WAPBinaryXML

WAPBinaryXML is a Swift encoder/decoder for WBXML 1.3. Intended for ActiveSync, but usable with other WBXML dialects too. The package has no dependencies and does not ship protocol code pages.

Swift 6.2 or later. Should work on all platforms, iOS 13 and peers and up.

Quick start

Declare the tag mappings used by the peer. This is a small subset of the ActiveSync AirSync page, not a complete ActiveSync schema:

import WAPBinaryXML

let airSync = WBXMLCodePage(page: 0, namespace: "AirSync:", tokens: [
  ( 0x05, "Sync"         ),
  ( 0x0B, "SyncKey"      ),
  ( 0x0E, "Status"       ),
  ( 0x0F, "Collection"   ),
  ( 0x12, "CollectionId" ),
  ( 0x1C, "Collections"  )
])
let codePages = WBXMLCodePage.Set([ airSync ])
try codePages.validate()

Build and encode a document using the page number from airSync. In the builder, children without an explicit page inherit their enclosing page:

let root = WBXMLBuilder.element("Sync", page: airSync.page) {
  $0.element("Collections") {
    $0.element("Collection") {
      $0.element("SyncKey", text: "0")
      $0.element("CollectionId", text: "inbox")
    }
  }
}

let encoder = WBXMLEncoder(codePages: codePages)
var bytes = [ UInt8 ]()
try encoder.encode(root, to: &bytes)

Decode from the encoded byte collection:

let decoder  = WBXMLDecoder(codePages: codePages)
let document = try decoder.decode(bytes)

let collection = document.root[tag: "Collections"]?[tag: "Collection"]
let syncKey    = collection?[tag: "SyncKey"]?.textValue

Both codecs can be reused for multiple documents. To select a charset or public identifier, wrap the root in WBXMLDocument and pass that to the encoder.

Supported WBXML constructs

Construct	Decode	Encode
WBXML 1.3 header version	Yes	Yes
Tag code pages and page switches	Yes	Yes
Attribute code pages and attributes	Yes	Yes
Inline and string-table text	Yes	Yes
Numeric and literal public identifiers	Yes	Yes
Literal tag and attribute names	Yes	Yes
Entities and opaque data	Yes	Yes
Processing instructions	Yes	Yes
WBXML extension tokens	Yes	Yes

Character sets

Latin 1 or UTF-8 only.

MIBenum	Interpretation
0	Unspecified value, treated as UTF-8
4	ISO-8859-1
106	UTF-8

Who

This is brought to you by Helge Heß / ZeeZide. We like feedback, GitHub stars, cool contract work, and any form of praise you can think of. Buy one of our apps, you don't have to use them! 😀