dsh-plc-decode 0.2.1 → 0.2.3

package/assets/wincc_i18n.py

@@ -1,504 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""WinCC flexible / SMART 项目文本导出表（.xls/.xlsx）多语言处理器。
-
-项目文本导出表的格式契约（WinCC flexible 导出 → Excel）：
-  - 第 1 行  $_Languages
-  - 第 2 行  语言名，如「4(2) 简体中文(中国大陆)」「9(2) 英语(英国)」；第 1 列＝源语言，第 2 列＝目标语言
-  - 第 3 行  $_Typ(Screens) / $_Attrib(MultiLanguage)
-  - 若干 // 注释行
-  - 之后每行：第 1 列＝源文本（画面上显示什么就以什么为键），第 2 列＝目标语言文本
-  - 文本可能被 <text>...</text> 包裹；换行是**字面反斜杠 n**（两个字符），不是真换行
-  - 目标列出现 $_ignore ＝ 该源文本在工程里存在多份不同译文，导出时被替换；导入时该行被忽略，必须原样保留
-
-三条硬规则（本脚本按它判定）：
-  1. 第 1 列是匹配键——改了就对不上，绝不能动
-  2. $_ignore 行照样要给译文（它是多译文冲突被导出挡掉的，不是不用翻）
-  3. 一源一译：同一句源文本全表只能有一个译文，否则导回后工程里又变成多译文、下次导出又变 $_ignore
-
-第四条是校对判据：源文本不含中文（纯编号/代号）而译文比源长 → 判 CODE_EXTRA，
-即别行的功能词串到这一行了（Q10 top box 这类），应只留编号。
-
-子命令：
-  scan   --files a.xls b.xls ... --outdir DIR        列出每行的分型（未译/空译/忽略/编号/已有译文），出清单与报告
-  apply  --files ... --dict 映射.tsv --outdir DIR    按映射写回目标列，出可直接导回的 xls（源是 xls 就出 xls）
-  check  --src a.xls --out a_译文.xls               自检：行数/键列/忽略行/中文残留/一源一译
-
-映射文件格式（TSV，制表符分隔，UTF-8，可带表头）：
-  源文本<TAB>译文
-  换行写 \n（字面两个字符）
-"""
-
-import argparse
-import json
-import os
-import re
-import shutil
-import subprocess
-import sys
-import tempfile
-
-TEXT_RE = re.compile(r"^<text>(.*)</text>\s*$", re.S)
-CJK_RE = re.compile(r"[\u4e00-\u9fff\u3000-\u303f\uff00-\uffef]")
-# 日期/时间格式串（yyyy-M-d H:mm:ss 之类）两列本来就不一样，不能算串位
-DATEFMT_RE = re.compile(r"^[yMdHmsAP:/.\s-]+$")
-IGNORE = "$_ignore"
-
-
-def looks_datefmt(s):
-    s = s.strip()
-    return len(s) >= 6 and bool(DATEFMT_RE.match(s))
-
-# 项目文本导出表里非正文的头部行数上限（语言行/类型行/注释行）
-HEADER_SCAN_LIMIT = 12
-
-LO_KIT_CANDIDATES = [
-    os.environ.get("WINCC_LO_KIT", ""),
-    r"D:\DSH\Desktop\resources\app.asar.unpacked\dsh\node_modules\@deepseek-ai\libreoffice-kit\lib\cli.js",
-]
-NODE_CANDIDATES = [
-    os.environ.get("WINCC_LO_NODE", ""),
-    r"D:\DSH\Desktop\resources\runtime\primary-runtime\dependencies\node\bin\node.exe",
-    r"D:\DSH\home\dsh-runtimes\dsh-primary-runtime\dependencies\node\bin\node.exe",
-]
-
-
-# ───────────────────────── 文件格式 ─────────────────────────
-
-def _find_lo_kit():
-    cli = next((p for p in LO_KIT_CANDIDATES if p and os.path.exists(p)), None)
-    node = next((p for p in NODE_CANDIDATES if p and os.path.exists(p)), None)
-    return (node, cli) if cli and node else (None, None)
-
-
-def convert(src, dst):
-    """借 LibreOffice Kit 转格式（xls <-> xlsx）。返回 dst。"""
-    node, cli = _find_lo_kit()
-    if not cli:
-        raise SystemExit(
-            f"需要 LibreOffice Kit 转换 {os.path.basename(src)}，但没找到 cli.js。\n"
-            "可设环境变量 WINCC_LO_KIT=<cli.js 路径> / WINCC_LO_NODE=<node 路径>。"
-        )
-    if os.path.exists(dst):
-        os.remove(dst)
-    out = subprocess.run([node, cli, "convert", "--input", src, "--output", dst],
-                         capture_output=True, text=True, encoding="utf-8", errors="replace")
-    if out.returncode != 0 or not os.path.exists(dst):
-        raise SystemExit(f"转换失败（{os.path.basename(src)} -> {os.path.basename(dst)}）:\n{out.stdout}\n{out.stderr}")
-    return dst
-
-
-def as_xlsx(path, workdir):
-    """拿到一个可被 openpyxl 读的 xlsx 路径。"""
-    if path.lower().endswith(".xlsx"):
-        return path
-    os.makedirs(workdir, exist_ok=True)
-    return convert(path, os.path.join(workdir, os.path.splitext(os.path.basename(path))[0] + ".xlsx"))
-
-
-# ───────────────────────── 解析 ─────────────────────────
-
-def parse(path, workdir):
-    import openpyxl
-    xlsx = as_xlsx(path, workdir)
-    wb = openpyxl.load_workbook(xlsx, data_only=True)
-    ws = wb.worksheets[0]
-
-    rows = []
-    for r in ws.iter_rows(values_only=True):
-        # 整行保留：导入是按列对语言的，多余列必须原样带回去
-        rows.append(["" if v is None else str(v) for v in r])
-    for row in rows:
-        while len(row) < 2:
-            row.append("")
-
-    # 头部：直到出现第一行非注释、非表头的内容
-    body_start = 0
-    langs = [rows[1][0] if len(rows) > 1 else "", rows[1][1] if len(rows) > 1 else ""]
-    typ = ""
-    for i, row in enumerate(rows[:HEADER_SCAN_LIMIT]):
-        s = row[0].strip()
-        if s.startswith("$_Languages"):
-            continue
-        if s.startswith("$_Typ(") or s.startswith("$_Attrib("):
-            typ = typ or s
-            continue
-        if s.startswith("//"):
-            continue
-        if re.match(r"^\d+\(\d+\)", s):
-            continue
-        body_start = i
-        break
-
-    return {"path": path, "rows": rows, "body_start": body_start, "langs": langs, "typ": typ,
-            "xlsx": xlsx, "sheet": ws.title}
-
-
-def core_of(s):
-    """去掉 <text> 包裹，返回 (是否包裹, 内容)。"""
-    m = TEXT_RE.match(s)
-    return (True, m.group(1)) if m else (False, s)
-
-
-def classify(src, dst):
-    """给一行分型。返回 (type, 说明)。"""
-    if not src.strip():
-        return ("EMPTY_KEY", "源文本为空")
-    wrapped, core = core_of(src)
-    if dst == IGNORE:
-        return ("IGNORE", "多译文冲突，导出时已替换，导入会被忽略")
-    if not dst.strip():
-        return ("UNTRANSLATED", "目标列空")
-    _, core_b = core_of(dst)
-    if CJK_RE.search(core_b):
-        return ("UNTRANSLATED", "目标列还是中文" if core_b.strip() != core.strip() else "目标列＝源文本")
-    if not CJK_RE.search(core) and core.strip() == core_b.strip():
-        return ("SAME_AS_SRC", "非中文文本，两列相同（编号/符号/日期格式）")
-    if not CJK_RE.search(core) and not looks_datefmt(core) and len(core_b.strip()) > len(core.strip()):
-        # 源是纯编号/代号，译文却更长 → 别行的功能词串过来了（Q10 top box 这种）
-        return ("CODE_EXTRA", "源无中文，译文多出内容（疑串位，应只留编号）")
-    return ("HAS_TRANS", "已有译文，需人工核错译/串位")
-
-
-def scan(p):
-    out = []
-    for idx in range(p["body_start"], len(p["rows"])):
-        row = p["rows"][idx]
-        src, dst = row[0], row[1]
-        if not src.strip() and not dst.strip():
-            continue
-        t, note = classify(src, dst)
-        out.append({"row": idx + 1, "src": src, "dst": dst, "type": t, "note": note})
-    return out
-
-
-# ───────────────────────── 一源一译 ─────────────────────────
-
-def conflicts(items):
-    """同一源文本出现多个不同译文 → 导入后工程又变多译文。"""
-    seen = {}
-    for it in items:
-        if it["type"] == "IGNORE":
-            continue
-        _, core = core_of(it["src"])
-        if not core.strip():
-            continue
-        _, tgt = core_of(it["dst"])
-        if not tgt.strip():
-            continue
-        seen.setdefault(core, {}).setdefault(tgt, []).append(it["row"])
-    bad = {}
-    for core, m in seen.items():
-        if len(m) > 1:
-            bad[core] = {k: v for k, v in m.items()}
-    return bad
-
-
-# ───────────────────────── 命令 ─────────────────────────
-
-def cmd_scan(args):
-    os.makedirs(args.outdir, exist_ok=True)
-    work = os.path.join(args.outdir, "_work")
-    report = []
-    tsv_lines = ["文件\t行号\t分型\t源文本\t现有译文"]
-    stat = {}
-    for f in args.files:
-        p = parse(f, work)
-        items = scan(p)
-        name = os.path.basename(f)
-        c = {}
-        for it in items:
-            c[it["type"]] = c.get(it["type"], 0) + 1
-            tsv_lines.append("\t".join([name, str(it["row"]), it["type"], it["src"].replace("\t", " "), it["dst"].replace("\t", " ")]))
-        bad = conflicts(items)
-        stat[name] = c
-        report.append((name, p, items, bad))
-
-    list_path = os.path.join(args.outdir, "待译清单.tsv")
-    with open(list_path, "w", encoding="utf-8", newline="") as fh:
-        fh.write("\n".join(tsv_lines) + "\n")
-
-    md = ["# 项目文本分型报告", ""]
-    total_need = 0
-    for name, p, items, bad in report:
-        c = {}
-        for it in items:
-            c[it["type"]] = c.get(it["type"], 0) + 1
-        need = c.get("UNTRANSLATED", 0)
-        total_need += need
-        md.append(f"## {name}")
-        md.append("")
-        md.append(f"- 语言：{p['langs'][0]} → {p['langs'][1]}（{p['typ']}）")
-        md.append(f"- 正文行 {len(items)}：未译 {need} / 已有译文 {c.get('HAS_TRANS', 0)} / 忽略 {c.get('IGNORE', 0)} / 非中文同形 {c.get('SAME_AS_SRC', 0)} / 编号行挂词 {c.get('CODE_EXTRA', 0)}")
-        if bad:
-            md.append(f"- ⚠ 一源多译 {len(bad)} 组（导回后工程会产生多译文，必须统一）：")
-            for core, m in list(bad.items())[:20]:
-                md.append(f"    - {core!r} → " + "；".join(f"{k!r}(行{','.join(map(str, v))})" for k, v in m.items()))
-        md.append("")
-    md.insert(1, f"共 {len(args.files)} 个表，未译合计 {total_need} 条。清单：{list_path}")
-    md_path = os.path.join(args.outdir, "分型报告.md")
-    with open(md_path, "w", encoding="utf-8") as fh:
-        fh.write("\n".join(md) + "\n")
-
-    print("\n".join(md))
-    return 0
-
-
-def load_dict(path):
-    """映射：源文本 → 译文。支持 TSV（源\t译文）与 JSON（{"源":"译文"}）。"""
-    m = {}
-    if path.lower().endswith(".json"):
-        with open(path, encoding="utf-8") as fh:
-            raw = json.load(fh)
-        for k, v in raw.items():
-            m[k] = v
-        return m
-    with open(path, encoding="utf-8-sig") as fh:
-        for line in fh:
-            line = line.rstrip("\n").rstrip("\r")
-            if not line.strip() or line.startswith("#"):
-                continue
-            parts = line.split("\t")
-            if len(parts) < 2:
-                continue
-            k, v = parts[0], parts[1]
-            if k == "源文本":
-                continue
-            # 表里的换行是真换行（Excel 单元格内 Alt+Enter），映射文件一行放不下，
-            # 所以在映射里写 \n（字面两字符），读进来还原成真换行。
-            m[k.replace("\\n", "\n")] = v.replace("\\n", "\n")
-    return m
-
-
-def write_book(rows, path, sheet_name="Sheet1"):
-    """写 .xlsx（openpyxl）。"""
-    import openpyxl
-    wb = openpyxl.Workbook()
-    ws = wb.active
-    ws.title = sheet_name
-    for r in rows:
-        ws.append([("" if v is None else v) for v in r])
-    wb.save(path)
-    return path
-
-
-def write_xls(src, old_rows, rows, path, sheet_name="Sheet1"):
-    """写 Excel 97-2003 .xls。与源表同扩展名、同工作表名，导回软件时才认得出是哪类文本。
-
-    首选 Excel COM：拿源表复制一份，只改变化的格子——行数、列数、格式、其他列全原样
-    （xlwt 整表重建实测会吞行，394 行写完剩 393，之后整表错位）。
-    没装 Excel 才退到 xlwt，且写完必须回读核对行数。
-    """
-    errs = []
-    for fn in (_write_xls_via_com, _write_xls_via_xlwt):
-        try:
-            return fn(src, old_rows, rows, path, sheet_name)
-        except Exception as exc:  # noqa: BLE001 - 记录后试下一条路
-            errs.append(f"{fn.__name__}: {exc}")
-    raise SystemExit("写 .xls 两条路都不通：\n  " + "\n  ".join(errs))
-
-
-def _find_powershell():
-    for name in ("powershell.exe", "pwsh.exe", "powershell", "pwsh"):
-        p = shutil.which(name)
-        if p:
-            return p
-    return None
-
-
-_COM_SCRIPT = """
-param([string]$Xls, [string]$Plan)
-$ErrorActionPreference = 'Stop'
-$xl = New-Object -ComObject Excel.Application
-$xl.Visible = $false
-$xl.DisplayAlerts = $false
-try {
-  $wb = $xl.Workbooks.Open($Xls)
-  $ws = $wb.Worksheets.Item(1)
-  foreach ($line in Get-Content -LiteralPath $Plan -Encoding UTF8) {
-    if ($line.Trim() -eq '') { continue }
-    $i = $line.IndexOf("`t")
-    $r = [int]$line.Substring(0, $i)
-    $rest = $line.Substring($i + 1)
-    $j = $rest.IndexOf("`t")
-    $c = [int]$rest.Substring(0, $j)
-    $v = $rest.Substring($j + 1).Replace('\\n', "`n")
-    $ws.Cells.Item($r, $c).Value2 = $v
-  }
-  $wb.Save()
-  $wb.Close($true)
-} finally {
-  $xl.Quit()
-  [void][System.Runtime.InteropServices.Marshal]::ReleaseComObject($xl)
-}
-"""
-
-
-def _write_xls_via_com(src, old_rows, rows, path, sheet_name="Sheet1"):
-    """复制源表 → 用 Excel COM 只改变化的格子 → 保存。行数与格式天然不变。"""
-    ps = _find_powershell()
-    if not ps:
-        raise RuntimeError("找不到 powershell")
-    plan = []
-    for i in range(max(len(old_rows), len(rows))):
-        a = old_rows[i] if i < len(old_rows) else []
-        b = rows[i] if i < len(rows) else []
-        for c in range(max(len(a), len(b))):
-            va = a[c] if c < len(a) else ""
-            vb = b[c] if c < len(b) else ""
-            if str(va) != str(vb):
-                plan.append("%d\t%d\t%s" % (i + 1, c + 1, str(vb).replace("\n", "\\n")))
-    if not os.path.exists(path) or os.path.abspath(path) != os.path.abspath(src):
-        shutil.copyfile(src, path)
-    if not plan:
-        return path
-    tmpdir = tempfile.mkdtemp(prefix="wincc_xls_")
-    plan_path = os.path.join(tmpdir, "plan.tsv")
-    with open(plan_path, "w", encoding="utf-8", newline="") as fh:
-        fh.write("\n".join(plan) + "\n")
-    script = os.path.join(tmpdir, "apply.ps1")
-    with open(script, "w", encoding="utf-8-sig") as fh:
-        fh.write(_COM_SCRIPT)
-    out = subprocess.run([ps, "-NoProfile", "-ExecutionPolicy", "Bypass", "-File", script,
-                          "-Xls", path, "-Plan", plan_path],
-                         capture_output=True, text=True, encoding="utf-8", errors="replace")
-    if out.returncode != 0:
-        raise RuntimeError((out.stderr or out.stdout or "COM 写表失败").strip()[:400])
-    return path
-
-
-def _write_xls_via_xlwt(src, old_rows, rows, path, sheet_name="Sheet1"):
-    """退化路径：xlwt 整表重建。会让格式与未引用列走样，只在没有 Excel 时用。"""
-    import xlwt
-    book = xlwt.Workbook(encoding="utf-8")
-    # 工作表名必须与软件导出的一致（Screens / Others / Messages）——
-    # 写成默认的 Sheet1，导入时软件认不出表类型，会整份跳过。
-    sheet = book.add_sheet(sheet_name)
-    for r_i, r in enumerate(rows):
-        for c_i, v in enumerate(r):
-            # 空列也要写：原表有几列就保持几列，导入侧按列对语言
-            sheet.write(r_i, c_i, "" if v is None else str(v))
-    if os.path.exists(path):
-        os.remove(path)
-    book.save(path)
-    return path
-
-
-def cmd_apply(args):
-    os.makedirs(args.outdir, exist_ok=True)
-    work = os.path.join(args.outdir, "_work")
-    mapping = load_dict(args.dict)
-    if not mapping:
-        raise SystemExit(f"映射文件没有内容: {args.dict}")
-
-    log = []
-    applied = 0
-    missing = 0
-    for f in args.files:
-        p = parse(f, work)
-        rows = [list(r) for r in p["rows"]]
-        for idx in range(p["body_start"], len(rows)):
-            src, dst = rows[idx][0], rows[idx][1]
-            if not src.strip():
-                continue
-            wrapped, core = core_of(src)
-            # $_ignore 行照样要填：它是「工程里已有多个不同译文」被导出挡掉的，
-            # 留着 $_ignore 等于这段文字在表里仍是源语言，导回去屏上也不变。
-            if core in mapping:
-                new = mapping[core]
-                rows[idx][1] = f"<text>{new}</text>" if wrapped else new
-                applied += 1
-            else:
-                # 没给译文的行：源是中文、目标也是中文/空 → 记下来
-                _, core_b = core_of(dst)
-                if CJK_RE.search(core) and (not core_b.strip() or CJK_RE.search(core_b)):
-                    missing += 1
-                    log.append(f"{os.path.basename(f)} 行{idx+1} 未给译文: {core!r}")
-
-        stem = os.path.splitext(os.path.basename(f))[0]
-        if f.lower().endswith(".xls"):
-            # 源是 .xls 就出 .xls：WinCC flexible 的导入对话框按扩展名过滤，给 xlsx 会选不到
-            out_path = os.path.join(args.outdir, stem + ".xls")
-            write_xls(f, p["rows"], rows, out_path, p["sheet"])
-            # 写完回读核对：行数与第 1 列必须与源表一致，否则整表会错位
-            back = parse(out_path, work)
-            if len(back["rows"]) != len(p["rows"]):
-                raise SystemExit(f"写出的 {stem}.xls 行数 {len(back['rows'])} ≠ 源表 {len(p['rows'])}")
-            for i in range(len(p["rows"])):
-                if back["rows"][i][0] != p["rows"][i][0]:
-                    raise SystemExit(f"写出的 {stem}.xls 行{i+1} 键列与源表不一致")
-        else:
-            out_path = os.path.join(args.outdir, stem + ".xlsx")
-            write_book(rows, out_path, p["sheet"])
-        log.append(f"产出: {out_path}")
-
-    print(f"写回 {applied} 条；未给译文 {missing} 条。")
-    if log:
-        print("\n".join(log))
-    return 0
-
-
-def cmd_check(args):
-    work = os.path.join(args.outdir or tempfile.gettempdir(), "_work")
-    src = parse(args.src, work)
-    out = parse(args.out, work)
-    problems = []
-
-    if len(src["rows"]) != len(out["rows"]):
-        problems.append(f"行数不一致：原 {len(src['rows'])} / 产 {len(out['rows'])}")
-    n = min(len(src["rows"]), len(out["rows"]))
-    for i in range(n):
-        a1 = src["rows"][i][0]
-        a2 = out["rows"][i][0]
-        if a1 != a2:
-            problems.append(f"行{i+1} 键列（第 1 列）被改动：{a1!r} → {a2!r}")
-
-    items = scan(out)
-    for it in items:
-        if it["type"] == "UNTRANSLATED":
-            problems.append(f"行{it['row']} 未译/{it['note']}：{it['src'][:40]!r}")
-        elif it["type"] == "IGNORE":
-            # 填了才算译完；还留着 $_ignore 就是目标列仍是源语言
-            problems.append(f"行{it['row']} 仍是 $_ignore（未填译文）：{it['src'][:40]!r}")
-        elif it["type"] == "CODE_EXTRA":
-            problems.append(f"行{it['row']} 译文多出内容（疑串位，源无中文）：{it['src'][:30]!r} → {it['dst'][:40]!r}")
-    bad = conflicts(items)
-    for core, m in bad.items():
-        problems.append("一源多译：" + repr(core) + " → " + "；".join(map(repr, m.keys())))
-
-    if problems:
-        print(f"✗ 自检发现 {len(problems)} 处问题：")
-        for x in problems:
-            print("  - " + x)
-        return 1
-    print(f"✓ 自检通过：{n} 行，键列未动，无未译、无 $_ignore 残留、无串位、无一源多译。")
-    return 0
-
-
-def main(argv=None):
-    ap = argparse.ArgumentParser(description="WinCC flexible 项目文本导出表多语言处理")
-    sub = ap.add_subparsers(dest="cmd", required=True)
-
-    s = sub.add_parser("scan", help="分型出清单")
-    s.add_argument("--files", nargs="+", required=True)
-    s.add_argument("--outdir", required=True)
-    s.set_defaults(func=cmd_scan)
-
-    a = sub.add_parser("apply", help="按映射写回")
-    a.add_argument("--files", nargs="+", required=True)
-    a.add_argument("--dict", required=True)
-    a.add_argument("--outdir", required=True)
-    a.set_defaults(func=cmd_apply)
-
-    c = sub.add_parser("check", help="自检产物")
-    c.add_argument("--src", required=True)
-    c.add_argument("--out", required=True)
-    c.add_argument("--outdir", default="")
-    c.set_defaults(func=cmd_check)
-
-    args = ap.parse_args(argv)
-    return args.func(args)
-
-
-if __name__ == "__main__":
-    sys.exit(main())

package/invariant.js

@@ -1,10 +0,0 @@
-const PACKAGE_NAME = 'dsh-plc-decode'
-
-/** Cordis companion plugin name. */
-export const name = 'dsh-plc-decode-invariant'
-/** Service required before the companion can reserve package ownership. */
-export const inject = ['invariants']
-
-const install = () => {}
-
-export const apply = (ctx) => Promise.resolve(ctx.invariants.register(PACKAGE_NAME, install))