Files
buuh/packages/core/lib/router.js
T
Bendik Aagaard LynghaugandClaude Fable 5 501e8f5927 fix(core): URL normalization that survives non-ASCII and malformed input
Replace nanorouter's regex+decodeURI normalization with the WHATWG URL
parser. Path segments are percent-decoded exactly once in the trie, with
a keep-raw fallback instead of the URIError that crashed v7 on a literal
'%'. Route definitions and locations are NFC-normalized so 'café' matches
regardless of composition. Hash-to-slash rewriting now handles every hash,
not just the first, and state.href is decoded for humans (raw on failure).

Behavior changes from v7: params are no longer double-decoded (%2540 →
'%40', not '@'), and unmatchable encodings fall back to raw segments
instead of silently 404ing.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014NgfSjHE11oFpoSnLVKLXd
2026-09-08 16:50:34 +02:00

252 lines
7.0 KiB
JavaScript

// Ported from nanorouter 4.0.0 and wayfarer 7.0.1 with its trie (MIT)
// https://github.com/choojs/nanorouter — https://github.com/yoshuawuyts/wayfarer
// Consolidated into a single module: trie-based router with :params and
// * wildcards, plus URL normalization on top.
//
// Normalization differs deliberately from v7: locations are parsed with the
// WHATWG URL parser and each path segment is percent-decoded exactly once,
// falling back to the raw segment instead of throwing on malformed input
// (v7's decodeURI crashed on a literal '%'). Both route definitions and
// incoming segments are NFC-normalized so 'café' matches 'café' regardless
// of how the é was composed.
import { ok, equal, notEqual } from './assert.js'
function has (object, property) {
return Object.prototype.hasOwnProperty.call(object, property)
}
// Decode a path segment exactly once; keep the raw segment when it isn't
// valid percent-encoding. Always NFC-normalize.
export function decodeSegment (segment) {
try {
return decodeURIComponent(segment).normalize('NFC')
} catch (e) {
return segment.normalize('NFC')
}
}
class Trie {
constructor () {
this.trie = { nodes: {} }
}
// create a node on the trie at route and return it
create (route) {
equal(typeof route, 'string', 'route should be a string')
// strip leading '/' and split routes
const routes = route.replace(/^\//, '').split('/')
const self = this
function createNode (index, trie) {
const thisRoute = has(routes, index) && routes[index]
if (thisRoute === false) return trie
let node = null
if (/^:|^\*/.test(thisRoute)) {
// if node is a name match, set name and append to ':' node
if (!has(trie.nodes, '$$')) {
node = { nodes: {} }
trie.nodes.$$ = node
} else {
node = trie.nodes.$$
}
if (thisRoute[0] === '*') {
trie.wildcard = true
}
trie.name = thisRoute.replace(/^:|^\*/, '')
} else {
// Literal segment: store it decoded + NFC-normalized so route
// definitions and incoming URLs compare in one canonical form.
const key = decodeSegment(thisRoute)
if (!has(trie.nodes, key)) {
node = { nodes: {} }
trie.nodes[key] = node
} else {
node = trie.nodes[key]
}
}
return createNode(index + 1, node)
}
return createNode(0, self.trie)
}
// match a route on the trie and return the node
match (route) {
equal(typeof route, 'string', 'route should be a string')
const routes = route.replace(/^\//, '').split('/')
const params = {}
function search (index, trie) {
// either there's no match, or we're done searching
if (trie === undefined) return undefined
if (routes[index] === undefined) return trie
// Segments arrive percent-encoded from the URL parser; decode each
// exactly once (create() stored literal keys in the same form).
const thisRoute = decodeSegment(routes[index])
if (has(trie.nodes, thisRoute)) {
// match regular routes first
return search(index + 1, trie.nodes[thisRoute])
} else if (trie.name) {
// match named routes
params[trie.name] = thisRoute
return search(index + 1, trie.nodes.$$)
} else if (trie.wildcard) {
// match wildcards; decode per segment so an encoded '/' can't
// change the segment structure mid-decode
params.wildcard = routes.slice(index).map(decodeSegment).join('/')
// return early, or else search may keep recursing through the wildcard
return trie.nodes.$$
} else {
// no matches found
return search(index + 1)
}
}
let node = search(0, this.trie)
if (!node) return undefined
node = Object.assign({}, node)
node.params = params
return node
}
// mount a trie onto a node at route
mount (route, trie) {
equal(typeof route, 'string', 'route should be a string')
equal(typeof trie, 'object', 'trie should be a object')
const split = route.replace(/^\//, '').split('/')
let node = null
if (split.length === 1) {
node = this.create(split[0])
} else {
node = this.create(split.join('/'))
}
Object.assign(node.nodes, trie.nodes)
if (trie.name) node.name = trie.name
// delegate properties from '/' to the new node
// '/' cannot be reached once mounted
if (node.nodes['']) {
Object.keys(node.nodes['']).forEach(function (key) {
if (key === 'nodes') return
node[key] = node.nodes[''][key]
})
Object.assign(node.nodes, node.nodes[''].nodes)
delete node.nodes[''].nodes
}
}
}
function wayfarer (dft) {
const _default = (dft || '').replace(/^\//, '')
const _trie = new Trie()
emit._trie = _trie
emit.on = on
emit.emit = emit
emit.match = match
emit._wayfarer = true
return emit
function on (route, cb) {
equal(typeof route, 'string')
equal(typeof cb, 'function')
route = route || '/'
if (cb._wayfarer && cb._trie) {
_trie.mount(route, cb._trie.trie)
} else {
const node = _trie.create(route)
node.cb = cb
node.route = route
}
return emit
}
function emit (route, ...args) {
const matched = match(route)
return matched.cb(matched.params, ...args)
}
function match (route) {
notEqual(route, undefined, "'route' must be defined")
const matched = _trie.match(route)
if (matched && matched.cb) return new Route(matched)
const dft = _trie.match(_default)
if (dft && dft.cb) return new Route(dft)
throw new Error("route '" + route + "' did not match")
}
function Route (matched) {
this.cb = matched.cb
this.route = matched.route
this.params = matched.params
}
}
// Reduce a location (path, full URL, or electron file:// URL) to a
// matchable path: pathname plus any hash segments rewritten to slashes,
// query dropped. Percent-decoding is NOT done here — the trie decodes per
// segment — so a malformed sequence can never throw during routing.
function pathname (routename) {
let url
try {
url = new URL(routename, 'http://localhost')
} catch (e) {
return routename
}
let path = url.pathname
// electron support: file:///path/to/index.html routes as '/'
if (url.protocol === 'file:') {
path = path.replace(/\/[^/]*\.html?$/, '') || '/'
}
const hash = url.hash.replace(/^#/, '/').replaceAll('#', '/')
return path + hash
}
export default class Nanorouter {
constructor (opts) {
opts = opts || {}
this.router = wayfarer(opts.default || '/404')
}
on (routename, listener) {
equal(typeof routename, 'string')
routename = routename.replace(/^[#/]/, '')
this.router.on(routename, listener)
}
emit (routename) {
equal(typeof routename, 'string')
routename = pathname(routename)
return this.router.emit(routename)
}
match (routename) {
equal(typeof routename, 'string')
routename = pathname(routename)
return this.router.match(routename)
}
}
export { Trie, wayfarer, ok }