A high-performance, schema-based XML parser for TypeScript with strong type safety.
- 🚀 High Performance - Significantly faster than popular alternatives
- đź”’ Type Safety - Full TypeScript support with schema-based validation
- 📝 Schema Definition - Define your XML structure with a simple, declarative API
- âś… Validation - Automatic validation with detailed error messages
- 🎯 Zero Dependencies - Lightweight with no external dependencies
npm install fast-typed-xmlimport * as tx from "fast-typed-xml";
// Define your schema
const bookSchema = tx.object({
id: tx.string("id", "attribute"),
title: tx.string("title", "element"),
author: tx.string("author", "element"),
price: tx.number("price", "element"),
});
// Parse XML with type safety
const xml = `
<book id="bk001">
<title>The Great Gatsby</title>
<author>F. Scott Fitzgerald</author>
<price>10.99</price>
</book>
`;
const result = bookSchema.parse(xml);
// result is fully typed!
console.log(result.title); // "The Great Gatsby"The chart compares the throughput (operations per second) of three XML parsers. fast-xml-parser achieves 3,817 ops/s, and xml2js follows with 3,402 ops/s. In contrast, fast-typed-xml reaches 28,550 ops/s (with validation) and 41,837 ops/s (parse only), which is about 7.5Ă— faster than fast-xml-parser and 8.4Ă— faster than xml2js in parse-only mode. This clearly shows that fast-typed-xml delivers far superior performance, providing high-speed XML parsing suitable for demanding real-time or large-scale applications.
fast-typed-xml- Main schema APIfast-typed-xml/parser- Low-level XML parserfast-typed-xml/schema- Schema types and utilitiesfast-typed-xml/util- Utility functions
Parse XML element or attribute as a string.
// Element: <title>Book Title</title>
tx.string("title", "element")
// Attribute: <book id="123">
tx.string("id", "attribute")Parse XML element or attribute as a number.
// Element: <price>19.99</price>
tx.number("price", "element")
// Attribute: <item count="5">
tx.number("count", "attribute")Parse XML element or attribute as a boolean.
// Element: <available>true</available>
tx.boolean("available", "element")
// Attribute: <item inStock="false">
tx.boolean("inStock", "attribute")Parse XML element with nested structure.
const personSchema = tx.object({
name: tx.string("name", "element"),
age: tx.number("age", "element"),
email: tx.string("email", "element"),
});
// Parses:
// <person>
// <name>John Doe</name>
// <age>30</age>
// <email>john@example.com</email>
// </person>Parse multiple XML elements as an array.
// Array with element name
const genresSchema = tx.array("genre", tx.string());
// Parses: <genres><genre>Fiction</genre><genre>Drama</genre></genres>
// Array without element name (for root elements)
const booksSchema = tx.array(
tx.object({
title: tx.string("title", "element"),
})
);
// Parses: <books><book><title>The Great Gatsby</title></books>Tip
fast-typed-xml is a library focused on type-safe XML parsing. For more robust validation, it is safer and more efficient to use dedicated validation libraries like zod, arktype, or valibot on the parsed results.
You can define custom types with validation by extending the schema API. Here's an example of creating a custom userState type:
import * as tx from "fast-typed-xml";
import { AttributeSchema, ElementSchema, ValueSchema } from "fast-typed-xml/schema";
// 1. Define your custom type
type UserState = "active" | "inactive";
// 2. Create a parser function
const parseUserState = (v: string): UserState => {
switch (v) {
case "active":
case "inactive":
return v;
default:
throw new Error(`Invalid user state: ${v}`);
}
};
// 3. Define schema factories
const userStateAttributeSchema = <Optional extends boolean>(
name: string,
optional: Optional,
) => new AttributeSchema<UserState, Optional>(name, optional, parseUserState);
const userStateValueSchema = new ValueSchema<UserState>(parseUserState);
// 4. Create overloaded function with flexible signatures
function userState(): ValueSchema<UserState>;
function userState(
name: string,
kind: "attribute",
): AttributeSchema<UserState, false>;
function userState(
name: string,
kind: "attribute",
optional: false,
): AttributeSchema<UserState, false>;
function userState(
name: string,
kind: "attribute",
optional: true,
): AttributeSchema<UserState, true>;
function userState<Optional extends boolean>(
name: string,
kind: "attribute",
optional: Optional,
): AttributeSchema<UserState, Optional>;
function userState(
name: string,
kind: "element",
): ElementSchema<UserState, false>;
function userState(
name: string,
kind: "element",
optional: false,
): ElementSchema<UserState, false>;
function userState(
name: string,
kind: "element",
optional: true,
): ElementSchema<UserState, true>;
function userState<Optional extends boolean>(
name: string,
kind: "element",
optional: Optional,
): ElementSchema<UserState, Optional>;
function userState<Optional extends boolean>(
name?: string,
kind?: "attribute" | "element",
optional?: Optional,
):
| AttributeSchema<UserState, Optional>
| ValueSchema<UserState>
| ElementSchema<UserState, Optional> {
if (name !== undefined) {
switch (kind) {
case "attribute": {
return userStateAttributeSchema(name, optional as Optional);
}
case "element": {
return new ElementSchema(
name,
userStateValueSchema,
optional as Optional,
);
}
}
}
return userStateValueSchema;
}
// 5. Use your custom type in schemas
const schema = tx.object({
state: userState("state", "element")
});
const result = schema.parse("<user><state>active</state></user>"); // result: { state: "active" }This approach gives you:
- Type Safety: The return type is correctly inferred based on the parameters
- Validation: Custom parsing logic with error handling
- Flexibility: Can be used as an element, attribute, or standalone value
- Consistency: Same API pattern as built-in types like
tx.string()andtx.number()
Contributions are welcome! Please feel free to submit a Pull Request.
