#node #xml-document #dom #xml-parser #name #traits #specification


A Rust crate providing a reasonably faithful implementation of the W3C DOM Core

11 releases

0.2.6 May 31, 2022
0.2.5 Mar 22, 2021
0.2.3 Apr 16, 2020
0.2.0 Mar 31, 2020
0.1.4 Mar 31, 2020

#6 in #xml-document

Download history 165/week @ 2023-12-14 45/week @ 2023-12-21 56/week @ 2023-12-28 151/week @ 2024-01-04 155/week @ 2024-01-11 164/week @ 2024-01-18 215/week @ 2024-01-25 153/week @ 2024-02-01 73/week @ 2024-02-08 66/week @ 2024-02-15 44/week @ 2024-02-22 55/week @ 2024-02-29 30/week @ 2024-03-07 23/week @ 2024-03-14 7/week @ 2024-03-21 32/week @ 2024-03-28

99 downloads per month
Used in tiles7800

MIT license

4.5K SLoC

Crate xml_dom

A Rust crate providing a reasonably faithful implementation of the W3C Document Object Model Core, Level 2.

MIT License Minimum Rust Version crates.io docs.rs Build Audit GitHub stars

This crate provides a trait-based implementation of the DOM with minimal changes to the style and semantics defined in the Level 2 specification. The specific mapping from the IDL in the specification is described in the documentation, however from a purely style point of view the implementation has the following characteristics:

  1. It maintains a reasonable separation between the node type traits and the tree implementation using opaque an RefNode reference type.
  2. Where possible the names from IDL are used with minimal conversion.
  3. All IDL attributes become trait functions (attribute "foo" becomes foo(), set_foo(), and unset_foo()).

This leads to a replication of the typical DOM programmer experience where casting between the node traits is required. This is supported by the xml_dom::convert module.


use xml_dom::level2::*;
use xml_dom::level2::convert::*;

// Bootstrap; get an instance of `DOMImplementation`. The mechanism for this is
// intentionally undefined by the specification.
let implementation = get_implementation();

// Create a `DocumentType` instance.
let document_type = implementation
        Some("-//W3C//DTD XHTML 1.0 Transitional//EN"),

// Create a new `Document` using the document type defined above. Note that this 
// also has the side-effect of creating the document's root element named "html".
let mut document_node = implementation
    .create_document(Some("http://www.w3.org/1999/xhtml"), Some("html"), Some(document_type))

// Cast the returned document `RefNode` into a `RefDocument` trait reference
let document = as_document_mut(&mut document_node).unwrap();

// Fetch the document's root element as a node, then cast to `RefElement`.
let mut root_node = document.document_element().unwrap();
let root = as_element_mut(&mut root_node).unwrap();

// Create an `Attribute` instance on the root element.
root.set_attribute("lang", "en");

// Create two child `Element`s of "html".
let _head = root.append_child(document.create_element("head").unwrap());
let _body = root.append_child(document.create_element("body").unwrap());

// Display as XML.
let xml = document_node.to_string();
println!("document 2: {}", xml);

This should result in the following XML; note that formatting was added for this document, the provided implementation of Display for RefNode does not format the output.

  PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 
  SYSTEM "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html lang="en">


Currently only one feature, quick_parser, is provided which provides a new module parser with the single public function. This feature is enabled by default.

pub fn read_xml(xml: &str) -> Result<RefNode>;

This will parse the document and return a new RefNode that corresponds to the Document trait.


Version 0.2.6

  • Updated quick-xml dependency.
    • Had to update the handling of errors from the underlying parser.
    • Had to handle the CData parser explicitly.

Version 0.2.5

  • Added parser::from_reader function alongside the existing parser::from_str to allow for streaming input of the underlying source.

Version 0.2.4

Version 0.2.3

  • Mostly documentation updates/fixes.
  • Made namespace support respect ProcessingOptions.

Version 0.2.2

  • Mostly documentation updates/fixes.
  • Fixed a defect in handling of xml:id.

Version 0.2.1

  • Bug Fixes:
    • Fixed a publishing error in Travis
    • Separated Attribute::owner_element from Node::parent_node, they aren't the same.
    • Fixed handling of owner_element in Attribute tests
    • Fixed the implementation of WrongDocument error in Node::insert_before and used the same in Element::set_attribute_node.
    • Fixed escaping of values to happen on get not set.
    • Implemented attribute value normalization and end-of-line handling from the XML 1.1 spec.
      • Required added a dependency on regex
      • Added EntityResolver trait for callback into DOM.
      • Expansion of Entity node children is not yet complete.
  • Code clean-up:
    • Added some macros in trait_impls for a number of common patterns.
    • Simplified some existing match expressions.

Version 0.2.0

  • Cleaned up documentation.
  • Stable release.

Version 0.1.4

  • BREAKING refactored to add a level2 module, allowing other levels to be added at a later time. Also moved extensions into level2::ext module.
  • BREAKING renamed methods to conform with DOM names:
    • Node::name to Node::node_name;
    • CharacterData::substring to CharacterData::substring_data;
    • CharacterData::append to CharacterData::append_data;
    • CharacterData::insert to CharacterData::insert_data;
    • CharacterData::delete to CharacterData::delete_data;
    • CharacterData::replace to CharacterData::replace_data.
  • Implemented the following methods:
    • Node::clone_node;
    • Node::normalize;
    • Namespaced::normalize_mappings.
  • Added the following DOM methods:
    • Attribute::owner_element;
    • Node::local_name;
    • Node::namespace_uri;
    • Node::prefix.
  • CI builds now working with Travis, rust-xml_dom.
  • Added quick_xml based text parser.
  • Make this the 0.2.0 candidate.

Version 0.1.3

  • More unit tests overall, especially for append/insert/replace child
  • Add support for xml declaration (XmlDecl, XmlVersion), not reusing processing instruction
  • Support the last Level 2 extended interfaces (Entity, EntityReference, and Notation).
    • Also, add create_notation, create_entity, and create_internal_entity to dom_impl.
  • Implement an options (ProcessingOptions and DOMImplementation::create_document_with_options) capability to turn on extended processing behaviors.
  • Fixed some nested borrow issues.

Version 0.1.2

  • Focus on feature completion:
    • implement all core trait features
    • implement extended trait features for currently supported traits
    • unescaping text
    • refactor NodeImpl for extended traits
  • Unit tests, lot's of unit tests

Version 0.1.1

  • Focus on API, separate the traits from implementation more cleanly.
  • Better Display formatting
  • Better append_child rule support
  • Have support for namespace resolution
  • Have support for text escaping on setting values
  • More examples, fleshing out more of the common methods.
  • Note, this is NOT YET ready for production usage.

Version 0.1.0

  • Focus on modeling as traits, not all methods actually implemented.
  • Note, this is NOT YET ready for production usage.


  1. More tests required.
  2. More doc test/examples.
  3. Ensure correct notation replacement.


~60K SLoC