A pure Python HTML5 parser that just works.
Verified details
These details have been verified by PyPIProject links
GitHub Statistics
Maintainers
Project description
JustHTML
A pure Python HTML5 parser that just works. No C extensions to compile. No system dependencies to install. No complex API to learn.
Why use JustHTML?
1. Just... Correct ✅
It implements the official WHATWG HTML5 specification exactly. If a browser can parse it, JustHTML can parse it. It handles all the complex error-handling rules that browsers use.
- Verified Compliance: Passes all 8,500+ tests in the official html5lib-tests suite (used by browser vendors).
- 100% Coverage: Every line and branch of code is covered by integration tests.
- Fuzz Tested: Has parsed 3 million randomized broken HTML documents to ensure it never crashes or hangs (see benchmarks/fuzz.py).
- Living Standard: It tracks the living standard, not a snapshot from 2012.
2. Just... Python 🐍
JustHTML has zero dependencies. It's pure Python.
- Just Install: No C extensions to compile, no system libraries (like libxml2) required. Works on PyPy, WASM (Pyodide) (yes, it's in the test matrix), and anywhere Python runs.
- No dependency upgrade hassle: Some libraries depend on a large set of libraries, all which require upgrades to avoid security issues.
- Debuggable: It's just Python code. You can step through it with a debugger to understand exactly how your HTML is being parsed.
- Returns plain python objects: Other parsers return lxml or etree trees which means you have another API to learn. JustHTML returns a set of nested objects you can iterate over. Simple.
3. Just... Query 🔍
Find elements with CSS selectors. Just one method to learn - query() - and it uses CSS syntax you already know.
doc.query("div.container > p.intro") # Familiar CSS syntax
doc.query("#main, .sidebar") # Selector groups
doc.query("li:nth-child(2n+1)") # Pseudo-classes
4. Just... Fast Enough ⚡
If you need to parse terabytes of data, use a C or Rust parser (like html5ever). They are 10x-20x faster.
But for most use cases, JustHTML is fast enough. It parses the Wikipedia homepage in ~0.1s. It is the fastest pure-Python HTML5 parser available, outperforming html5lib and BeautifulSoup.
Comparison to other parsers
| Parser | HTML5 Compliance | Pure Python? | Speed | Query API | Notes |
|---|---|---|---|---|---|
| JustHTML | ✅ 100% | ✅ Yes | ⚡ Fast | ✅ CSS selectors | It just works. Correct, easy to install, and fast enough. |
html5lib |
🟡 88% | ✅ Yes | 🐢 Slow | ❌ None | The reference implementation. Very correct but quite slow. |
html5_parser |
🟡 84% | ❌ No | 🚀 Very Fast | 🟡 XPath (lxml) | C-based (Gumbo). Fast and mostly correct. |
selectolax |
🟡 68% | ❌ No | 🚀 Very Fast | ✅ CSS selectors | C-based (Lexbor). Very fast but less compliant. |
BeautifulSoup |
🔴 4% | ✅ Yes | 🐢 Slow | 🟡 Custom API | Wrapper around html.parser. Not spec compliant. |
html.parser |
🔴 4% | ✅ Yes | ⚡ Fast | ❌ None | Standard library. Chokes on malformed HTML. |
lxml |
🔴 1% | ❌ No | 🚀 Very Fast | 🟡 XPath | C-based (libxml2). Fast but not HTML5 compliant. |
Compliance scores from running the html5lib-tests suite (1,743 tree-construction tests). See benchmarks/correctness.py.
Installation
Requires Python 3.10 or later.
pip install justhtml
Example usage
Python API
from justhtml import JustHTML
html = "<html><body><div id='main'><p>Hello, <b>world</b>!</p></div></body></html>"
doc = JustHTML(html)
# 1. Traverse the tree
# The tree is made of SimpleDomNode objects.
# Each node has .name, .attrs, .children, and .parent
root = doc.root # #document
html_node = root.children[0] # html
body = html_node.children[1] # body (children[0] is head)
div = body.children[0] # div
print(f"Tag: {div.name}")
print(f"Attributes: {div.attrs}")
# 2. Query with CSS selectors
# Find elements using familiar CSS selector syntax
paragraphs = doc.query("p") # All <p> elements
main_div = doc.query("#main")[0] # Element with id="main"
bold = doc.query("div > p b") # <b> inside <p> inside <div>
# 3. Pretty-print HTML
# You can serialize any node back to HTML
print(div.to_html())
# Output:
# <div id="main">
# <p>
# Hello,
# <b>world</b>
# !
# </p>
# </div>
# 4. Streaming API (extremely fast and memory efficient)
# For massive files or when you don't need the full DOM tree.
# NOTE: Does not build a tree and _only_ runs the html5-compatible tokenizer
from justhtml import stream
for event, data in stream(html):
if event == "start":
tag, attrs = data
print(f"Start: {tag} with {attrs}")
elif event == "text":
print(f"Text: {data}")
elif event == "end":
print(f"End: {data}")
Supported CSS Selectors
JustHTML supports a comprehensive subset of CSS selectors:
| Selector | Example | Description |
|---|---|---|
| Tag | div |
Elements by tag name |
| Class | .intro |
Elements with class |
| ID | #main |
Element with ID |
| Universal | * |
All elements |
| Attribute | [href] |
Elements with attribute |
| Attr value | [type="text"] |
Exact attribute match |
| Attr prefix | [href^="https"] |
Attribute starts with |
| Attr suffix | [href$=".pdf"] |
Attribute ends with |
| Attr contains | [href*="example"] |
Attribute contains |
| Descendant | div p |
<p> inside <div> |
| Child | div > p |
Direct child |
| Adjacent | h1 + p |
Immediately after |
| Sibling | h1 ~ p |
Any sibling after |
| First child | :first-child |
First child element |
| Last child | :last-child |
Last child element |
| Nth child | :nth-child(2n+1) |
Nth child (odd, even, formula) |
| Not | :not(.hidden) |
Negation |
| Groups | h1, h2, h3 |
Multiple selectors |
Command Line Interface
You can also use JustHTML from the command line to pretty-print HTML files:
# Parse a file
python -m justhtml index.html
# Parse from stdin (great for piping)
curl -s https://example.com | python -m justhtml -
Develop locally and run the tests
-
Clone the repository:
git clone git@github.com:EmilStenstrom/justhtml.git cd justhtml
-
Install the library locally:
pip install -e ".[dev]"
-
Run the tests:
python run_tests.pyFor verbose output showing diffs on failures:
python run_tests.py -v
-
Run the benchmarks:
python benchmarks/performance.py
License
MIT. Free to use both for commercial and non-commercial use.
Project details
Verified details
These details have been verified by PyPIProject links
GitHub Statistics
Maintainers
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file justhtml-0.4.0.tar.gz.
File metadata
- Download URL: justhtml-0.4.0.tar.gz
- Upload date:
- Size: 115.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
8aa4d19a2bc4aab55cc9ed42a85a4e51e1e4c4c74ccd1aa359a44bab241eb621
|
|
| MD5 |
bc7d0ac7237701ff642d0e2c6ead6ac3
|
|
| BLAKE2b-256 |
a8a010765a87cc9bdd785ead76d594ad2831228c3060c8ccb37bb86a7d6abdf0
|
Provenance
The following attestation bundles were made for justhtml-0.4.0.tar.gz:
Publisher:
publish.yml on EmilStenstrom/justhtml
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
justhtml-0.4.0.tar.gz -
Subject digest:
8aa4d19a2bc4aab55cc9ed42a85a4e51e1e4c4c74ccd1aa359a44bab241eb621 - Sigstore transparency entry: 747188151
- Sigstore integration time:
-
Permalink:
EmilStenstrom/justhtml@214091105b65dacc08dd70201e21c0ea51d80fe3 -
Branch / Tag:
refs/tags/v0.4.0 - Owner: https://github.com/EmilStenstrom
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@214091105b65dacc08dd70201e21c0ea51d80fe3 -
Trigger Event:
release
-
Statement type:
File details
Details for the file justhtml-0.4.0-py3-none-any.whl.
File metadata
- Download URL: justhtml-0.4.0-py3-none-any.whl
- Upload date:
- Size: 59.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
bd8ba203b85537f1acd3b1d86da378a9189cfd2a08ca294d4ab274235bc7cfe2
|
|
| MD5 |
406e62b5dca4646f4d498293e8c0a00a
|
|
| BLAKE2b-256 |
57feb2be0d23fd7a2b7c671d3d9cd58963689ca7cc0450f7f3a7bf18f9f0cb8e
|
Provenance
The following attestation bundles were made for justhtml-0.4.0-py3-none-any.whl:
Publisher:
publish.yml on EmilStenstrom/justhtml
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
justhtml-0.4.0-py3-none-any.whl -
Subject digest:
bd8ba203b85537f1acd3b1d86da378a9189cfd2a08ca294d4ab274235bc7cfe2 - Sigstore transparency entry: 747188157
- Sigstore integration time:
-
Permalink:
EmilStenstrom/justhtml@214091105b65dacc08dd70201e21c0ea51d80fe3 -
Branch / Tag:
refs/tags/v0.4.0 - Owner: https://github.com/EmilStenstrom
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@214091105b65dacc08dd70201e21c0ea51d80fe3 -
Trigger Event:
release
-
Statement type:
