HTML+CSS to image renderer based on litehtml

These details have been verified by PyPI

Project links

GitHub Statistics

Maintainers

tyql688

These details have not been verified by PyPI

Project description

pylitehtml

HTML+CSS → PNG/JPEG 图像渲染器。轻量级，无需无头浏览器，线程安全。需要 Python ≥ 3.10。

效果展示

examples/ 下有若干独立可运行的示例，各渲染一张图（无浏览器、纯 pylitehtml，与 tests/ 分开）。

示例	说明	渲染结果
`showcase_html.py`	HTML/CSS 能力参考：排版、颜色/渐变、表格、代码、列表、进度条 + 真实 HTTPS 图片 + `data:` URI 图标
`showcase_css.py`	More CSS：conic/radial 渐变、边框样式、`position`、`::before`、`text-overflow:ellipsis`、`white-space:pre`、`vertical-align`
`showcase_markdown.py`	内置零依赖 Markdown 转换器的端到端输出
`showcase_jinja2.py`	Jinja2 数据驱动模板（循环 / 条件 / 过滤器 / autoescape，一张订单发票）
`markdown_full_converter.py`	用 markdown-it-py 接入完整 GFM（脚注 / 定义列表 / Pygments 语法高亮；需 `pip install markdown-it-py mdit-py-plugins pygments`）
`render_markdown_doc.py`	渲染一篇真实复杂文档（`markdown.md`：多表格 / 嵌套列表 / 多语言代码 / Mermaid 源码 / 数学 TeX / 行号引用代码块）

中英文混排共享同一基线（不会出现中文“飞起来”的错位）；文本测量为子像素精度；模板变量经 autoescape 安全转义；抓取失败的远程图片自动跳过，不影响整图渲染。

python examples/showcase_html.py      # 生成 assets/showcase_html.png（需联网加载真实图片）
python examples/showcase_css.py
python examples/showcase_markdown.py
python examples/showcase_jinja2.py

性能（对比 Playwright）

同一篇文档各渲染 30 次（macOS arm64，width=720；基准脚本 bench/bench_vs_playwright.py）：

指标	pylitehtml	Playwright（无头 Chromium）
单次渲染（warm，均值）	≈ 15.5 ms	≈ 21 ms
首屏 / 冷启动	≈ 0.2 s	≈ 0.25 s（OS 已缓存）/ 首次可达 ~3.4 s
额外依赖	无（仅系统 Cairo/Pango 等）	需下载 Chromium ~190 MB+，每次渲染有浏览器进程内存开销
JavaScript	❌ 不执行	✅ 可执行（KaTeX/Mermaid 等）

要点：稳态吞吐 pylitehtml 约快 1.4×；真正的差距在部署体积与冷启动 —— 无需 ~190MB 的 Chromium、无浏览器进程，特别适合容器 / Serverless / 边缩缩容场景。需要执行 JS（数学排版、Mermaid 绘图）时仍应选 Playwright。

pip install playwright && playwright install chromium   # 或: uv pip install --group bench
python bench/bench_vs_playwright.py 30

安装

pip install pylitehtml

快速上手

import pylitehtml

# 最简单：HTML 片段 → PNG（自动套用一套干净的默认样式）
png = pylitehtml.html_to_png("<h1>Hello</h1><p>世界</p>")
open("out.png", "wb").write(png)

# Markdown → PNG（零依赖内置转换器）
from pylitehtml.markdown import markdown_to_png
png = markdown_to_png("# 标题\n\n- 列表\n- [x] 任务")

# 需要精细控制时用底层接口：render / Renderer（不注入默认样式）
png = pylitehtml.render("<h1>Hello</h1>", width=800)
r = pylitehtml.Renderer(width=800)                       # 复用，字体只加载一次
jpg = r.render("<h1>Hello</h1>", fmt="jpeg", quality=90)

# 本地 HTML 文件（自动解析相对路径的 CSS/图片）+ Jinja2 模板
png = r.render_file("project/index.html")
png = r.render_file("project/order.html", title="订单", items=[...])

# 异步（不阻塞事件循环）
import asyncio
png = await asyncio.to_thread(r.render, "<h1>Hello</h1>")

高级封装：html_to_png / markdown_to_png

render() 是最底层接口（HTML 字符串 → 图片）。在它之上提供了两个高层便捷函数。

`html_to_png` — 带默认样式的 HTML 渲染（核心、零额外依赖）

from pylitehtml import html_to_png, html_to_image, wrap_html, DEFAULT_CSS

# 传入“片段”时，自动套用一套干净的 GitHub 风默认样式
png = html_to_png("<h1>标题</h1><p>正文 <code>code</code></p>", width=720)

# 传入完整文档（含 <html>/<body>）时，按原样渲染，不再注入默认样式
png = html_to_png("<html><body style='background:#000'>…</body></html>")

# HiDPI：scale 同时放大画布宽度与（默认样式的）根字号，em 布局等比放大
png = html_to_png(fragment, width=720, scale=2)

# 追加/覆盖样式；或拿原始 RGBA 像素
png = html_to_png(fragment, extra_css="body{background:#0d1117;color:#e6edf3}")
raw = html_to_image(fragment, fmt="raw")          # → RawResult(.data/.width/.height)
doc = wrap_html("<p>x</p>", css=DEFAULT_CSS)       # 仅生成完整 HTML 文档字符串

主要参数：width、scale（HiDPI 倍数，默认 1）、wrap（None 自动判断片段/整文档；True/False 强制）、 css / extra_css、fmt（"png"/"jpeg"）、quality、height、shrink_to_fit、locale、fonts、images。

`markdown_to_png` — Markdown → HTML → 图片（可选、默认零依赖）

“用 HTML 渲染 Markdown”：内置一个纯标准库的轻量 Markdown→HTML 转换器，再交给 html_to_png。 不引入 markdown-it-py、pygments 等任何第三方依赖。

from pylitehtml.markdown import markdown_to_png, markdown_to_html

png = markdown_to_png("# 标题\n\n- 列表\n- [x] 任务\n\n> 引用", width=720)  # scale 默认 2
html = markdown_to_html("**bold**")   # → '<p><strong>bold</strong></p>'

内置转换器覆盖：标题、粗/斜/***粗斜***/删除线/行内代码、反斜杠转义、链接/图片/裸 URL、行内 raw HTML、硬换行、围栏代码块（不高亮）、引用、有序（含 start）/无序/嵌套/任务列表、带对齐的 GFM 表格、分隔线。不支持：setext 标题、引用式链接、脚注、定义列表、缩进代码块。

需要完整 CommonMark/GFM（脚注、定义列表、语法高亮）时传入 converter= 接入 markdown-it-py，见 examples/markdown_full_converter.py：

from markdown_it import MarkdownIt
from mdit_py_plugins.footnote import footnote_plugin
from mdit_py_plugins.deflist import deflist_plugin
from pylitehtml.markdown import markdown_to_png

md = MarkdownIt("gfm-like").use(footnote_plugin).use(deflist_plugin)
png = markdown_to_png(text, converter=md.render)

不执行 JavaScript：数学公式（KaTeX）与 Mermaid 无法排版/绘制，需调用前自行预渲染或接入完整引擎。

支持 / 不支持

litehtml = CSS 2.1 + 部分 CSS3。下表为实测结论：

能力	状态
PNG / JPEG / RAW 输出	✅
盒模型、边框（solid/dashed/dotted/double）、`border-radius`、背景色	✅
`linear-gradient` / `conic-gradient`（`radial-gradient` ⚠️ 仅颜色对）	✅
字体、`color`、`text-align`、`line-height`、`text-shadow`、`text-transform`	✅
`text-decoration`、`<mark>`/`<sub>`/`<sup>`/`<kbd>`、`white-space:pre/nowrap`	✅
列表（嵌套 / `list-style-type` / 任务列表）、定义列表、表格（对齐 / `:nth-child`）	✅
`float`、`inline-block`、`position:relative/absolute`、`text-overflow:ellipsis`、`::before/::after`	✅
图片 `<img>`：`data:`(PNG/JPEG/WebP/SVG) / 本地 / HTTP·HTTPS、`@import` CSS	✅
中文 / 多语言（内置 Noto Sans + SC，中英共享基线、子像素测量）	✅
多线程并发、`asyncio.to_thread`	✅
Flexbox（简单行 / 列）	⚠️ 部分
`background-image:url()`、`box-shadow`、`transform`、`opacity`、`letter-spacing`	❌
CSS Grid、`var()`、`@font-face`、动画 / 过渡、`filter`、JavaScript	❌

图片抓取失败会自动跳过（不报错）；官方 wheel（Linux manylinux_2_34 / macOS / Windows）均含 OpenSSL ≥ 3.0，支持 https。

API

# 高层（推荐）：自动默认样式、片段/整文档自动判别、HiDPI scale
html_to_png(html, *, width=720, scale=1, wrap=None, css=DEFAULT_CSS, extra_css="",
            fmt="png", quality=85, height=0, shrink_to_fit=True,
            locale="en-US", fonts=None, images=None) -> bytes
html_to_image(...)   # 同参数，fmt 可取 "raw" → RawResult(.data/.width/.height，RGBA 行主序)

# 底层：不注入默认样式
Renderer(width, *, locale="en-US", dpi=96.0, device_height=600, fonts=None, images=None)
Renderer.render(html, *, fmt="png", quality=85, height=0, shrink_to_fit=True)
Renderer.render_file(path, *, fmt="png", quality=85, height=0, shrink_to_fit=True, **template_data)
render(html, width, ...) / render_file(path, width, ...)   # 一次性便捷函数

FontConfig(default="Noto Sans", size=16, extra=[])          # 已内置 Noto Sans/SC + DejaVu Sans
ImageConfig(cache_mb=64.0, timeout_ms=5000, max_mb=10.0, allow_http=True)

fmt="raw" 返回 RGBA：Image.frombytes("RGBA",(raw.width,raw.height),raw.data) 或 np.frombuffer(...)。
scale 等比放大画布与（默认样式的）根字号；shrink_to_fit 把画布收窄到内容宽度。
内置字体使 sans-serif/serif/monospace 与中文在无系统字体时也能渲染；dpi 影响 pt 换算。

render_file 与 Jinja2

按标准网页组织目录，自动解析相对路径的 CSS/图片；传入关键字参数即作为 Jinja2 模板渲染（变量 / 循环 / 条件 / 过滤器 / include / extends）：

png = r.render_file("project/index.html")                                       # 纯文件
png = r.render_file("order.html", title="订单", items=[{"name": "苹果", "price": 5}])   # 模板

autoescape 默认开启：.html 模板里的 {{ 变量 }} 会转义 </&/"，不可信数据也不会注入。 本地文件安全：file:// 与根路径资源仅在 render_file() 下加载；直接 render() 字符串时忽略。

线程 / 异步

先在单线程构造 Renderer，之后可多线程并发 render()（渲染期间释放 GIL）；异步用 await asyncio.to_thread(r.render, html)。带 extra_fonts 的构造会改写进程级字体配置，勿在渲染进行中于其它线程构造新 Renderer。

从源码构建

Ubuntu / Debian

sudo apt-get install -y libcairo2-dev libpango1.0-dev libfontconfig1-dev \
  libwebp-dev libjpeg-turbo8-dev libssl-dev cmake ninja-build pkg-config
pip install -e ".[dev]" --no-build-isolation

macOS

brew install cairo pango fontconfig webp jpeg-turbo openssl@3 cmake ninja
CC=/usr/bin/clang CXX=/usr/bin/clang++ pip install -e ".[dev]" --no-build-isolation

运行测试 / 代码检查

pytest tests/ -v
ruff check .        # lint
ruff format .       # 格式化（CI 用 `ruff format --check` 校验）

基于

litehtml · Cairo · Pango · FontConfig · pybind11

MIT License

Project details

These details have been verified by PyPI

Project links

GitHub Statistics

Maintainers

tyql688

These details have not been verified by PyPI

Release history Release notifications | RSS feed

This version

0.2.1

May 29, 2026

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pylitehtml-0.2.1.tar.gz (16.9 MB view details)

Uploaded May 29, 2026 Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

pylitehtml-0.2.1-cp314-cp314-win_amd64.whl (17.2 MB view details)

Uploaded May 29, 2026 CPython 3.14Windows x86-64

pylitehtml-0.2.1-cp314-cp314-manylinux_2_34_x86_64.whl (17.0 MB view details)

Uploaded May 29, 2026 CPython 3.14manylinux: glibc 2.34+ x86-64

pylitehtml-0.2.1-cp314-cp314-manylinux_2_34_aarch64.whl (16.5 MB view details)

Uploaded May 29, 2026 CPython 3.14manylinux: glibc 2.34+ ARM64

pylitehtml-0.2.1-cp314-cp314-macosx_15_0_arm64.whl (16.1 MB view details)

Uploaded May 29, 2026 CPython 3.14macOS 15.0+ ARM64

pylitehtml-0.2.1-cp313-cp313-win_amd64.whl (17.2 MB view details)

Uploaded May 29, 2026 CPython 3.13Windows x86-64

pylitehtml-0.2.1-cp313-cp313-manylinux_2_34_x86_64.whl (17.0 MB view details)

Uploaded May 29, 2026 CPython 3.13manylinux: glibc 2.34+ x86-64

pylitehtml-0.2.1-cp313-cp313-manylinux_2_34_aarch64.whl (16.5 MB view details)

Uploaded May 29, 2026 CPython 3.13manylinux: glibc 2.34+ ARM64

pylitehtml-0.2.1-cp313-cp313-macosx_15_0_arm64.whl (16.1 MB view details)

Uploaded May 29, 2026 CPython 3.13macOS 15.0+ ARM64

pylitehtml-0.2.1-cp312-cp312-win_amd64.whl (17.2 MB view details)

Uploaded May 29, 2026 CPython 3.12Windows x86-64

pylitehtml-0.2.1-cp312-cp312-manylinux_2_34_x86_64.whl (17.0 MB view details)

Uploaded May 29, 2026 CPython 3.12manylinux: glibc 2.34+ x86-64

pylitehtml-0.2.1-cp312-cp312-manylinux_2_34_aarch64.whl (16.5 MB view details)

Uploaded May 29, 2026 CPython 3.12manylinux: glibc 2.34+ ARM64

pylitehtml-0.2.1-cp312-cp312-macosx_15_0_arm64.whl (16.1 MB view details)

Uploaded May 29, 2026 CPython 3.12macOS 15.0+ ARM64

pylitehtml-0.2.1-cp311-cp311-win_amd64.whl (17.2 MB view details)

Uploaded May 29, 2026 CPython 3.11Windows x86-64

pylitehtml-0.2.1-cp311-cp311-manylinux_2_34_x86_64.whl (17.0 MB view details)

Uploaded May 29, 2026 CPython 3.11manylinux: glibc 2.34+ x86-64

pylitehtml-0.2.1-cp311-cp311-manylinux_2_34_aarch64.whl (16.5 MB view details)

Uploaded May 29, 2026 CPython 3.11manylinux: glibc 2.34+ ARM64

pylitehtml-0.2.1-cp311-cp311-macosx_15_0_arm64.whl (16.1 MB view details)

Uploaded May 29, 2026 CPython 3.11macOS 15.0+ ARM64

pylitehtml-0.2.1-cp310-cp310-win_amd64.whl (17.2 MB view details)

Uploaded May 29, 2026 CPython 3.10Windows x86-64

pylitehtml-0.2.1-cp310-cp310-manylinux_2_34_x86_64.whl (17.0 MB view details)

Uploaded May 29, 2026 CPython 3.10manylinux: glibc 2.34+ x86-64

pylitehtml-0.2.1-cp310-cp310-manylinux_2_34_aarch64.whl (16.5 MB view details)

Uploaded May 29, 2026 CPython 3.10manylinux: glibc 2.34+ ARM64

pylitehtml-0.2.1-cp310-cp310-macosx_15_0_arm64.whl (16.1 MB view details)

Uploaded May 29, 2026 CPython 3.10macOS 15.0+ ARM64

File details

Details for the file pylitehtml-0.2.1.tar.gz.

File metadata

Download URL: pylitehtml-0.2.1.tar.gz
Upload date: May 29, 2026
Size: 16.9 MB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for pylitehtml-0.2.1.tar.gz
Algorithm	Hash digest
SHA256	`90d5b0fddff0613efa82024802cd485e604557998ba20e4cb782d34d4f348967`
MD5	`a299333b2ace0f637240676b1780160b`
BLAKE2b-256	`c0b926cf5973c928314a648f3e112c285a0444dac29a051dd5244828f7a7a6f0`

Algorithm	Hash digest
SHA256	`073fa49fad9dff8f04de42e73b82331170745c4e274b498541943ca447a6d7bd`
MD5	`fcb9de51f68fc3c25f70364bc11f7d28`
BLAKE2b-256	`a1820dfd1802021a8560ed5e124f06aff24e8aaebd16cf3f844e76af8cb71f7c`

Algorithm	Hash digest
SHA256	`df656db14572dd676e6a857c12633b7bc01625d7a00b59d8d987673d1f0e4bf6`
MD5	`ff01fadfe3998ff174e2081846041e4f`
BLAKE2b-256	`d759e1a13255bf0ca56fee2a9ea58dce0c75e31e38070f3d5f69483c3725810f`

Algorithm	Hash digest
SHA256	`9b9f91e91ed0ef2db95b7ea212716b8ccd1e1c6ced4428b720aa84cb64166223`
MD5	`ad2f3c5b995572fa651156875484ac33`
BLAKE2b-256	`aa356e2271e2c3adac5ee44f90548acbb6641ab72100090c2692dd484dacb94b`

Algorithm	Hash digest
SHA256	`6f428d30229344c3eff27848be69a19e255722b157a3bb74c12f45bd44cca2fa`
MD5	`cd0aeac6c02284bca71ab98b0dcab446`
BLAKE2b-256	`4919476cf8d758822f85b60d0bc6985dc3f93ac6ec71df48bdedb830eeac3b55`

Algorithm	Hash digest
SHA256	`10d7b17071116567ea727b21f48725dff564b1e4b5eb27d88432ee2b5072431c`
MD5	`0ad95c22a4360b097be34058a7022937`
BLAKE2b-256	`2b0256b2e03ab2e8ce0cd93a2a4da6fe55df8680a7668de4d37ca2892c16e7cf`

Algorithm	Hash digest
SHA256	`3e7473cbf753fb634b0206e8d299de36cd9d968e4f05312a2a3d4ee4be1fc4f9`
MD5	`3ff1b679584f0cfc9eab0d51eb5579c3`
BLAKE2b-256	`6d20fcbd84a2d0f5f76ad08b0280dc23689bfad1ff3275a3999e93ae893edcda`

Algorithm	Hash digest
SHA256	`af142d9d055785b2140f9172de79079cd4618992d00ea4637647a57b4ceb179e`
MD5	`38e9823b676b7c1d7329ed3811609844`
BLAKE2b-256	`834da6f24c1d3b9cd63d6e6639efc2430eb197e71dbb3cb18f7453a8fcf6132a`

Algorithm	Hash digest
SHA256	`bab58f80da7f77682eb7fbbbf666ac9cffdf3eacbd014e9954eb440777a570a9`
MD5	`4f4f987857ac3f34f6b7b9be6c840026`
BLAKE2b-256	`6c892982acff8bdc66d24ef9e25c4790fb3a93af617078073c82532f7cc4b8f1`

Algorithm	Hash digest
SHA256	`f0cb41133badd39d587e340c7e1c6a5120c36164630eb2653149f08dc4859922`
MD5	`42bcca439ba1db8224afe5896d415430`
BLAKE2b-256	`65a045916eb2eb2db3fbef6caddee75fc4c4c82e03108e1ac7dd43dcedde55c4`

Algorithm	Hash digest
SHA256	`3663fa1aed2857efcbd0e8d6b4d53fe053735a113a02fe69a61618c3bc69cdd8`
MD5	`23ef15609436c508eb8d926b2d65f765`
BLAKE2b-256	`b72072279fb60be1e4cc6f079ca73166196780ac91bc9e9181efd1f83ada3d47`

Algorithm	Hash digest
SHA256	`f8f0c337f8299a0f577118399a87e70c9df70b3a1b2b879e8ef7729d5bb70444`
MD5	`85390fb67aa706d6de983c6c06a1b4e7`
BLAKE2b-256	`7db04777de5db655f409a489dbbfeb287ff9257152d806cdf686dd17728446eb`

Algorithm	Hash digest
SHA256	`00dc276cf505094aa99f9c0ed42f5c61bedc856355df88f2a22129a158d96c1b`
MD5	`3947406ac3a95db0d86ce45aab32bff9`
BLAKE2b-256	`c1ef23817a638f02179485bb7331718dbe9e2494b049ea49e32c0a7908636510`

Algorithm	Hash digest
SHA256	`2eba6767c1f04be4c18f2e4de8a73ae87483023658c16a08ab454cfc05439c0a`
MD5	`1268027979eb64f2c0e0f3a6a286445e`
BLAKE2b-256	`44aa79971b4858e09bed16460425ba99171e3e68305f4a6621e2fd20a446601c`

Algorithm	Hash digest
SHA256	`08784bffd3b167428c7fbb3bc8d8b613c2dafa18f1b691623ef9e1885f836a6a`
MD5	`b9029a8919cc2a8d1a48440db2c685c5`
BLAKE2b-256	`76a08aacb2f73b860d618f3a0bfda5d54ccd34f4814e36d32036d198a7cf23dc`

Algorithm	Hash digest
SHA256	`7fab16ed8709effc8a2b7740e47e58751f9abb4d0a24f128cbc4fa055f8c6a45`
MD5	`7cd707a9f45f52c368a78ead2480d4a6`
BLAKE2b-256	`8fbeb8d746b7f7d636492c51c7704b8f27469a649653f56f7998331a74820db1`

Algorithm	Hash digest
SHA256	`15d0f71de781fae9d9d56badf869bbdc2f28baf85b7edce4265bde59e8c1f5ac`
MD5	`dfcf1775b7ee135d6a0c1c7d8180b126`
BLAKE2b-256	`e2246dbfeb72b7ff4589cdf906a2109c9f4a145ac00cdbaa3e884bf13dec163d`

Algorithm	Hash digest
SHA256	`5ed584bb41d6f5f7826f4e32f57283909fdb4dc4c15022328cf313300dba1919`
MD5	`17eba6c2157ae1965ea47f99f41478d1`
BLAKE2b-256	`c3f27b950a3b685e3966435d3b442fec97b27c87863e20a172a27a5d5e58388c`

Algorithm	Hash digest
SHA256	`c041103e592980e3639e20b651aa9e2a07332dc8671651dcf72c878bf31b35c6`
MD5	`f01f10d28ef92ea460c78f464bfc966c`
BLAKE2b-256	`796c0f8d11495405db9eefabadeb920f2944eff7631521ef34bc6005f7cc06f3`

Algorithm	Hash digest
SHA256	`6d9fa9f3f15410212da584eccce7f10b1e4858f3a90dfd2143d367ed4403940c`
MD5	`a51779769ef445a420af91eb6d1540f0`
BLAKE2b-256	`83bf3547bcc5d34baeb5773ebaaee64835cd83d6c7cf2b9cefe933a965690225`

Algorithm	Hash digest
SHA256	`2d9205f2f213e5ed92e42419bb299d45a43ca55eb53d35d3822f6ffec2441c70`
MD5	`076106de9365031cd4d9289fe6d0b1ef`
BLAKE2b-256	`5bb365938bf51dfef7aed0558948a0df4697ce11fae4f048021b936fd83517c1`

pylitehtml 0.2.1

Navigation

Verified details

Project links

GitHub Statistics

Maintainers

Unverified details

Meta

Classifiers

Project description

pylitehtml

效果展示

性能（对比 Playwright）

安装

快速上手

高级封装：html_to_png / markdown_to_png

html_to_png — 带默认样式的 HTML 渲染（核心、零额外依赖）

markdown_to_png — Markdown → HTML → 图片（可选、默认零依赖）

支持 / 不支持

API

render_file 与 Jinja2

线程 / 异步

从源码构建

基于

Project details

Verified details

Project links

GitHub Statistics

Maintainers

Unverified details

Meta

Classifiers

Release history Release notifications | RSS feed

Download files

Source Distribution

Built Distributions

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

`html_to_png` — 带默认样式的 HTML 渲染（核心、零额外依赖）

`markdown_to_png` — Markdown → HTML → 图片（可选、默认零依赖）