@deepwhale-cn/dsh-doc-batch 0.0.0-stage → 0.1.4

package/lib/xlsx_tool.py

@@ -0,0 +1,554 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+dsh-xlsx-batch 的真实处理逻辑（openpyxl）
+
+由 bin/cli.mjs 通过 spawnSync 调用：
+    python3 lib/xlsx_tool.py <action> '<json payload>'
+约定：
+  - stdout 只输出一行 JSON（最后一行），人类可读文案由 Node 侧负责
+  - stderr 输出异常栈（调试用）
+  - 退出码：0 成功 / 1 运行失败 / 2 参数或输入问题 / 3 缺依赖
+  - 只读输入文件，所有产物写进 --out 目录（默认 <输入目录>/_输出）
+  - 读取用 data_only=True：公式取 Excel 缓存的计算结果；输出为值（不是公式）
+"""
+
+import csv
+import json
+import os
+import re
+import sys
+import tempfile
+import traceback
+from pathlib import Path
+
+try:
+    from openpyxl import Workbook, load_workbook
+    from openpyxl.styles import Font
+except ImportError as exc:  # 缺依赖必须明确提示安装命令，不能静默
+    print(json.dumps({
+        "ok": False,
+        "code": 3,
+        "message": "缺少 Python 依赖：%s" % getattr(exc, "name", exc),
+        "hint": "请先安装：\n  python3 -m pip install openpyxl\n"
+                "如果 python3 也没有：\n  macOS: brew install python3\n  Ubuntu: sudo apt install python3",
+    }, ensure_ascii=False))
+    sys.exit(3)
+
+XLSX_EXTS = {".xlsx", ".xlsm"}
+SKIP_DIRS = {"_输出", "node_modules", ".git", "__pycache__", ".venv", "venv"}
+SOURCE_FILE = "来源文件"
+SOURCE_SHEET = "来源工作表"
+NUM_RE = re.compile(r"^-?\d{1,3}(?:,\d{3})*(?:\.\d+)?$|^-?\d+(?:\.\d+)?$")
+
+
+# ── 错误 ──────────────────────────────────────────────────────────
+
+class ToolError(Exception):
+    def __init__(self, message, hint="", code=2):
+        super().__init__(message)
+        self.message = message
+        self.hint = hint
+        self.code = code
+
+
+def emit(obj):
+    print(json.dumps(obj, ensure_ascii=False))
+
+
+# ── 路径与文件收集 ────────────────────────────────────────────────
+
+def natural_key(text):
+    return [int(t) if t.isdigit() else t.lower() for t in re.split(r"(\d+)", text)]
+
+
+def resolve_out(payload, input_path):
+    out = payload.get("out")
+    if out:
+        return Path(out).expanduser().resolve()
+    base = Path(input_path).expanduser().resolve()
+    if not base.is_dir():
+        base = base.parent
+    return base / "_输出"
+
+
+def collect_xlsx(input_path, out_dir=None):
+    if not input_path:
+        raise ToolError("缺少 --input", "例：merge --input ./报表目录 --out ./汇总")
+    root = Path(input_path).expanduser().resolve()
+    if not root.exists():
+        raise ToolError(
+            "输入不存在：%s" % input_path,
+            "请检查路径是否正确；传目录会递归处理其中的 .xlsx 文件。",
+        )
+    skip_out = set()
+    if out_dir:
+        try:
+            skip_out = {str(out_dir.resolve())}
+        except OSError:
+            skip_out = set()
+    if root.is_file():
+        if root.suffix.lower() not in XLSX_EXTS:
+            raise ToolError(
+                "输入不是 Excel 文件：%s" % root.name,
+                "本插件只处理 .xlsx / .xlsm；老格式 .xls 请先另存为 .xlsx。",
+            )
+        return [root]
+    files = []
+    for dirpath, dirnames, filenames in os.walk(root):
+        dirnames[:] = sorted(
+            [d for d in dirnames if d not in SKIP_DIRS and not d.startswith("~$")],
+            key=natural_key,
+        )
+        if str(Path(dirpath).resolve()) in skip_out:
+            dirnames[:] = []
+            continue
+        for name in filenames:
+            if name.startswith("~$"):
+                continue  # Excel 打开时的临时文件
+            p = Path(dirpath) / name
+            if p.suffix.lower() in XLSX_EXTS:
+                files.append(p)
+    files.sort(key=lambda p: natural_key(str(p.relative_to(root))))
+    if not files:
+        raise ToolError(
+            "目录里没有 .xlsx / .xlsm 文件：%s" % root,
+            "确认目录选对了；.xls 老格式请先另存为 .xlsx。",
+        )
+    return files
+
+
+def unique_path(directory, filename):
+    directory = Path(directory)
+    target = directory / filename
+    if not target.exists():
+        return target
+    stem, suffix = target.stem, target.suffix
+    n = 1
+    while (directory / ("%s_%d%s" % (stem, n, suffix))).exists():
+        n += 1
+    return directory / ("%s_%d%s" % (stem, n, suffix))
+
+
+def safe_name(text, default="sheet"):
+    name = re.sub(r'[\\/:*?"<>|\[\]]', "_", str(text)).strip()
+    return name or default
+
+
+# ── 单元格判断与表格读取 ──────────────────────────────────────────
+
+def is_empty(value):
+    return value is None or (isinstance(value, str) and value.strip() == "")
+
+
+def trim_text(value):
+    if isinstance(value, str):
+        value = value.strip()
+        return value if value else None
+    return value
+
+
+def to_number(value):
+    """数值列判定：数字原样；纯数字字符串（可带千分位）转数字；其余不算。"""
+    if isinstance(value, bool):
+        return None
+    if isinstance(value, (int, float)):
+        return value
+    if isinstance(value, str):
+        text = value.strip()
+        if NUM_RE.fullmatch(text):
+            try:
+                return float(text.replace(",", ""))
+            except ValueError:
+                return None
+    return None
+
+
+def trim_trailing(rows):
+    while rows and all(is_empty(v) for v in rows[-1]):
+        rows.pop()
+    width = 0
+    for row in rows:
+        for i, value in enumerate(row):
+            if not is_empty(value):
+                width = max(width, i + 1)
+    return [(row + [None] * width)[:width] for row in rows]
+
+
+def read_sheet_rows(ws):
+    rows = [list(row) for row in ws.iter_rows(values_only=True)]
+    return trim_trailing(rows)
+
+
+def split_header(rows):
+    """第一条非空行当表头，其余当数据。"""
+    for idx, row in enumerate(rows):
+        if any(not is_empty(v) for v in row):
+            return row, rows[idx + 1:]
+    return [], []
+
+
+def header_names(header):
+    names, used = [], {}
+    for i, value in enumerate(header):
+        base = str(value).strip() if not is_empty(value) else "列%d" % (i + 1)
+        count = used.get(base, 0) + 1
+        used[base] = count
+        names.append(base if count == 1 else "%s#%d" % (base, count))
+    return names
+
+
+# ── 各动作 ────────────────────────────────────────────────────────
+
+def action_merge(payload):
+    files = collect_xlsx(payload.get("input"))
+    out = resolve_out(payload, payload.get("input"))
+    out.mkdir(parents=True, exist_ok=True)
+
+    sheets = []  # (file, sheet_name, names, data_rows)
+    union, seen = [], set()
+    for src in files:
+        wb = load_workbook(str(src), data_only=True)
+        for ws in wb.worksheets:
+            rows = read_sheet_rows(ws)
+            header, data = split_header(rows)
+            if not header:
+                continue
+            names = header_names(header)
+            data = [r for r in data if any(not is_empty(v) for v in r)]
+            sheets.append((src, ws.title, names, data))
+            for name in names:
+                if name not in seen:
+                    seen.add(name)
+                    union.append(name)
+
+    header_row = [SOURCE_FILE, SOURCE_SHEET] + union
+    wb_out = Workbook()
+    ws_out = wb_out.active
+    ws_out.title = "合并"
+    ws_out.append(header_row)
+    for cell in ws_out[1]:
+        cell.font = Font(bold=True)
+
+    total_rows = 0
+    file_results = []
+    for src, sheet_name, names, data in sheets:
+        for row in data:
+            mapping = dict(zip(names, row))
+            ws_out.append([src.name, sheet_name] + [mapping.get(name) for name in union])
+            total_rows += 1
+        file_results.append({
+            "input": str(src),
+            "input_name": src.name,
+            "sheet": sheet_name,
+            "rows": len(data),
+        })
+
+    ws_out.freeze_panes = "A2"
+    if ws_out.max_row >= 2:
+        ws_out.auto_filter.ref = ws_out.dimensions
+    target = unique_path(out, "merged.xlsx")
+    wb_out.save(str(target))
+
+    return {
+        "action": "merge",
+        "input": str(Path(payload["input"]).expanduser().resolve()),
+        "out": str(out),
+        "files": file_results,
+        "output": str(target),
+        "output_name": target.name,
+        "rows": total_rows,
+        "sheets": len(sheets),
+        "summary": "合并 %d 个文件 / %d 个工作表，写入 %d 行数据" % (len(files), len(sheets), total_rows),
+        "artifacts": [str(target)],
+    }
+
+
+def normalize_rows(rows):
+    """trim 文本 + 统计被 trim 的单元格数。"""
+    trimmed, out = 0, []
+    for row in rows:
+        new_row = []
+        for value in row:
+            if isinstance(value, str):
+                stripped = value.strip()
+                if stripped != value:
+                    trimmed += 1
+                new_row.append(stripped if stripped else None)
+            else:
+                new_row.append(value)
+        out.append(new_row)
+    return out, trimmed
+
+
+def action_clean(payload):
+    files = collect_xlsx(payload.get("input"))
+    out = resolve_out(payload, payload.get("input"))
+    out.mkdir(parents=True, exist_ok=True)
+    results = []
+    for src in files:
+        wb = load_workbook(str(src), data_only=True)
+        wb_out = Workbook()
+        wb_out.remove(wb_out.active)
+        total_in = total_empty = total_dup = total_trim = total_out = 0
+        for ws in wb.worksheets:
+            rows = read_sheet_rows(ws)
+            header, data = split_header(rows)
+            header, trim_h = normalize_rows([header] if header else [])
+            header = header[0] if header else []
+            data, trim_d = normalize_rows(data)
+            total_trim += trim_h + trim_d
+
+            seen, kept = set(), []
+            empty = dup = 0
+            for row in data:
+                if all(is_empty(v) for v in row):
+                    empty += 1
+                    continue
+                key = tuple(row)
+                if key in seen:
+                    dup += 1
+                    continue
+                seen.add(key)
+                kept.append(row)
+
+            total_in += len(data)
+            total_empty += empty
+            total_dup += dup
+            total_out += len(kept)
+
+            new_ws = wb_out.create_sheet(title=safe_name(ws.title)[:31])
+            if header:
+                new_ws.append(header)
+                for cell in new_ws[1]:
+                    cell.font = Font(bold=True)
+            for row in kept:
+                new_ws.append(row)
+            new_ws.freeze_panes = "A2"
+
+        target = unique_path(out, src.stem + ".xlsx")
+        wb_out.save(str(target))
+        results.append({
+            "input": str(src),
+            "input_name": src.name,
+            "output": str(target),
+            "output_name": target.name,
+            "rows_in": total_in,
+            "rows_out": total_out,
+            "empty_removed": total_empty,
+            "duplicates_removed": total_dup,
+            "cells_trimmed": total_trim,
+        })
+    return {
+        "action": "clean",
+        "input": str(Path(payload["input"]).expanduser().resolve()),
+        "out": str(out),
+        "files": results,
+        "summary": "清洗 %d 个文件：删空行 %d、去重 %d、trim %d 个单元格" % (
+            len(files),
+            sum(r["empty_removed"] for r in results),
+            sum(r["duplicates_removed"] for r in results),
+            sum(r["cells_trimmed"] for r in results),
+        ),
+        "artifacts": [r["output"] for r in results],
+    }
+
+
+def action_summary(payload):
+    files = collect_xlsx(payload.get("input"))
+    out = resolve_out(payload, payload.get("input"))
+    out.mkdir(parents=True, exist_ok=True)
+    wb_out = Workbook()
+    ws_out = wb_out.active
+    ws_out.title = "汇总"
+    ws_out.append(["文件", "工作表", "列名", "非空计数", "数值个数", "求和"])
+    for cell in ws_out[1]:
+        cell.font = Font(bold=True)
+    report_rows = 0
+    file_results = []
+
+    for src in files:
+        wb = load_workbook(str(src), data_only=True)
+        sheet_count = 0
+        for ws in wb.worksheets:
+            rows = read_sheet_rows(ws)
+            header, data = split_header(rows)
+            if not header:
+                continue
+            names = header_names(header)
+            sheet_count += 1
+            width = len(header)
+            for col in range(width):
+                values = [row[col] for row in data if col < len(row)]
+                values = [v for v in values if not is_empty(v)]
+                numbers = [n for n in (to_number(v) for v in values) if n is not None]
+                ws_out.append([
+                    src.name,
+                    ws.title,
+                    names[col],
+                    len(values),
+                    len(numbers),
+                    sum(numbers) if numbers else None,
+                ])
+                report_rows += 1
+        file_results.append({
+            "input": str(src),
+            "input_name": src.name,
+            "sheets": sheet_count,
+        })
+
+    ws_out.freeze_panes = "A2"
+    if ws_out.max_row >= 2:
+        ws_out.auto_filter.ref = ws_out.dimensions
+    for letter, width in (("A", 26), ("B", 22), ("C", 24), ("D", 12), ("E", 12), ("F", 14)):
+        ws_out.column_dimensions[letter].width = width
+    target = unique_path(out, "summary.xlsx")
+    wb_out.save(str(target))
+
+    return {
+        "action": "summary",
+        "input": str(Path(payload["input"]).expanduser().resolve()),
+        "out": str(out),
+        "files": file_results,
+        "output": str(target),
+        "output_name": target.name,
+        "rows": report_rows,
+        "summary": "汇总 %d 个文件，输出 %d 行列级统计" % (len(files), report_rows),
+        "artifacts": [str(target)],
+    }
+
+
+def action_csv(payload):
+    files = collect_xlsx(payload.get("input"))
+    out = resolve_out(payload, payload.get("input"))
+    out.mkdir(parents=True, exist_ok=True)
+    results = []
+    for src in files:
+        wb = load_workbook(str(src), data_only=True)
+        for ws in wb.worksheets:
+            rows = read_sheet_rows(ws)
+            filename = "%s__%s.csv" % (safe_name(src.stem), safe_name(ws.title))
+            target = unique_path(out, filename)
+            with target.open("w", encoding="utf-8-sig", newline="") as fh:
+                writer = csv.writer(fh)
+                for row in rows:
+                    writer.writerow(["" if v is None else v for v in row])
+            results.append({
+                "input": str(src),
+                "input_name": src.name,
+                "sheet": ws.title,
+                "output": str(target),
+                "output_name": target.name,
+                "rows": len(rows),
+            })
+    return {
+        "action": "csv",
+        "input": str(Path(payload["input"]).expanduser().resolve()),
+        "out": str(out),
+        "files": results,
+        "summary": "导出 %d 个 CSV（UTF-8 带 BOM，Excel 可直接打开）" % len(results),
+        "artifacts": [r["output"] for r in results],
+    }
+
+
+# ── demo ─────────────────────────────────────────────────────────
+
+def _demo_book(path, sheets):
+    wb = Workbook()
+    wb.remove(wb.active)
+    for name, rows in sheets:
+        ws = wb.create_sheet(title=name)
+        for row in rows:
+            ws.append(row)
+    wb.save(str(path))
+
+
+def action_demo(payload):
+    root = Path(tempfile.mkdtemp(prefix="dsh-xlsx-batch-demo-"))
+    inputs = root / "样例"
+    inputs.mkdir(parents=True, exist_ok=True)
+    _demo_book(inputs / "销售A.xlsx", [
+        ("销售", [
+            ["姓名", "部门", "金额"],
+            ["张三", " 销售部 ", "100"],
+            ["李四", "销售部", "200"],
+            [None, None, None],
+            ["张三", " 销售部 ", "100"],
+        ]),
+        ("库存", [["品名", "数量"], ["A", 5], ["B", 7]]),
+    ])
+    _demo_book(inputs / "销售B.xlsx", [
+        ("销售", [["部门", "姓名", "金额"], ["市场部", "王五", "300"]]),
+        ("备注", [["说明"], ["样例数据"]]),
+    ])
+    out = root / "输出"
+    steps, artifacts = [], []
+
+    def run(name, fn):
+        result = fn({"input": str(inputs), "out": str(out / name)})
+        steps.append({"step": name, "action": result.get("action"),
+                      "summary": result.get("summary", ""), "out": result.get("out")})
+        artifacts.extend(result.get("artifacts", []))
+
+    run("合并", action_merge)
+    run("清洗", action_clean)
+    run("汇总", action_summary)
+    run("CSV", action_csv)
+
+    return {
+        "action": "demo",
+        "demoRoot": str(root),
+        "steps": steps,
+        "summary": "演示完成：4 个动作全部跑通，产物 %d 个" % len(artifacts),
+        "artifacts": artifacts,
+    }
+
+
+ACTIONS = {
+    "merge": action_merge,
+    "clean": action_clean,
+    "summary": action_summary,
+    "csv": action_csv,
+    "demo": action_demo,
+}
+
+
+def main(argv):
+    if len(argv) < 3:
+        emit({"ok": False, "code": 2,
+              "message": "内部调用参数不足：需要 <action> 和 JSON payload",
+              "hint": "请通过 bin/cli.mjs 调用本脚本。"})
+        return 2
+    action, raw = argv[1], argv[2]
+    try:
+        payload = json.loads(raw) if raw else {}
+    except json.JSONDecodeError as exc:
+        emit({"ok": False, "code": 2, "message": "参数 JSON 解析失败：%s" % exc, "hint": ""})
+        return 2
+    fn = ACTIONS.get(action)
+    if not fn:
+        emit({"ok": False, "code": 2, "message": "未知动作：%s" % action,
+              "hint": "可用动作：%s" % " / ".join(k for k in ACTIONS if k != "demo")})
+        return 2
+    try:
+        result = fn(payload)
+        result.setdefault("ok", True)
+        emit(result)
+        return 0
+    except ToolError as exc:
+        emit({"ok": False, "code": exc.code, "message": exc.message, "hint": exc.hint})
+        return exc.code
+    except PermissionError as exc:
+        emit({"ok": False, "code": 1,
+              "message": "没有权限读写文件：%s" % exc,
+              "hint": "检查目标文件是否被 Excel 占用、或输出目录是否有写权限。"})
+        return 1
+    except Exception as exc:  # noqa: BLE001 —— 兜底，保证永远输出 JSON
+        traceback.print_exc(file=sys.stderr)
+        emit({"ok": False, "code": 1, "message": "处理失败：%s" % exc,
+              "hint": "文件可能已损坏、被加密或格式不是标准 xlsx；可先用 Excel 另存一次再试。"})
+        return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main(sys.argv))

package/package.json

@@ -1,6 +1,67 @@
 {
   "name": "@deepwhale-cn/dsh-doc-batch",
-  "version": "0.0.0-stage",
-  "stub": true,
-  "description": "Temporary package placeholder for staged publishing"
-}
+  "version": "0.1.4",
+  "description": "深鲸文档批量处理：Word / Excel / PDF 一次处理整目录（替换、提取、合并、拆分、压缩、水印、导出）",
+  "type": "module",
+  "license": "MIT",
+  "keywords": [
+    "dsh",
+    "dsh-plugin",
+    "deepseek-harness",
+    "skill",
+    "tool",
+    "doc-batch",
+    "word",
+    "excel",
+    "pdf"
+  ],
+  "files": [
+    "lib",
+    "bin",
+    "skill",
+    "cordis.patch.yml",
+    "README.md",
+    "LICENSE",
+    "assets"
+  ],
+  "main": "./lib/index.js",
+  "exports": {
+    ".": "./lib/index.js",
+    "./client": "./lib/client.js",
+    "./package.json": "./package.json"
+  },
+  "bin": {
+    "dsh-doc-batch": "./bin/cli.mjs"
+  },
+  "engines": {
+    "dsh": ">=0.1.0-rc.6 <0.3.0",
+    "node": ">=18"
+  },
+  "dsh": {
+    "bundle": {
+      "patch": "./cordis.patch.yml"
+    },
+    "client": {
+      "inject": [
+        "@deepseek-ai/dsh-client-runtime"
+      ],
+      "platform": "web",
+      "immediately": true
+    }
+  },
+  "dependencies": {
+    "@deepwhale-cn/dsh-guard-core": "^0.5.1"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/dsh-skill": "*"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/feely0208/dsh-doc-batch.git"
+  },
+  "homepage": "https://www.deepwhale.org.cn/skills.html",
+  "bugs": {
+    "url": "https://github.com/feely0208/dsh-doc-batch/issues"
+  },
+  "author": "温州深鲸智能科技有限公司 <contact@deepwhale.org.cn>"
+}