Multipart encoded file attachment names are not properly decoded

[karladler]

When using attachment headers that include multiple parts of encoded strings the parser fails to return the proper file names.

I think the behavior should be like in emailjs-mime-codec mimeWordsDecode function.

Here are a few unit tests for validation:

  it('should return an empty string for empty input', () => {
    expect(mimeWordsDecode('')).toBe('')
  })

  it('should return the original string if not MIME-encoded', () => {
    const plain = 'simple-filename.txt'
    expect(mimeWordsDecode(plain)).toBe(plain)
  })

  it('should decode a single Base64-encoded word', () => {
    const encoded = '=?UTF-8?Q?See_on_=C3=B5hin_test?='
    const decoded = mimeWordsDecode(encoded)
    expect(decoded).toBe('See on õhin test')
  })

  it('should decode multipart Base64-encoded words into a single string', () => {
    const encoded
      = '=?utf-8?B?UmVwb3J0X0ZpbmFuY2lhbF9TdW1t?= =?utf-8?B?YXJ5XzIwMjVfSmFuLnBkZg==?='
    const decoded = mimeWordsDecode(encoded)
    expect(decoded).toBe('Report_Financial_Summary_2025_Jan.pdf')
  })

  it('should handle different encodings in the same header', () => {
    const plain = `=?utf-8?B?anVzdCBhbiBleGFtcGxlX18=?=
 =?utf-8?Q?unencoded.xlsx?=`
    expect(mimeWordsDecode(plain)).toBe('just an example__unencoded.xlsx')
  })

[MQpeng]

Contributions are welcome!