Sync third-party and MCP marketplace plugins
Constraint: Public skills are published only by explicit administrator action unless they are tracked third-party market sources. Confidence: high Scope-risk: narrow Directive: Keep private/internal skills out of the public marketplace and preserve normal incremental market Git history. Tested: Marketplace validation passed.
This commit is contained in:
KeyInfo Bot committed
1 parent
be7ce9bda2
commit
36c089fee9
298 files changed
+32935
-11681
No files matched your search
@@ -0,0 +1,52 @@
|
||||
"""Unpack or repack an OOXML document without shell-specific ZIP commands."""
|
||||
|
||||
import argparse
|
||||
import os
|
||||
import tempfile
|
||||
import zipfile
|
||||
from pathlib import Path, PurePosixPath
|
||||
|
||||
|
||||
def unpack(source, destination):
|
||||
root = Path(destination).resolve()
|
||||
if root.exists() and any(root.iterdir()):
|
||||
raise ValueError("Unpack destination must be empty")
|
||||
with zipfile.ZipFile(source) as archive:
|
||||
for info in archive.infolist():
|
||||
name = PurePosixPath(info.filename.replace("\\", "/"))
|
||||
if name.is_absolute() or ".." in name.parts or any(":" in part for part in name.parts):
|
||||
raise ValueError(f"Unsafe ZIP entry: {info.filename}")
|
||||
if (info.external_attr >> 16) & 0o170000 == 0o120000:
|
||||
raise ValueError(f"ZIP symlink is not supported: {info.filename}")
|
||||
archive.extractall(root)
|
||||
|
||||
|
||||
def pack(source, destination):
|
||||
root = Path(source).resolve()
|
||||
output = Path(destination).resolve()
|
||||
if output.is_relative_to(root):
|
||||
raise ValueError("Output must be outside the unpacked directory")
|
||||
if not (root / "[Content_Types].xml").is_file():
|
||||
raise ValueError("Source must be the package root containing [Content_Types].xml")
|
||||
output.parent.mkdir(parents=True, exist_ok=True)
|
||||
fd, temporary = tempfile.mkstemp(suffix=output.suffix, dir=output.parent)
|
||||
os.close(fd)
|
||||
try:
|
||||
with zipfile.ZipFile(temporary, "w", zipfile.ZIP_DEFLATED) as archive:
|
||||
for file in sorted(root.rglob("*")):
|
||||
if file.is_symlink():
|
||||
raise ValueError(f"Package symlink is not supported: {file}")
|
||||
if file.is_file():
|
||||
archive.write(file, file.relative_to(root).as_posix())
|
||||
os.replace(temporary, output)
|
||||
finally:
|
||||
Path(temporary).unlink(missing_ok=True)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("action", choices=("unpack", "pack"))
|
||||
parser.add_argument("input")
|
||||
parser.add_argument("output")
|
||||
args = parser.parse_args()
|
||||
{"unpack": unpack, "pack": pack}[args.action](args.input, args.output)
|
||||
@@ -0,0 +1,41 @@
|
||||
# 简体中文与 Windows 环境
|
||||
|
||||
以下命令中的 `$skill` 是当前已安装 Skill 的真实目录；不要硬编码开发者机器路径。脚本使用 Python 3.10+。先检查现有 Python/Node 环境，缺少哪个依赖再安装哪个。
|
||||
|
||||
```powershell
|
||||
$env:PYTHONUTF8 = "1"
|
||||
[Console]::OutputEncoding = [System.Text.UTF8Encoding]::new($false)
|
||||
python --version
|
||||
node --version
|
||||
```
|
||||
|
||||
若只有 `py`，把命令中的 `python` 替换为 `py -3`。PowerShell 调用带空格的可执行文件路径使用 `& "路径"`；Python 脚本路径作为独立参数传入，不拼接 shell 命令。中文路径、空格路径均需引号。
|
||||
|
||||
需要独立依赖环境时，在任务工作目录创建 `.venv`，使用 `& ".\.venv\Scripts\python.exe" -m pip install <缺失包>`。无需改变系统执行策略或激活脚本。Node 库安装到任务目录，不默认全局安装。不要假设 Codex、Claude 或容器已预装所有依赖。
|
||||
|
||||
## 字体与文本
|
||||
|
||||
- 用户未指定语言时，中文用户的说明、标题、图例和交付文案使用简体中文；引文、产品名、代码和原文专有名词保持原样。
|
||||
- 新文件优先使用本机可用的微软雅黑、等线、宋体或思源黑体。先查字体；声明字体名不等于字体已安装。Word 配置 `w:rFonts/@w:eastAsia` 与 `zh-CN`，PPT 配置主题字体和文本框字体，Excel 使用 `Microsoft YaHei` 等实际存在的字体。
|
||||
- PDF 优先嵌入有再分发许可的中文字体；ReportLab 简单中文文档可用 `UnicodeCIDFont('STSong-Light')`，交付前仍检查目标阅读器效果。不要用 Helvetica 绘制中文，也不要把方块字当正常输出。
|
||||
- CSV 面向 Windows Excel 时可用 `utf-8-sig`；先保留已有文件编码，不要自动改写所有输入。XML、JSON 和新建脚本使用 UTF-8。
|
||||
- 标点、单位、日期、人民币格式、身份证号和前导零按用户语义保留；不要默认套用美式日期或把长编号转为数字。
|
||||
|
||||
## 转换与预览
|
||||
|
||||
Office 文档转 PDF 使用随附的跨平台 `scripts/office/soffice.py`。它按 `SOFFICE_PATH`、PATH、Windows Program Files、macOS 应用目录寻找 LibreOffice，并为转换创建独立临时 profile。
|
||||
|
||||
```powershell
|
||||
# 仅在自动检测不到时设置；不要求添加系统 PATH
|
||||
$env:SOFFICE_PATH = "C:\Program Files\LibreOffice\program\soffice.com"
|
||||
python "$skill\scripts\office\soffice.py" --headless --convert-to pdf --outdir ".\preview" ".\中文 文档.docx"
|
||||
python "$skill\scripts\render_pdf.py" ".\preview\中文 文档.pdf" ".\preview\pages"
|
||||
```
|
||||
|
||||
按实际需要安装 LibreOffice、`pymupdf`、`defusedxml`、`lxml` 等缺失依赖。PDF 的 `scripts/convert_pdf_to_images.py` 与其他 Skill 的 `scripts/render_pdf.py` 均使用 PyMuPDF，不需要 Poppler。`pptx` 上游缩略图脚本仍需 Poppler；无 Poppler 时走 PDF 渲染流程。
|
||||
|
||||
使用 Python `scripts/office_package.py unpack <输入> <空目录>` 和 `pack <包根目录> <输出>` 处理 Office ZIP，无需 `unzip`、`zip`、`rm` 或 Bash。输出必须在解包目录之外。
|
||||
|
||||
## 验证与交付
|
||||
|
||||
生成或编辑后保存到新文件；先检查内容、公式或 OOXML 结构，再渲染并查看页面，检查中文缺字、表格越界、文本遮挡和分页。转换失败或字体缺失时说明具体原因，不把“文件保存成功”当成渲染通过。保留输入，除非用户明确要求覆盖。
|
||||
@@ -0,0 +1,29 @@
|
||||
"""Render PDF pages without Poppler; requires PyMuPDF."""
|
||||
|
||||
import argparse
|
||||
from pathlib import Path
|
||||
|
||||
import pymupdf
|
||||
|
||||
|
||||
def convert(pdf_path, output_dir, max_dim=1600):
|
||||
if max_dim <= 0:
|
||||
raise ValueError("max_dim must be positive")
|
||||
output = Path(output_dir)
|
||||
output.mkdir(parents=True, exist_ok=True)
|
||||
with pymupdf.open(pdf_path) as document:
|
||||
for index, page in enumerate(document):
|
||||
scale = min(max_dim / page.rect.width, max_dim / page.rect.height)
|
||||
pixmap = page.get_pixmap(matrix=pymupdf.Matrix(scale, scale), alpha=False)
|
||||
target = output / f"page_{index + 1}.png"
|
||||
pixmap.save(str(target))
|
||||
print(target)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("input")
|
||||
parser.add_argument("output")
|
||||
parser.add_argument("--max-dim", type=int, default=1600)
|
||||
args = parser.parse_args()
|
||||
convert(args.input, args.output, args.max_dim)
|
||||
@@ -0,0 +1,61 @@
|
||||
"""Portable LibreOffice runner for Windows, macOS and Linux."""
|
||||
|
||||
import os
|
||||
import platform
|
||||
import shutil
|
||||
import subprocess
|
||||
import sys
|
||||
import tempfile
|
||||
from pathlib import Path
|
||||
|
||||
|
||||
def resolve_soffice_binary():
|
||||
explicit = os.environ.get("SOFFICE_PATH")
|
||||
if explicit:
|
||||
candidate = Path(explicit).expanduser()
|
||||
if candidate.is_file():
|
||||
return str(candidate)
|
||||
raise FileNotFoundError(f"SOFFICE_PATH does not exist: {candidate}")
|
||||
names = ("soffice.com", "soffice.exe", "soffice") if platform.system() == "Windows" else ("soffice",)
|
||||
for name in names:
|
||||
binary = shutil.which(name)
|
||||
if binary:
|
||||
return binary
|
||||
candidates = [Path("/Applications/LibreOffice.app/Contents/MacOS/soffice")]
|
||||
if platform.system() == "Windows":
|
||||
candidates = [
|
||||
Path(root) / "LibreOffice/program" / name
|
||||
for root in (os.environ.get("ProgramFiles"), os.environ.get("ProgramFiles(x86)")) if root
|
||||
for name in ("soffice.com", "soffice.exe")
|
||||
]
|
||||
for candidate in candidates:
|
||||
if candidate.is_file():
|
||||
return str(candidate)
|
||||
raise FileNotFoundError("LibreOffice not found. Install it or set SOFFICE_PATH to its executable.")
|
||||
|
||||
|
||||
def get_soffice_env():
|
||||
env = os.environ.copy()
|
||||
env["PYTHONUTF8"] = "1"
|
||||
if platform.system() == "Linux":
|
||||
env.setdefault("SAL_USE_VCLPLUGIN", "svp")
|
||||
return env
|
||||
|
||||
|
||||
def run_soffice(args, **kwargs):
|
||||
binary = resolve_soffice_binary()
|
||||
kwargs.setdefault("env", get_soffice_env())
|
||||
kwargs.setdefault("timeout", 120)
|
||||
if any(str(arg).startswith("-env:UserInstallation=") for arg in args):
|
||||
return subprocess.run([binary, *args], **kwargs)
|
||||
with tempfile.TemporaryDirectory(prefix="skill-lo-", ignore_cleanup_errors=True) as profile:
|
||||
profile_arg = f"-env:UserInstallation={Path(profile).resolve().as_uri()}"
|
||||
return subprocess.run([binary, profile_arg, *args], **kwargs)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
try:
|
||||
raise SystemExit(run_soffice(sys.argv[1:]).returncode)
|
||||
except (FileNotFoundError, subprocess.TimeoutExpired) as exc:
|
||||
print(str(exc), file=sys.stderr)
|
||||
raise SystemExit(1)
|
||||
@@ -0,0 +1,38 @@
|
||||
---
|
||||
name: docx
|
||||
description: 创建、读取、编辑和检查 Word 文档或模板（.docx/.dotx），包括中文排版、表格、批注和修订；适用于简体中文与 Windows PowerShell 环境。
|
||||
license: Proprietary. LICENSE.txt has complete terms
|
||||
---
|
||||
|
||||
# Word 文档
|
||||
|
||||
开始操作前读取 [references/platform.md](references/platform.md)，确认 Python、Node、中文字体及需要的转换工具。命令中的脚本路径均相对本 Skill。
|
||||
|
||||
| 任务 | 选择 |
|
||||
| --- | --- |
|
||||
| 新建文档 | `docx`（Node）或 `python-docx`，按所需版式选择 |
|
||||
| 提取正文或表格 | `python-docx`；需要查看修订时用 Pandoc 或检查 OOXML |
|
||||
| 编辑复杂模板、批注或修订 | 保留 OOXML 结构，使用随附脚本 |
|
||||
| .doc 旧文件 | LibreOffice 先转换为 .docx |
|
||||
|
||||
## 中文版式
|
||||
|
||||
未指定页面时，新文档采用 A4；已有文档沿用其页面和样式。正文中文字体设置 East Asian 字体，语言为 `zh-CN`，标题使用真正的标题样式。保持全角标点、编号、金额和原文引文；中文表格不要依赖空格对齐。
|
||||
|
||||
`docx` 创建时注意：表格同时指定列宽和单元格宽度；`ShadingType.CLEAR` 避免黑底；列表用 numbering 配置；`ImageRun` 指定类型；分页符放进 Paragraph。每个逻辑段落独立创建，不用换行模拟段落。需要 API 细节时按需读取 [upstream-guide.md](upstream-guide.md)。
|
||||
|
||||
## 编辑与审阅
|
||||
|
||||
先确认要清稿、批注还是保留修订；不要因文档类型自动强制红线。原有修订与批注保持不变，除非用户要求接受、拒绝或删除。
|
||||
|
||||
```powershell
|
||||
python "$skill\scripts\office_package.py" unpack ".\输入.docx" ".\unpacked"
|
||||
python "$skill\scripts\merge_runs.py" ".\unpacked"
|
||||
# 按需编辑 word/document.xml，保持命名空间与节点顺序
|
||||
python "$skill\scripts\office_package.py" pack ".\unpacked" ".\输出.docx"
|
||||
python "$skill\scripts\office\validate.py" ".\输出.docx" --original ".\输入.docx"
|
||||
```
|
||||
|
||||
`scripts/comment.py` 支持为文档添加批注；输出的定位节点需放到对应正文位置才能显示。`scripts/accept_changes.py` 仅在用户要求接受全部修订时使用，输出新文件；超时、文件未保存或仍有修订均算失败。
|
||||
|
||||
最终转换为 PDF，用 `scripts/render_pdf.py` 生成预览并查看中文字体、表格、页眉页脚、分页和页码。保留可编辑 Word 文件与用户需要的最终副本。
|
||||
@@ -0,0 +1,75 @@
|
||||
"""Accept tracked changes into a new DOCX using an isolated LibreOffice profile."""
|
||||
|
||||
import argparse
|
||||
import shutil
|
||||
import subprocess
|
||||
import tempfile
|
||||
import zipfile
|
||||
from pathlib import Path
|
||||
|
||||
from defusedxml import ElementTree
|
||||
from office.soffice import run_soffice
|
||||
|
||||
MACRO = '''<?xml version="1.0" encoding="UTF-8"?>
|
||||
<!DOCTYPE script:module PUBLIC "-//OpenOffice.org//DTD OfficeDocument 1.0//EN" "module.dtd">
|
||||
<script:module xmlns:script="http://openoffice.org/2000/script" script:name="Module1" script:language="StarBasic">
|
||||
Sub AcceptAllTrackedChanges()
|
||||
Dim dispatcher As Object
|
||||
dispatcher = createUnoService("com.sun.star.frame.DispatchHelper")
|
||||
dispatcher.executeDispatch(ThisComponent.CurrentController.Frame, ".uno:AcceptAllTrackedChanges", "", 0, Array())
|
||||
ThisComponent.store()
|
||||
ThisComponent.close(True)
|
||||
End Sub
|
||||
</script:module>'''
|
||||
WORD_NS = "{http://schemas.openxmlformats.org/wordprocessingml/2006/main}"
|
||||
|
||||
|
||||
def accept_changes(input_file, output_file, timeout=120):
|
||||
source = Path(input_file).resolve()
|
||||
output = Path(output_file).resolve()
|
||||
if source == output or not source.is_file() or source.suffix.lower() != ".docx":
|
||||
return None, "Input must be an existing DOCX and output must be a different file"
|
||||
output.parent.mkdir(parents=True, exist_ok=True)
|
||||
try:
|
||||
with tempfile.TemporaryDirectory(prefix="docx-accept-", ignore_cleanup_errors=True) as temporary:
|
||||
root = Path(temporary).resolve()
|
||||
profile = root / "profile"
|
||||
profile_arg = f"-env:UserInstallation={profile.as_uri()}"
|
||||
run_soffice([profile_arg, "--headless", "--terminate_after_init"], check=True, capture_output=True, timeout=timeout)
|
||||
macro_dir = profile / "user/basic/Standard"
|
||||
macro_dir.mkdir(parents=True, exist_ok=True)
|
||||
(macro_dir / "Module1.xba").write_text(MACRO, encoding="utf-8")
|
||||
candidate = root / "accepted.docx"
|
||||
shutil.copy2(source, candidate)
|
||||
before = candidate.stat().st_mtime_ns
|
||||
run_soffice([
|
||||
profile_arg, "--headless", "--norestore",
|
||||
"vnd.sun.star.script:Standard.Module1.AcceptAllTrackedChanges?language=Basic&location=application",
|
||||
str(candidate),
|
||||
], check=True, capture_output=True, timeout=timeout)
|
||||
if candidate.stat().st_mtime_ns == before:
|
||||
return None, "LibreOffice did not save the document"
|
||||
with zipfile.ZipFile(candidate) as archive:
|
||||
for name in archive.namelist():
|
||||
if not name.startswith("word/") or not name.endswith(".xml"):
|
||||
continue
|
||||
document = ElementTree.fromstring(archive.read(name))
|
||||
if any(element.tag in {WORD_NS + "ins", WORD_NS + "del", WORD_NS + "moveFrom", WORD_NS + "moveTo"}
|
||||
or (element.tag.startswith(WORD_NS) and element.tag.endswith("Change"))
|
||||
for element in document.iter()):
|
||||
return None, f"Tracked changes remain in {name}"
|
||||
shutil.copy2(candidate, output)
|
||||
return None, f"Accepted tracked changes: {output}"
|
||||
except (OSError, subprocess.SubprocessError, zipfile.BadZipFile) as exc:
|
||||
return None, f"Failed to accept tracked changes: {exc}"
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("input")
|
||||
parser.add_argument("output")
|
||||
parser.add_argument("--timeout", type=int, default=120)
|
||||
args = parser.parse_args()
|
||||
_, message = accept_changes(args.input, args.output, args.timeout)
|
||||
print(message)
|
||||
raise SystemExit(0 if message.startswith("Accepted tracked changes:") else 1)
|
||||
@@ -0,0 +1,35 @@
|
||||
---
|
||||
name: pdf
|
||||
description: 读取、创建、合并、拆分、旋转、OCR 或填写 PDF 表单，处理简体中文字体与页面预览；适用于 Windows PowerShell 和跨平台终端。
|
||||
license: Proprietary. LICENSE.txt has complete terms
|
||||
---
|
||||
|
||||
# PDF 文档
|
||||
|
||||
先读取 [references/platform.md](references/platform.md)，确认中文字体、Python 与实际需要的依赖。PDF 不等于扫描图片，先判断页面是否有可提取文字。
|
||||
|
||||
| 任务 | 工具 |
|
||||
| --- | --- |
|
||||
| 文本、页面预览、坐标检查 | PyMuPDF |
|
||||
| 合并、拆分、旋转、元数据、加密 | pypdf |
|
||||
| 表格提取 | pdfplumber，提取后核对表头、合并单元格与换行 |
|
||||
| 新建中文 PDF | ReportLab 或可用的 Office/HTML 转 PDF 工具 |
|
||||
| 扫描页 OCR | 本机可用的 OCR；Tesseract 中文需 `chi_sim` 数据 |
|
||||
|
||||
## 中文与排版
|
||||
|
||||
新文档默认简体中文、A4；已有版式或用户指定版式优先。中文必须使用实际可用的中文字体。优先嵌入授权允许的字体；ReportLab 简单文档可注册 `UnicodeCIDFont('STSong-Light')`。复制文本不乱码、渲染无缺字都需检查。
|
||||
|
||||
OCR 只针对扫描页或提取失败的页面；中文可用 `chi_sim+eng`，未装语言包时报告缺失。OCR 结果需核对数字、单位、姓名与表格关系，不凭猜测补全。
|
||||
|
||||
## 表单
|
||||
|
||||
涉及填写时按需读取 [forms.md](forms.md)，先判断是否有 AcroForm 字段。可填表单使用随附 `extract_form_field_info.py`、`fill_fillable_fields.py`；非可填表单按页面坐标叠加，并检查原内容与填写区域。不要把签名图片视作数字签名。
|
||||
|
||||
## 预览与检查
|
||||
|
||||
```powershell
|
||||
python "$skill\scripts\convert_pdf_to_images.py" ".\输入 中文.pdf" ".\preview"
|
||||
```
|
||||
|
||||
此脚本使用 PyMuPDF，无需 Poppler，逐页渲染以避免一次加载整个大文档。合并或拆分后核对页面数量和顺序；填写后检查可见内容、字段值与页面边界。更多提取与转换方案见 [reference.md](reference.md)，旧平台专用命令以本地环境指导为准。
|
||||
@@ -0,0 +1,34 @@
|
||||
---
|
||||
name: pptx
|
||||
description: 创建、读取、编辑和检查 PowerPoint 演示文稿或模板（.pptx/.potx），保留模板与可编辑图表，适配简体中文字体和 Windows PowerShell。
|
||||
license: Proprietary. LICENSE.txt has complete terms
|
||||
---
|
||||
|
||||
# PowerPoint 演示文稿
|
||||
|
||||
先读取 [references/platform.md](references/platform.md)，确认字体、Node/Python 与需要的转换工具。先看用户模板或参考稿，再决定布局、配色与内容密度。
|
||||
|
||||
| 任务 | 选择 |
|
||||
| --- | --- |
|
||||
| 新建幻灯片 | `pptxgenjs`；简单任务也可使用 `python-pptx` |
|
||||
| 提取文本 | `python-pptx` 或 markitdown；检查备注与隐藏页 |
|
||||
| 复杂模板、母版或原生图表编辑 | OOXML 加随附脚本 |
|
||||
| .ppt 旧格式 | LibreOffice 先转 .pptx |
|
||||
|
||||
## 中文与画布
|
||||
|
||||
新建稿未指定比例时用 16:9，设置画布后再添加元素。用户模板比例和主题优先。使用本机可用的微软雅黑、等线或思源黑体，同时设置主题字体与正文文本字体；简体中文语言标记使用 `zh-CN`。长中文标题主动换行，表格与数据标签保留足够空间，不能只缩小字号隐藏溢出。
|
||||
|
||||
`pptxgenjs` 颜色使用六位 HEX，不带 `#`；每次添加元素使用独立选项对象。文本框 margin 与字体测量要一致；原生图表、表格和文本保持可编辑。坐标单位为英寸，必须落在实际画布内。库限制、组合图表和模板 OOXML 细节按需读取 [upstream-guide.md](upstream-guide.md)。
|
||||
|
||||
## 模板编辑
|
||||
|
||||
先完成复制、删除或重排幻灯片，再改正文。使用 `scripts/add_slide.py` 注册复制页，最终页序确定后运行 `scripts/clean.py` 清理失去引用的部分。不要只复制 slide XML；关系、媒体、母版和内容类型必须一起处理。用 `scripts/office_package.py` 解包和重打包，避免依赖 Bash ZIP 命令。
|
||||
|
||||
```powershell
|
||||
python "$skill\scripts\office\validate.py" ".\输出.pptx" --original ".\模板.pptx"
|
||||
python "$skill\scripts\office\soffice.py" --headless --convert-to pdf --outdir ".\preview" ".\输出.pptx"
|
||||
python "$skill\scripts\render_pdf.py" ".\preview\输出.pdf" ".\preview\pages"
|
||||
```
|
||||
|
||||
没有模板时省略 `--original`。每页检查文字溢出、缺字、对齐、遮挡、图表标签、图片分辨率和页序。需要模板缩略图且 Poppler 可用时运行 `scripts/thumbnail.py` 并指定独立输出前缀；否则用上面的免 Poppler 逐页预览。
|
||||
@@ -0,0 +1,35 @@
|
||||
---
|
||||
name: xlsx
|
||||
description: 创建、读取、清理、编辑和检查 Excel 工作簿或 CSV/TSV，处理公式、图表、中文表头及 Windows Excel 编码，保留现有格式与宏。
|
||||
license: Proprietary. LICENSE.txt has complete terms
|
||||
---
|
||||
|
||||
# Excel 工作簿
|
||||
|
||||
先读取 [references/platform.md](references/platform.md)，确认 Python、中文字体及是否需要公式计算引擎。复杂工作簿先检查公式、外部链接、宏、图表和现有样式，再选择处理方法。
|
||||
|
||||
- `openpyxl` 用于普通 .xlsx 单元格、公式和样式；`.xlsm` 读取时用 `keep_vba=True`，保留扩展名。不保证复杂 Excel 功能无损，必须检查实际需要保留的部分。
|
||||
- `pandas` 适合表格数据清理；不要用整表重写处理需要保留样式、公式或图表的原文件。
|
||||
- `.xls` 旧格式先使用兼容工具读取或转换；不要直接交给 openpyxl。
|
||||
|
||||
## 简体中文与数据类型
|
||||
|
||||
用户未指定语言时用简体中文表头、图例与说明。新文件使用可用的中文字体，表头、行高与列宽按中文实际长度设置；已有工作簿沿用其约定。身份证号、订单号、手机号和前导零存为文本，日期存真实日期值并设置明确格式，金额单位按用户要求标注。
|
||||
|
||||
Windows Excel 打开的新 CSV 可写 `utf-8-sig`；读取旧 CSV 先判断编码和分隔符。CSV 导出不保留多表、公式和样式，转换前明确所需交付格式。
|
||||
|
||||
## 公式与缓存
|
||||
|
||||
按用户语义写公式，验证引用范围与预期结果；不要用硬编码数值代替可重算公式。跨表引用含空格或中文标点时正确引用工作表名。合并区域只写左上角。
|
||||
|
||||
openpyxl 不计算公式。检查时分别用 `data_only=False` 读取公式、`data_only=True` 读取缓存；不能保存 `data_only=True` 的工作簿，否则公式会被数值覆盖。
|
||||
|
||||
```powershell
|
||||
python "$skill\scripts\recalc.py" ".\输出 中文.xlsx" 120
|
||||
```
|
||||
|
||||
重算会改写指定工作簿，先保留输入。脚本使用独立 LibreOffice profile，自动识别 Windows 的安装路径。检查 JSON 的 `status`、`total_errors` 和具体单元格；`errors_found` 即使进程退出码为零也未通过。含 `error` 的结果表示没有完成重算。
|
||||
|
||||
有外部链接、宏、动态数组或 Excel 专用函数时，优先使用目标 Excel 环境验证，不能盲目用 LibreOffice 往返保存。兼容公式、缓存丢失和链接风险按需读取 [upstream-guide.md](upstream-guide.md)。无法运行计算引擎时说明“尚未重算”，不要声称公式已验证。
|
||||
|
||||
交付前核对数值、公式、中文显示、筛选、冻结窗格、打印区域和图表。用新输出文件保留输入；抽查代表行、合计和异常情况，避免只检查文件能够打开。
|
||||
@@ -0,0 +1,89 @@
|
||||
"""Recalculate an XLSX through a temporary LibreOffice import/export, without macros."""
|
||||
|
||||
import argparse
|
||||
import json
|
||||
import os
|
||||
import shutil
|
||||
import subprocess
|
||||
import tempfile
|
||||
import zipfile
|
||||
from pathlib import Path
|
||||
|
||||
from openpyxl import load_workbook
|
||||
from office.soffice import run_soffice
|
||||
|
||||
|
||||
def inspect_workbook(path):
|
||||
formulas = load_workbook(path, data_only=False)
|
||||
values = load_workbook(path, data_only=True)
|
||||
try:
|
||||
formula_cells = set()
|
||||
errors = {}
|
||||
for sheet in formulas:
|
||||
if not hasattr(sheet, "iter_rows"):
|
||||
continue
|
||||
for row in sheet.iter_rows():
|
||||
for cell in row:
|
||||
location = f"{sheet.title}!{cell.coordinate}"
|
||||
if cell.data_type == "f":
|
||||
formula_cells.add(location)
|
||||
cached = values[sheet.title][cell.coordinate]
|
||||
if cached.data_type == "e":
|
||||
errors.setdefault(str(cached.value), []).append(location)
|
||||
return formula_cells, errors
|
||||
finally:
|
||||
formulas.close()
|
||||
values.close()
|
||||
|
||||
|
||||
def recalc(filename, timeout=120):
|
||||
source = Path(filename).expanduser().resolve()
|
||||
if not source.is_file() or source.suffix.lower() != ".xlsx":
|
||||
return {"error": "Input must be an existing .xlsx file; use Excel for macro-enabled workbooks"}
|
||||
try:
|
||||
with zipfile.ZipFile(source) as archive:
|
||||
if any(name.startswith("xl/externalLinks/") for name in archive.namelist()):
|
||||
return {"error": "External workbook links require validation in Excel; original file was not modified"}
|
||||
original_formulas, _ = inspect_workbook(source)
|
||||
with tempfile.TemporaryDirectory(prefix="xlsx-recalc-", dir=source.parent, ignore_cleanup_errors=True) as temporary:
|
||||
root = Path(temporary)
|
||||
input_dir, output_dir = root / "input", root / "output"
|
||||
input_dir.mkdir()
|
||||
output_dir.mkdir()
|
||||
candidate = input_dir / source.name
|
||||
shutil.copy2(source, candidate)
|
||||
result = run_soffice([
|
||||
"--headless", "--norestore", "--convert-to", "xlsx:Calc MS Excel 2007 XML",
|
||||
"--outdir", str(output_dir), str(candidate),
|
||||
], capture_output=True, text=True, encoding="utf-8", errors="replace", timeout=timeout)
|
||||
recalculated = output_dir / source.name
|
||||
if result.returncode or not recalculated.is_file():
|
||||
return {"error": "LibreOffice did not export a recalculated workbook", "details": result.stderr or result.stdout}
|
||||
formula_cells, errors = inspect_workbook(recalculated)
|
||||
if original_formulas != formula_cells:
|
||||
return {"error": "Formula locations changed during conversion; original file was not modified"}
|
||||
error_count = sum(len(locations) for locations in errors.values())
|
||||
report = {
|
||||
"status": "errors_found" if error_count else "success",
|
||||
"total_formulas": len(formula_cells),
|
||||
"total_errors": error_count,
|
||||
"error_summary": {
|
||||
error: {"count": len(locations), "locations": locations[:100]}
|
||||
for error, locations in errors.items()
|
||||
},
|
||||
}
|
||||
if not errors:
|
||||
os.replace(recalculated, source)
|
||||
return report
|
||||
except (OSError, subprocess.SubprocessError, zipfile.BadZipFile, ValueError) as exc:
|
||||
return {"error": str(exc)}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("input")
|
||||
parser.add_argument("timeout", nargs="?", type=int, default=120)
|
||||
args = parser.parse_args()
|
||||
report = recalc(args.input, args.timeout)
|
||||
print(json.dumps(report, ensure_ascii=False, indent=2))
|
||||
raise SystemExit(1 if "error" in report else 0)
|
||||
Reference in new issue
Block a user