Skip to content

fix: decode HTTP headers as latin1 instead of utf8 #4768

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mcollina merged 1 commit into from
Jan 25, 2026
+51 −22
Merged

fix: decode HTTP headers as latin1 instead of utf8 #4768

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 3 additions & 22 deletions lib/core/util.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -431,22 +431,17 @@ function parseHeaders (headers, obj) {
val = [val]
obj[key] = val
}
val.push(headers[i + 1].toString('utf8'))
val.push(headers[i + 1].toString('latin1'))
} else {
const headersValue = headers[i + 1]
if (typeof headersValue === 'string') {
obj[key] = headersValue
} else {
obj[key] = Array.isArray(headersValue) ? headersValue.map(x => x.toString('utf8')) : headersValue.toString('utf8')
obj[key] = Array.isArray(headersValue) ? headersValue.map(x => x.toString('latin1')) : headersValue.toString('latin1')
}
}
}

// See https://github.com/nodejs/node/pull/46528
if ('content-length' in obj && 'content-disposition' in obj) {
obj['content-disposition'] = Buffer.from(obj['content-disposition']).toString('latin1')
}

return obj
}

Expand All @@ -461,34 +456,20 @@ function parseRawHeaders (headers) {
*/
const ret = new Array(headersLength)

let hasContentLength = false
let contentDispositionIdx = -1
let key
let val
let kLen = 0

for (let n = 0; n < headersLength; n += 2) {
key = headers[n]
val = headers[n + 1]

typeof key !== 'string' && (key = key.toString())
typeof val !== 'string' && (val = val.toString('utf8'))
typeof val !== 'string' && (val = val.toString('latin1'))

kLen = key.length
if (kLen === 14 && key[7] === '-' && (key === 'content-length' || key.toLowerCase() === 'content-length')) {
hasContentLength = true
} else if (kLen === 19 && key[7] === '-' && (key === 'content-disposition' || key.toLowerCase() === 'content-disposition')) {
contentDispositionIdx = n + 1
}
ret[n] = key
ret[n + 1] = val
}

// See https://github.com/nodejs/node/pull/46528
if (hasContentLength && contentDispositionIdx !== -1) {
ret[contentDispositionIdx] = Buffer.from(ret[contentDispositionIdx]).toString('latin1')
}

return ret
}

Expand Down
48 changes: 48 additions & 0 deletions test/node-test/util.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,11 +87,59 @@ test('parseHeaders', () => {
assert.deepEqual(util.parseHeaders([Buffer.from('key'), [Buffer.from('value1'), Buffer.from('value2'), Buffer.from('value3')]]), { key: ['value1', 'value2', 'value3'] })
})

test('parseHeaders decodes values as latin1, not utf8', () => {
// These bytes (0xE2, 0x80, 0xA6) are the UTF-8 encoding of U+2026 (ellipsis)
// When decoded as latin1, they should be 3 separate characters: â, €, ¦
// When incorrectly decoded as UTF-8, they would be a single character: …
const latin1Bytes = Buffer.from([0xe2, 0x80, 0xa6])
const result = util.parseHeaders([Buffer.from('x-test'), latin1Bytes])

assert.strictEqual(result['x-test'].length, 3)
assert.strictEqual(result['x-test'].charCodeAt(0), 0xe2)
assert.strictEqual(result['x-test'].charCodeAt(1), 0x80)
assert.strictEqual(result['x-test'].charCodeAt(2), 0xa6)
})

test('parseHeaders decodes duplicate header values as latin1', () => {
const latin1Bytes = Buffer.from([0xe2, 0x80, 0xa6])
const result = util.parseHeaders([
Buffer.from('x-test'), Buffer.from('first'),
Buffer.from('x-test'), latin1Bytes
])

assert.deepEqual(result['x-test'][0], 'first')
assert.strictEqual(result['x-test'][1].length, 3)
assert.strictEqual(result['x-test'][1].charCodeAt(0), 0xe2)
})

test('parseHeaders decodes array header values as latin1', () => {
const latin1Bytes = Buffer.from([0xe2, 0x80, 0xa6])
const result = util.parseHeaders([Buffer.from('x-test'), [latin1Bytes, latin1Bytes]])

assert.strictEqual(result['x-test'].length, 2)
assert.strictEqual(result['x-test'][0].length, 3)
assert.strictEqual(result['x-test'][0].charCodeAt(0), 0xe2)
})

test('parseRawHeaders', () => {
assert.deepEqual(util.parseRawHeaders(['key', 'value', Buffer.from('key'), Buffer.from('value')]), ['key', 'value', 'key', 'value'])
assert.deepEqual(util.parseRawHeaders(['content-length', 'value', 'content-disposition', 'form-data; name="fieldName"']), ['content-length', 'value', 'content-disposition', 'form-data; name="fieldName"'])
})

test('parseRawHeaders decodes values as latin1, not utf8', () => {
// These bytes (0xE2, 0x80, 0xA6) are the UTF-8 encoding of U+2026 (ellipsis)
// When decoded as latin1, they should be 3 separate characters
// When incorrectly decoded as UTF-8, they would be a single character
const latin1Bytes = Buffer.from([0xe2, 0x80, 0xa6])
const result = util.parseRawHeaders([Buffer.from('x-test'), latin1Bytes])

assert.strictEqual(result[0], 'x-test')
assert.strictEqual(result[1].length, 3)
assert.strictEqual(result[1].charCodeAt(0), 0xe2)
assert.strictEqual(result[1].charCodeAt(1), 0x80)
assert.strictEqual(result[1].charCodeAt(2), 0xa6)
})

test('serializePathWithQuery', () => {
const tests = [
[{ id: BigInt(123456) }, 'id=123456'],
Expand Down
Loading