ZSCGR/open-kimi-ppt-skill
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: export_pptx.py

Browse Source 
Related to #4
This commit is contained in:
binaryify
2026-08-06 18:02:20 +08:00
parent 066cbc33a5
commit f77ee8dfcb
7 changed files with 204 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.npmignore
+3 -1
View File
@@ -1,8 +1,10 @@
# When .npmignore exists, npm ignores .gitignore for packing.
# Keep packaging exclusions here even if they already appear in .gitignore.
__pycache__/
**/__pycache__/
**/__pycache__/**
*.py[cod]
**/*.py[cod]
*$py.class
.pytest_cache/
CHANGELOG.md
+8
View File
@@ -4,6 +4,14 @@
本项目遵循 [语义化版本](https://semver.org/lang/zh-CN/)。
## [1.1.2] - 2026-08-06
### 修复
- 修复中文 Windows 下 `export_pptx.py` / `export_images.py` 导出卡住或 GBK 解码失败（stdout 改走临时文件 + UTF-8）
- 规避 agent-browser `--download-path` 导致 Chrome 静默取消下载：改为点击下载并轮询默认 Downloads
- 修复 npm 包误打入 `__pycache__/*.pyc`（`files` 仅列出脚本源文件）
## [1.1.1] - 2026-08-06
### 新增
CHANGELOG_EN.md
+8
View File
@@ -4,6 +4,14 @@
This project follows [Semantic Versioning](https://semver.org/).
## [1.1.2] - 2026-08-06
### Fixed
- Fix Chinese-locale Windows export hang / GBK decode errors in `export_pptx.py` / `export_images.py` (capture stdout via temp file + UTF-8)
- Work around agent-browser `--download-path` silently canceling Chrome downloads: click download and poll the default Downloads folder
- Stop shipping `__pycache__/*.pyc` in the npm package (list script sources explicitly in `files`)
## [1.1.1] - 2026-08-06
### Added
package.json
+4 -2
View File
@@ -1,6 +1,6 @@
{
"name": "open-kimi-ppt-skills",
"version": "1.1.1",
"version": "1.1.2",
"description": "Unofficial reverse-engineered Kimi Slides skill for AI coding agents with local PPTD editing and PPTX export.",
"license": "MIT",
"author": "binaryify",
@@ -35,7 +35,9 @@
"lib/",
"skills/open-kimi-ppt/SKILL.md",
"skills/open-kimi-ppt/reference/",
"skills/open-kimi-ppt/scripts/",
"skills/open-kimi-ppt/scripts/export_host.html",
"skills/open-kimi-ppt/scripts/export_images.py",
"skills/open-kimi-ppt/scripts/export_pptx.py",
"README.md",
"README_EN.md",
"theme.md",
skills/open-kimi-ppt/scripts/export_images.py
+18 -18
View File
@@ -16,7 +16,6 @@ import re
import shutil
import subprocess
import sys
import tempfile
import time
import uuid
import zipfile
@@ -28,12 +27,15 @@ from export_pptx import (
BrowserSession,
ExportError,
build_payload,
default_downloads_dir,
ensure_agent_browser,
find_download,
find_manifest,
log,
ref_by_name,
run_command,
serve,
temporary_directory,
wait_for_export_dialog,
)
@@ -51,11 +53,8 @@ def ensure_pillow() -> Tuple[Any, Any, Any]:
return Image, ImageDraw, ImageFont
except ImportError:
log("Pillow is required for stitching; installing pillow with pip --user")
process = subprocess.run(
process = run_command(
[sys.executable, "-m", "pip", "install", "--user", "pillow"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=300,
)
if process.returncode != 0:
@@ -187,11 +186,8 @@ def ensure_websocket() -> Any:
return websocket
except ImportError:
log("websocket-client is required for dialog automation; installing with pip --user")
process = subprocess.run(
process = run_command(
[sys.executable, "-m", "pip", "install", "--user", "websocket-client"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=300,
)
if process.returncode != 0:
@@ -307,7 +303,7 @@ def export_images(
image_cls, draw_cls, image_font = ensure_pillow()
log(f"manifest: {manifest}")
with tempfile.TemporaryDirectory(prefix="open-kimi-ppt-images-") as temp_name:
with temporary_directory(prefix="open-kimi-ppt-images-") as temp_name:
temp_dir = Path(temp_name)
download_dir = temp_dir / "downloads"
download_dir.mkdir()
@@ -318,6 +314,7 @@ def export_images(
server, thread, url = serve(temp_dir)
session = f"open-kimi-ppt-images-{os.getpid()}-{uuid.uuid4().hex[:8]}"
browser = BrowserSession(agent_browser, session, temp_dir, download_dir)
downloads = default_downloads_dir()
try:
log("opening the public Kimi slide editor")
browser.open(url)
@@ -338,17 +335,15 @@ def export_images(
select_image_format(browser)
dialog = wait_for_export_dialog(browser)
started_at = time.time() - 1.0
download_ref = ref_by_name(dialog, "下载", "button")
log("rendering page images in the browser")
result = browser.run(
["download", f"@{download_ref}", str(temp_dir / "browser-output.zip")],
timeout=300,
check=False,
)
if result.returncode != 0:
log("download capture reported a timeout; checking browser output files")
browser.run(["click", f"@{download_ref}"], timeout=300)
downloaded = find_download(
(download_dir, temp_dir), timeout=240, accept=is_image_zip
(downloads, download_dir, temp_dir),
timeout=240,
accept=is_image_zip,
since=started_at,
)
finally:
browser.close()
@@ -362,6 +357,11 @@ def export_images(
images = unzip_images(downloaded, output / "pages")
if keep_download:
shutil.copy2(downloaded, output / "browser-raw.zip")
try:
if downloaded.resolve().parent == downloads.resolve():
downloaded.unlink(missing_ok=True)
except OSError:
pass
overview = stitch_overview(
images, output / "overview.jpg", image_cls, draw_cls, image_font
)
skills/open-kimi-ppt/scripts/export_pptx.py
+101 -43
View File
@@ -63,16 +63,83 @@ def log(message: str) -> None:
print(f"[open-kimi-ppt] {message}", file=sys.stderr, flush=True)
def run_command(
command: Sequence[str],
*,
cwd: Optional[Path] = None,
env: Optional[Dict[str, str]] = None,
timeout: int = 90,
) -> subprocess.CompletedProcess[str]:
"""Capture merged stdout/stderr via a temp file.
On Windows, agent-browser's detached daemon can inherit a PIPE handle and
prevent EOF, deadlocking ``subprocess.run(stdout=PIPE)``. Decoding with the
system locale (GBK on zh-CN Windows) can also raise UnicodeDecodeError.
Writing to a UTF-8 file avoids both failures.
"""
handle, sink_path = tempfile.mkstemp(prefix="open-kimi-ppt-", suffix=".log")
os.close(handle)
sink = Path(sink_path)
output = ""
try:
with sink.open("w", encoding="utf-8", errors="replace") as out:
returncode = subprocess.call(
list(command),
cwd=str(cwd) if cwd is not None else None,
env=env,
stdout=out,
stderr=subprocess.STDOUT,
timeout=timeout,
)
output = sink.read_text(encoding="utf-8", errors="replace")
except subprocess.TimeoutExpired as exc:
try:
output = sink.read_text(encoding="utf-8", errors="replace")
except OSError:
output = ""
raise subprocess.TimeoutExpired(
cmd=list(command),
timeout=timeout,
output=output,
) from exc
finally:
try:
sink.unlink(missing_ok=True)
except OSError:
# WinError 32: daemon may still hold the log file handle.
pass
return subprocess.CompletedProcess(list(command), returncode, output, None)
def temporary_directory(prefix: str) -> Any:
# ignore_cleanup_errors avoids masking the real export error when a Windows
# browser daemon still holds files under the temp tree (Python 3.10+).
try:
return tempfile.TemporaryDirectory(prefix=prefix, ignore_cleanup_errors=True)
except TypeError:
return tempfile.TemporaryDirectory(prefix=prefix)
def default_downloads_dir() -> Path:
home = Path.home()
candidates: List[Path] = []
user_profile = os.environ.get("USERPROFILE")
if user_profile:
candidates.append(Path(user_profile) / "Downloads")
candidates.extend((home / "Downloads", home / "下载"))
for path in candidates:
if path.is_dir():
return path
return home / "Downloads"
def ensure_pyyaml() -> Any:
try:
import yaml
except ImportError:
log("PyYAML is required; installing pyyaml with pip --user")
process = subprocess.run(
process = run_command(
[sys.executable, "-m", "pip", "install", "--user", "pyyaml"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=300,
)
if process.returncode != 0:
@@ -103,13 +170,7 @@ def parse_node_version(output: str) -> Tuple[int, int, int]:
def read_agent_browser_version(executable: str) -> Tuple[int, int, int]:
process = subprocess.run(
[executable, "--version"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=30,
)
process = run_command([executable, "--version"], timeout=30)
if process.returncode != 0:
raise ExportError(f"agent-browser --version failed:\n{process.stdout[-2000:]}")
return parse_version(process.stdout)
@@ -120,13 +181,7 @@ def ensure_nodejs() -> str:
if not executable:
raise ExportError(f"Node.js is not installed or not on PATH. {NODE_INSTALL_HINT}")
process = subprocess.run(
[executable, "--version"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=30,
)
process = run_command([executable, "--version"], timeout=30)
if process.returncode != 0:
raise ExportError(f"node --version failed:\n{process.stdout[-2000:]}")
@@ -168,11 +223,8 @@ def ensure_agent_browser() -> str:
current = "not installed" if version is None else ".".join(map(str, version))
minimum = ".".join(map(str, MIN_AGENT_BROWSER_VERSION))
log(f"agent-browser {current} is below {minimum}; installing agent-browser@latest")
process = subprocess.run(
process = run_command(
[npm, "install", "-g", "agent-browser@latest"],
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=300,
)
if process.returncode != 0:
@@ -304,6 +356,8 @@ class BrowserSession:
self.executable = executable
self.session = session
self.cwd = cwd
# Kept as a search root fallback; not passed to agent-browser. On Windows,
# --download-path can be rewritten to a \\?\ path that cancels Chrome downloads.
self.download_dir = download_dir
self.env = os.environ.copy()
self.env.setdefault("AGENT_BROWSER_DEFAULT_TIMEOUT", "60000")
@@ -317,15 +371,7 @@ class BrowserSession:
check: bool = True,
) -> subprocess.CompletedProcess[str]:
command = [self.executable, "--session", self.session, *args]
process = subprocess.run(
command,
cwd=self.cwd,
env=self.env,
text=True,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
timeout=timeout,
)
process = run_command(command, cwd=self.cwd, env=self.env, timeout=timeout)
if check and process.returncode != 0:
raise ExportError(
f"agent-browser command failed ({process.returncode}): "
@@ -334,9 +380,9 @@ class BrowserSession:
return process
def open(self, url: str) -> None:
self.run(
["--download-path", str(self.download_dir), "open", url], timeout=90
)
# Avoid --download-path: agent-browser ≤0.33.2 + Chrome may cancel downloads
# when given a verbatim Windows path. Files land in the default Downloads folder.
self.run(["open", url], timeout=90)
def snapshot(self) -> Dict[str, Any]:
process = self.run(["snapshot", "-i", "-C", "--json"])
@@ -408,6 +454,8 @@ def find_download(
search_roots: Iterable[Path],
timeout: float = 150.0,
accept: Callable[[Path], bool] = is_pptx,
*,
since: Optional[float] = None,
) -> Path:
deadline = time.monotonic() + timeout
last_sizes: Dict[Path, int] = {}
@@ -420,7 +468,10 @@ def find_download(
candidates.extend(path for path in root.rglob("*") if path.is_file())
for path in sorted(candidates, key=lambda item: item.stat().st_mtime, reverse=True):
try:
size = path.stat().st_size
stat = path.stat()
size = stat.st_size
if since is not None and stat.st_mtime < since:
continue
except OSError:
continue
if size == last_sizes.get(path) and size > 0:
@@ -588,7 +639,7 @@ def export_pptx(
f"defaults: transition={transition}, embed_fonts={'on' if embed_fonts else 'off'}"
)
with tempfile.TemporaryDirectory(prefix="open-kimi-ppt-export-") as temp_name:
with temporary_directory(prefix="open-kimi-ppt-export-") as temp_name:
temp_dir = Path(temp_name)
download_dir = temp_dir / "downloads"
download_dir.mkdir()
@@ -599,6 +650,7 @@ def export_pptx(
server, thread, url = serve(temp_dir)
session = f"open-kimi-ppt-export-{os.getpid()}-{uuid.uuid4().hex[:8]}"
browser = BrowserSession(agent_browser, session, temp_dir, download_dir)
downloads = default_downloads_dir()
try:
log("opening the public Kimi slide editor")
browser.open(url)
@@ -627,16 +679,17 @@ def export_pptx(
elif embed_fonts:
log("warning: the official export dialog exposed no font switch")
# Plain click (not agent-browser `download`) so Chrome saves to the
# default Downloads folder; --download-path is broken on some Windows setups.
started_at = time.time() - 1.0
download_ref = ref_by_name(dialog, "下载", "button")
log("generating PPTX in the browser")
result = browser.run(
["download", f"@{download_ref}", str(temp_dir / "browser-output.pptx")],
timeout=180,
check=False,
browser.run(["click", f"@{download_ref}"], timeout=180)
downloaded = find_download(
(downloads, download_dir, temp_dir),
timeout=90,
since=started_at,
)
if result.returncode != 0:
log("download capture reported a timeout; checking browser output files")
downloaded = find_download((download_dir, temp_dir), timeout=90)
shutil.copy2(downloaded, output)
if keep_download:
debug_copy = output.with_name(f"{output.stem}.browser-raw.pptx")
@@ -645,6 +698,11 @@ def export_pptx(
f"raw debug output already exists (pass --force): {debug_copy}"
)
shutil.copy2(downloaded, debug_copy)
try:
if downloaded.resolve().parent == downloads.resolve():
downloaded.unlink(missing_ok=True)
except OSError:
pass
finally:
browser.close()
server.shutdown()
skills/open-kimi-ppt/tests/test_export_pptx.py
+62 -13
View File
@@ -1,6 +1,8 @@
#!/usr/bin/env python3
import importlib.util
import os
import tempfile
import time
import unittest
import zipfile
from pathlib import Path
@@ -19,9 +21,9 @@ class ExportPptxTests(unittest.TestCase):
self.assertEqual(MODULE.parse_version("agent-browser 0.33.2"), (0, 33, 2))
self.assertEqual(MODULE.parse_version("v1.4.0-beta.1"), (1, 4, 0))
@patch.object(MODULE.subprocess, "run")
@patch.object(MODULE, "run_command")
@patch.object(MODULE.shutil, "which")
def test_old_agent_browser_is_upgraded(self, which, run):
def test_old_agent_browser_is_upgraded(self, which, run_command):
which.side_effect = [
"/bin/node",
"/bin/npm",
@@ -29,38 +31,38 @@ class ExportPptxTests(unittest.TestCase):
"/bin/npm",
"/bin/agent-browser",
]
run.side_effect = [
run_command.side_effect = [
MODULE.subprocess.CompletedProcess([], 0, "v22.11.0\n"),
MODULE.subprocess.CompletedProcess([], 0, "agent-browser 0.17.1\n"),
MODULE.subprocess.CompletedProcess([], 0, "changed 1 package\n"),
MODULE.subprocess.CompletedProcess([], 0, "agent-browser 0.33.2\n"),
]
self.assertEqual(MODULE.ensure_agent_browser(), "/bin/agent-browser")
self.assertEqual(run.call_args_list[2].args[0], [
self.assertEqual(run_command.call_args_list[2].args[0], [
"/bin/npm", "install", "-g", "agent-browser@latest"
])
@patch.object(MODULE.subprocess, "run")
@patch.object(MODULE, "run_command")
@patch.object(MODULE.shutil, "which")
def test_missing_nodejs_raises_clear_error(self, which, run):
def test_missing_nodejs_raises_clear_error(self, which, run_command):
which.return_value = None
with self.assertRaisesRegex(MODULE.ExportError, "Node.js is not installed"):
MODULE.ensure_nodejs()
run.assert_not_called()
run_command.assert_not_called()
@patch.object(MODULE.subprocess, "run")
@patch.object(MODULE, "run_command")
@patch.object(MODULE.shutil, "which")
def test_old_nodejs_raises_clear_error(self, which, run):
def test_old_nodejs_raises_clear_error(self, which, run_command):
which.return_value = "/bin/node"
run.return_value = MODULE.subprocess.CompletedProcess([], 0, "v16.20.2\n")
run_command.return_value = MODULE.subprocess.CompletedProcess([], 0, "v16.20.2\n")
with self.assertRaisesRegex(MODULE.ExportError, "Node.js 18\\+ is required"):
MODULE.ensure_nodejs()
@patch.object(MODULE.subprocess, "run")
@patch.object(MODULE, "run_command")
@patch.object(MODULE.shutil, "which")
def test_missing_npm_raises_clear_error(self, which, run):
def test_missing_npm_raises_clear_error(self, which, run_command):
which.side_effect = ["/bin/node", None]
run.return_value = MODULE.subprocess.CompletedProcess([], 0, "v22.11.0\n")
run_command.return_value = MODULE.subprocess.CompletedProcess([], 0, "v22.11.0\n")
with self.assertRaisesRegex(MODULE.ExportError, "npm is not installed"):
MODULE.ensure_nodejs()
@@ -128,6 +130,53 @@ class ExportPptxTests(unittest.TestCase):
slide = archive.read("ppt/slides/slide1.xml")
self.assertIn(b"<p:fade/>", slide)
@patch.object(MODULE.subprocess, "call", return_value=0)
def test_run_command_captures_utf8_via_temp_file(self, call):
def write_sink(*_args, **kwargs):
kwargs["stdout"].write("agent-browser 0.33.2\n")
return 0
call.side_effect = write_sink
process = MODULE.run_command(["agent-browser", "--version"], timeout=5)
self.assertEqual(process.returncode, 0)
self.assertIn("0.33.2", process.stdout)
self.assertEqual(call.call_args.kwargs["stderr"], MODULE.subprocess.STDOUT)
def test_find_download_ignores_files_older_than_since(self):
with tempfile.TemporaryDirectory() as name:
root = Path(name)
old = root / "old.pptx"
new = root / "new.pptx"
for path in (old, new):
with zipfile.ZipFile(path, "w") as archive:
archive.writestr(
"[Content_Types].xml",
'<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">'
'<Override PartName="/ppt/presentation.xml" '
f'ContentType="{MODULE.PPTX_CONTENT_TYPE}"/></Types>',
)
archive.writestr("ppt/presentation.xml", "<p:presentation/>")
older = time.time() - 60
os.utime(old, (older, older))
since = time.time() - 5
found = MODULE.find_download([root], timeout=2.0, since=since)
self.assertEqual(found.resolve(), new.resolve())
def test_browser_open_does_not_pass_download_path(self):
session = MODULE.BrowserSession(
"/bin/agent-browser",
"test-session",
Path("."),
Path("/tmp/downloads"),
)
with patch.object(session, "run") as run:
session.open("http://127.0.0.1:9/export_host.html")
run.assert_called_once_with(
["open", "http://127.0.0.1:9/export_host.html"],
timeout=90,
)
if __name__ == "__main__":
unittest.main()