Crate nom_xml

Expand description

nom-xml is a crate for parsing XML documents using the nom parser combinator crate.

This crate was initially created to be able to parse the following XML pattern which was troublesome in other Rust XML parsers explored at the time due to the limitations of the Serde crate:

<root>
    <header>
        <header_field1>Value1</header_field1>
        <header_field2>Value2</header_field2>
    </header>
    <body>
        <body_field1>BodyValue1</body_field1>
        <body_field2>BodyValue2</body_field2>
    </body>
    <header>
        <header_field1>Value1</header_field1>
        <header_field2>Value2</header_field2>
    </header>
    <body>
        <body_field1>BodyValue1</body_field1>
        <body_field2>BodyValue2</body_field2>
    </body>
</root>

It eventually evolved into implementing the XML 1.0 Specification - Fifth Edition as closely as possible. Nom was chosen specifically for its combinator parsing style which allowed for the implementation of the XML specification rules from their lowest level up to parsing the full document step-by-step. There is still a decent way to go to get to full compliance but the ultimate goal is to be able to parse any XML document, validate on schema, and write compliant XML documents. Unless complicated external entities are involved, this crate should already be able to parse most XML documents.

§Key Data Structure:

§`Document`

This enum encapsulates all of the top level types that comprise an XML document. The core variant is the Element(Tag,Box<Document>,Tag) type which allows recursive parsing of nested tags and their content.

pub enum Document {
    Prolog {
        xml_decl: Option<XmlDecl>,
        misc: Option<Vec<Misc>>,
        doc_type: Option<DocType>,
    },
    Element(Tag, Box<Document>, Tag),
    Content(Option<String>),
    Nested(Vec<Document>),
    Empty,
    EmptyTag(Tag),
    ProcessingInstruction(ProcessingInstruction),
    Comment(String),
    CDATA(String),
}

§Key Methods:

§Document::parse

The main way to parse an entire XML &str.

§Example:

use nom_xml::{parse::Parse, config::Config, Document};

fn main() {
    let xml = "<root><child>Content</child></root>";
    let (_, doc) = Document::parse(xml, &Config::default()).unwrap();
    println!("{doc:?}");
}

§Output:

Element(
    Tag {
        name:
            Name {
                prefix: None,
                local_part: "root",
            },
        attributes: None,
        state: Start,
    },
    Nested([
        Element(
            Tag {
                name:
                    Name {
                        prefix: None,
                        local_part: "child",
                    },
                attributes: None,
                state: Start,
            },
            Content("Content"),
            Tag {
                name:
                    Name {
                        prefix: None,
                        local_part: "child",
                    },
                attributes: None,
                state: End,
            },
        ),
    ]),
    Tag {
        name:
            Name {
                prefix: None,
                local_part: "root",
            },
        attributes: None,
        state: End,
    },
)

§Document::iter_with_depth

A method for iterating to a specific depth of an XML tree. See the ’extract_information_manual` example for more details

§Introducing `nom-xml-derive`

As of nom-xml version 0.3.0, nom-xml-derive is available for use. The nom-xml-derive derive macro crate was created to reduce the boilerplate necessary for users to extract data into structs. See the differences between manual implementations and derived counterpart implementations in the examples.