Written Tutorial

Problem

You want to know the size of a non-binary document.

Solution

Applies to MarkLogic versions 7+

If we want to know the size of a binary, it’s easy: xdmp:binary-size($doc).

However, there isn’t a special function for non-binary documents. Instead, we can convert it the content to a different format where we can measure the size:

var doc = fn.doc('/test-uri1.xml')
xdmp.binarySize(fn.head(xdmp.unquote(xdmp.quote(doc), null, "format-binary")).root)

let $doc := fn:doc("/test-uri1.xml")
return
  xdmp:binary-size(xdmp:unquote(xdmp:quote($doc),(),"format-binary")/binary())

Discussion

You might be tempted to use fn:string-length(). In some cases, that will work. The critical thing to realize is that fn:string-length() gives you the number of characters, not the size of the document. These can be different, depending on the characters in the document. For instance, consider this simple example:

var doc =
  {
    zoo: {
      lions: "die Löwen"
    }
  }

result = {
  "string": fn.stringLength(xdmp.quote(doc)),
  "binary": xdmp.binarySize(fn.head(xdmp.unquote(xdmp.quote(doc), null, "format-binary")).root)
}

let $doc := 
  <zoo>
    <lions lang="de">die Löwen</lions>
  </zoo>
return (
  fn:string-length(xdmp:quote($doc)),
  xdmp:binary-size(xdmp:unquote(xdmp:quote($doc),(),"format-binary")/binary())
)

This returns (45, 46) for the XML document and {"string": 29, "binary": 30 } for the JSON document. Why? The ö is a 2-character letter.

Note that while the example code uses XML sample input documents, it works just the same for JSON.