#!/usr/bin/env python3 # -*- coding: utf-8 -*- """上游(TDC / TDS / ssw-common)测试用例**普查** + 与台账对账。 用法: python3 scripts/upstream-test-census.py # 打印普查结果(TSV) python3 scripts/upstream-test-census.py --check # 与台账对账(有差异 ⇒ 退出码 1) 上游克隆路径(可用环境变量覆盖): UPSTREAM_TDC_CLONE 默认 /tmp/sswtdc-probe UPSTREAM_TDS_CLONE 默认 /tmp/sswtds-probe 台账: scripts/upstream-test-ledger.tsv(默认) 🔴 **为什么要有它**:上游的测试用例**没有统一入口** —— `*Test.java` 靠 surefire 跑,`*SelfTest.java`(纯 main)surefire **不跑**、 上游自己靠 `tools/run-java-selftests.sh` 人工跑 ⇒ 「上游到底有哪些测试用例、我们搬没搬」这件事**必须有一份可复算的清单**, 否则上游新增一个类而我们没看见 ⇒ 静默漂移。 🔴 **克隆不在场 ⇒ 不判红**(只跳过该侧的完整性校验)—— 台账里登记的类**仍然**按台账校验(PORTED 行必须真有对位用例)。 """ import argparse import os import re import sys DEFAULT_TDC_CLONE = os.environ.get("UPSTREAM_TDC_CLONE", "/tmp/sswtdc-probe") DEFAULT_TDS_CLONE = os.environ.get("UPSTREAM_TDS_CLONE", "/tmp/sswtds-probe") DEFAULT_LEDGER = os.path.join( os.path.dirname(os.path.abspath(__file__)), "upstream-test-ledger.tsv" ) REPO_LABEL = {"tdc": "TDC", "tds": "TDS"} def scan_clone(root, label): """扫一个克隆根,返回 {类全名: (仓库标签, 相对路径, kind)}。""" found = {} if not os.path.isdir(root): return found, False present = True for dirpath, _dirs, files in os.walk(root): if "/src/test/java" not in dirpath.replace(os.sep, "/"): continue for f in files: if not f.endswith(".java"): continue path = os.path.join(dirpath, f) rel = os.path.relpath(path, root) try: with open(path, encoding="utf-8", errors="replace") as fh: text = fh.read() except OSError: continue m = re.search(r"^\s*package\s+([\w.]+)\s*;", text, re.M) pkg = m.group(1) if m else "?" cls = f[:-5] if "@Test" in text: kind = "JUNIT" elif re.search(r"public\s+static\s+void\s+main", text): kind = "SELFTEST" else: kind = "OTHER" found["%s.%s" % (pkg, cls)] = (label, rel, kind) return found, present def read_ledger(path): """读台账 ⇒ {类全名: 行字典}。""" rows = {} if not os.path.isfile(path): return rows with open(path, encoding="utf-8") as fh: for lineno, raw in enumerate(fh, 1): line = raw.rstrip("\n") if not line.strip() or line.lstrip().startswith("#"): continue parts = line.split("\t") if len(parts) < 8: sys.stderr.write( "⚠️ 台账 %s:%d 列数不足 8(应为 8 列)\n" % (path, lineno) ) continue rows[parts[0].strip()] = { "class": parts[0].strip(), "repo": parts[1].strip(), "file": parts[2].strip(), "kind": parts[3].strip(), "assertions": parts[4].strip(), "counterpart": parts[5].strip(), "status": parts[6].strip(), "reason": parts[7].strip(), "lineno": lineno, } return rows def main(): ap = argparse.ArgumentParser() ap.add_argument("--tdc-clone", default=DEFAULT_TDC_CLONE) ap.add_argument("--tds-clone", default=DEFAULT_TDS_CLONE) ap.add_argument("--ledger", default=DEFAULT_LEDGER) ap.add_argument("--check", action="store_true") args = ap.parse_args() census = {} missing_clones = [] for root, label in ((args.tdc_clone, "TDC"), (args.tds_clone, "TDS")): found, present = scan_clone(root, label) if not present: missing_clones.append("%s(%s)" % (label, root)) for cls, (repo, rel, kind) in found.items(): if cls in census: # ssw-common 在两个克隆里都有 ⇒ 合并仓库标签 prev_repo = census[cls][0] if prev_repo != repo: census[cls] = ("COMMON", census[cls][1], census[cls][2]) continue census[cls] = (repo, rel, kind) if not args.check: print("class\trepo\tfile\tkind") for cls in sorted(census): repo, rel, kind = census[cls] print("%s\t%s\t%s\t%s" % (cls, repo, rel, kind)) return 0 ledger = read_ledger(args.ledger) problems = [] # ① 台账里的每一行 ⇒ 上游文件必须真的在(克隆在场时) for cls, row in sorted(ledger.items()): root = args.tdc_clone if row["repo"] in ("TDC", "COMMON") else args.tds_clone if not os.path.isdir(root): continue path = os.path.join(root, row["file"]) if not os.path.isfile(path): problems.append( "台账 %s:%d 引用的上游文件不存在:%s" % (args.ledger, row["lineno"], path) ) # ② 上游有的测试类 ⇒ 台账必须登记(克隆在场时) if census: for cls in sorted(census): if cls not in ledger: repo, rel, kind = census[cls] problems.append( "上游新增测试类未登记台账:%s(%s,%s)⇒ 请在 %s 补一行并判定搬/不搬" % (cls, repo, rel, args.ledger) ) # ③ 台账里登记了、上游却找不到了(两侧克隆都在场时) if census and not missing_clones: for cls in sorted(ledger): if cls not in census: problems.append( "台账登记了上游已不存在的测试类:%s(台账 %s:%d)" % (cls, args.ledger, ledger[cls]["lineno"]) ) print("普查:上游测试类 %d 个 · 台账登记 %d 个" % (len(census), len(ledger))) if missing_clones: print( "⚠️ 上游克隆不在场(%s)⇒ 跳过「上游有、台账无」与「上游文件存在」两项校验" % "、".join(missing_clones) ) if problems: print("❌ 对账不通过:") for p in problems: print(" · %s" % p) return 1 print("✅ 对账通过:上游测试类与台账一致") return 0 if __name__ == "__main__": sys.exit(main())