if score >= 0.85:
|
||||
return "B-轻度差异"
|
||||
return "C-显著差异"
|
||||
|
||||
|
||||
def main() -> int:
|
||||
ap = argparse.ArgumentParser()
|
||||
ap.add_argument("--json", action="store_true", help="额外写出 JSON")
|
||||
ap.add_argument(
|
||||
"--out",
|
||||
default=str(MV_ROOT / "tools/elevator_api_parity/report/V1-V2-DIFF-ISSUES.md"),
|
||||
help="Markdown 输出路径",
|
||||
)
|
||||
args = ap.parse_args()
|
||||
|
||||
v1_root_parent = Path(os.environ.get("V1_DECOMP_ROOT", _DEFAULT_V1)).resolve()
|
||||
if not v1_root_parent.is_dir():
|
||||
print(f"ERROR: V1 decompile dir missing: {v1_root_parent}", file=sys.stderr)
|
||||
return 1
|
||||
|
||||
records: list[dict] = []
|
||||
only_mv: list[tuple[str, str, str]] = []
|
||||
totals = {"paired": 0, "A": 0, "B": 0, "C": 0, "only_v1": 0, "only_mv": 0}
|
||||
|
||||
for mod_mvn, mod_v1dir in _PAIR:
|
||||
v1_elev = v1_root_parent / mod_v1dir / "cn/cloudwalk/elevator"
|
||||
mv_elev = MV_ROOT / mod_mvn / "src/main/java/cn/cloudwalk/elevator"
|
||||
if not v1_elev.is_dir() or not mv_elev.is_dir():
|
||||
continue
|
||||
|
||||
v1_files = {str(p.relative_to(v1_elev)).replace("\\", "/"): p for p in v1_elev.rglob("*.java")}
|
||||
mv_files = {str(p.relative_to(mv_elev)).replace("\\", "/"): p for p in mv_elev.rglob("*.java")}
|
||||
|
||||
for rel in sorted(set(v1_files) | set(mv_files)):
|
||||
p1, pm = v1_files.get(rel), mv_files.get(rel)
|
||||
if p1 and pm:
|
||||
totals["paired"] += 1
|
||||
try:
|
||||
t1 = p1.read_text(encoding="utf-8", errors="replace")
|
||||
t2 = pm.read_text(encoding="utf-8", errors="replace")
|
||||
s = similarity(t1, t2)
|
||||
except OSError:
|
||||
s = 0.0
|
||||
t1 = t2 = ""
|
||||
bk = bucket(s)
|
||||
if bk.startswith("A"):
|
||||
totals["A"] += 1
|
||||
elif bk.startswith("B"):
|
||||
totals["B"] += 1
|
||||
else:
|
||||
totals["C"] += 1
|
||||
pri = review_priority(rel)
|
||||
records.append(
|
||||
{
|
||||
"module": mod_mvn,
|
||||
"relative_path": rel,
|
||||
"similarity": round(s, 6),
|
||||
"bucket": bk,
|
||||
"review_priority": pri,
|
||||
"v1_file": str(p1),
|
||||
"v2_file": str(pm),
|
||||
}
|
||||
)
|
||||
elif p1:
|
||||
totals["only_v1"] += 1
|
||||
records.append(
|
||||
{
|
||||
"module": mod_mvn,
|
||||
"relative_path": rel,
|
||||
"similarity": None,
|
||||
"bucket": "ONLY_V1",
|
||||
"review_priority": review_priority(rel),
|
||||
"v1_file": str(p1),
|
||||
"v2_file": None,
|
||||
}
|
||||
)
|
||||
else:
|
||||
totals["only_mv"] += 1
|
||||
tg = tag_only_mv(rel)
|
||||
only_mv.append((mod_mvn, rel, tg))
|
||||
records.append(
|
||||
{
|
||||
"module": mod_mvn,
|
||||
"relative_path": rel,
|
||||
"similarity": None,
|
||||
"bucket": "ONLY_MAVEN",
|
||||
"review_priority": tg,
|
||||
"v1_file": None,
|
||||
"v2_file": str(pm),
|
||||
}
|
||||
)
|
||||
|
||||
out_md = Path(args.out)
|
||||
out_md.parent.mkdir(parents=True, exist_ok=True)
|
||||
ts = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
|
||||
|
||||
md: list[str] = []
|
||||
md.append("# V1 vs V2 源码差异与问题梳理（全自动比对）")
|
||||
md.append("")
|
||||
md.append(f"- **生成时间（UTC）**: {ts}")
|
||||
md.append(f"- **V1**: `{v1_root_parent}`（cw_lib jar → CFR）")
|
||||
md.append(f"- **V2**: `{MV_ROOT}/{{module}}/src/main/java/cn/cloudwalk/elevator`")
|
||||
md.append("- **相似度**: 去注释后全文 `SequenceMatcher.ratio`；**低分≠业务错误**（CFR 变量名/枚举顺序等）。")
|
||||
md.append("")
|
||||
md.append("## 1. 结论摘要")
|
||||
md.append("")
|
||||
md.append("| 维度 | 数量 | 说明 |")
|
||||
md.append("|------|------|------|")
|
||||
md.append(f"| 配对类 | {totals['paired']} | 同名路径 `.java` |")
|
||||
md.append(f"| A 桶 ≥0.98 | {totals['A']} | 文本高度接近 |")
|
||||
md.append(f"| B 桶 0.85–0.98 | {totals['B']} | 多为格式/注释/轻微差异，建议抽检 |")
|
||||
md.append(f"| C 桶 <0.85 | {totals['C']} | **优先 IDE diff**；枚举/DTO 常为 CFR 噪声 |")
|
||||
md.append(f"| 仅 V1 | {totals['only_v1']} | 当前四模块应为 0 |")
|
||||
md.append(f"| 仅 Maven | {totals['only_mv']} | V2 新增；访客楼层策略等为**已知扩展** |")
|
||||
md.append("")
|
||||
md.append("### 1.1 建议处置")
|
||||
md.append("")
|
||||
md.append("1. **仅 Maven**：`TenantVisitorFloor*` — 与 V1 对齐无关（产品已排除）；`package-info` — 忽略。")
|
||||
md.append("2. **C 桶 + 高优先级**（Controller / ServiceImpl / Feign）：优先人工对照业务分支与远程调用。")
|
||||
md.append("3. **C 桶 + 枚举/DTO**：优先视为 CFR 噪声，除非接口契约变更。")
|
||||
md.append("")
|
||||
|
||||
md.append("## 2. 仅 Maven 文件（问题/范围标注）")
|
||||
md.append("")
|
||||
for mod, rel, tg in sorted(only_mv, key=lambda x: (x[0], x[1])):
|
||||
md.append(f"- `{mod}/{rel}` — **{tg}**")
|
||||
md.append("")
|
||||
|
||||
md.append("## 3. C 桶（<0.85）按复核优先级分组")
|
||||
md.append("")
|
||||
c_records = [r for r in records if r.get("similarity") is not None and r["similarity"] < 0.85]
|
||||
by_pri: dict[str, list[dict]] = {}
|
||||
for r in c_records:
|
||||
by_pri.setdefault(r["review_priority"], []).append(r)
|
||||
for pri in sorted(by_pri.keys()):
|
||||
md.append(f"### {pri}")
|
||||
md.append("")
|
||||
md.append("| 路径 | 相似度 |")
|
||||
md.append("|------|--------|")
|
||||
for r in sorted(by_pri[pri], key=lambda x: -x["similarity"]):
|
||||
md.append(f"| `{r['module']}/{r['relative_path']}` | {r['similarity']:.4f} |")
|
||||
md.append("")
|
||||
|
||||
md.append("## 4. 完整配对表（全部类）")
|
||||
md.append("")
|
||||
md.append("| 模块 | 相对路径 | 相似度 | 分桶 | 复核优先级 |")
|
||||
md.append("|------|----------|--------|------|------------|")
|
||||
for r in sorted(
|
||||
[x for x in records if x.get("similarity") is not None],
|
||||
key=lambda x: (x["module"], x["relative_path"]),
|
||||
):
|
||||
md.append(
|
||||
f"| {r['module']} | `{r['relative_path']}` | {r['similarity']:.4f} | {r['bucket']} | {r['review_priority']} |"
|
||||
)
|
||||
md.append("")
|
||||
md.append("### 仅 V1 / 仅 Maven 行")
|
||||
md.append("")
|
||||
for r in records:
|
||||
if r["bucket"] in ("ONLY_V1", "ONLY_MAVEN"):
|
||||
md.append(f"- `{r['module']}/{r['relative_path']}` — **{r['bucket']}** — {r['review_priority']}")
|
||||
md.append("")
|
||||
|
||||
md.append("## 5. 生成命令")
|
||||
md.append("")
|
||||
md.append("```bash")
|
||||
md.append("cd maven-cw-elevator-application")
|
||||
md.append("./scripts/decompile_v1_elevator_jars.sh # 若尚未反编译 V1 jar")
|
||||
md.append("python3 tools/elevator_api_parity/scripts/v1_v2_diff_issues_report.py")
|
||||
md.append("```")
|
||||
md.append("")
|
||||
|
||||
out_md.write_text("\n".join(md) + "\n", encoding="utf-8")
|
||||
print(out_md)
|
||||
|
||||
if args.json:
|
||||
out_json = out_md.with_suffix(".json")
|
||||
out_json.write_text(json.dumps({"generated": ts, "totals": totals, "records": records}, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
|
||||
print(out_json)
|
||||
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
Reference in New Issue
Block a user