Java: XML canonicalization

Question

What's the easiest way to make a canonical form of a XML file in Java? Do you have some done code for that? I've found several links on the net, like this, this, and this, but I can't make it to work :/

Thanks,

Ivan

EDIT: I used the canonicalizer that was proposed down there, but I get strange results. To be more precize, this method doesn't delete white spaces between elements... This is what I get:

<Metric xmlns="http://www.ibm.com/wsla" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" name="total_memory_consumption_metric" type="double" unit="Mbit" xsi:schemaLocation="http://www.ibm.com/wsla WSLA.xsd">                        <Source>ServiceProvider</Source>                        <MeasurementDirective resultType="double" xsi:type="StatusRequest">                              <RequestURI> ***unused*** </RequestURI>                        </MeasurementDirective>                  </Metric>

Same problem, did you manage to solve it by now ? – pwillemet Feb 28 '18 at 17:30 — pwillemet, Feb 28 '18 at 17:30

score 27 · Accepted Answer · edited Nov 25 '20 at 03:21

27

The Canonicalizer class at Apache XML Security project.

Initialize the library.

org.apache.xml.security.Init.init();

Convert your XML.

Canonicalizer canon = Canonicalizer.getInstance(Canonicalizer.ALGO_ID_C14N_OMIT_COMMENTS);
byte canonXmlBytes[] = canon.canonicalize(yourXmlBytes);
String canonXmlString = new String(canonXmlBytes);

edited Nov 25 '20 at 03:21

Basil Bourque

303,325
100
852
1,154

answered Feb 25 '11 at 13:20

eolith

1,366
10
14

22

To anyone who hasn't worked with the Apache XML Security library before, you must initialize the library using the static method "org.apache.xml.security.Init.init();" before invoking any code from that library, or you'll get an error. – Stew Nov 15 '12 at 22:16
1

The point @Stew makes is quite important. It should really be part of the answer itself. – aroth Apr 11 '16 at 04:01
I find this solution doesn't work on emoji character which becomes "??" after canonicalized. – Solomon Tam Aug 22 '18 at 09:43
1

@Stew I added your library initialization to the original answer. Better late than never :) – eolith Feb 12 '20 at 11:57

score 2 · Answer 2 · answered Nov 04 '15 at 00:01

2

Another option is nu.xom.canonical.Canonicalizer if you're using XOM, or if you don't otherwise have a need for Apache XML Security.

answered Nov 04 '15 at 00:01

David Moles

48,006
27
136
235

This is the better option. In my case I called Serializer first and then directed the result to Canonicalizer. The result was a more readable canonical XML. – Georgios F. Apr 20 '20 at 16:37

Java: XML canonicalization

2 Answers2

Linked