markuptree

A modern, zero-dependency drop-in replacement for html5lib.

Installation

pip install markuptree

Quickstart

import markuptree

# Parse an HTML document
doc = markuptree.parse("<html><body><p>Hello <b>world</b></p></body></html>")

# Parse a fragment
fragments = markuptree.parseFragment("<p>Hello</p><p>World</p>")

# Serialize a tree back to HTML
html = markuptree.serialize(doc, tree="etree")

# Use the HTMLParser class directly
parser = markuptree.HTMLParser(tree="etree")
doc = parser.parse("<p>Hello</p>")
print(parser.errors)           # list of parse errors
print(parser.documentEncoding) # detected encoding

Tree Builders

Two built-in backends for constructing parse trees:

# Get a tree builder class
TB = markuptree.getTreeBuilder("etree")

Tree Walkers

Walk a parsed tree and yield serializer tokens:

Walker = markuptree.getTreeWalker("etree")
walker = Walker(doc)
for token in walker:
    print(token)

Serializer

from markuptree.serializer import HTMLSerializer

Walker = markuptree.getTreeWalker("etree")
s = HTMLSerializer(
    omit_optional_tags=True,
    quote_attr_values="always",
    minimize_boolean_attributes=True,
    alphabetical_attributes=True,
)
html = s.render(Walker(doc))

Serializer Options

Filters

Token stream filters that sit between a tree walker and the serializer:

from markuptree.treewalkers.etree import TreeWalker
from markuptree.filters.sanitizer import Filter as SanitizerFilter
from markuptree.serializer import HTMLSerializer

walker = TreeWalker(doc)
safe = SanitizerFilter(walker)
html = HTMLSerializer().render(safe)

Migration from html5lib

# Change this:
import html5lib

# To this:
import markuptree as html5lib

Or use the compatibility shim:

from markuptree._compat import *

License

MIT