Removed dependency on 'org.apache.commons.codec'

[cweitkamp]

• Removed dependency on 'org.apache.commons.codec' to fix OH3 add-ons build: https://ci.openhab.org/job/openHAB-Addons-port/108/

Related to openhab/openhab-core#1436

Signed-off-by: Christoph Weitkamp [email protected]

[lolodomo]

I have the feeling it will break the Freebox binding.

[lolodomo]

I have the feeling this will not be the same result.
The new call will add a space between each hexa value while the previous code did not.

[cweitkamp]

It will work - check it out. I wrote a unit test for each change.

    @Test
    public void testFreebox() {
        byte[] rawSubject = "test".getBytes(StandardCharsets.UTF_8);

        byte[] hexBytes = new Hex().encode(rawSubject);
        String subject = new String(hexBytes, StandardCharsets.UTF_8);

        assertThat(subject, is(HexUtils.bytesToHex(rawSubject)));
    }

[lolodomo]

Ok, if you are sure, that is fine for me.

[kaikreuzer]

@lolodomo Then please remove your "changes requested" review :-)

[lolodomo]

I don't know how to suppress it !
I approved it even if I check only the part relative to the Freebox binding.

[kaikreuzer]

Seems you made it :-)

[lolodomo]

Yes, I approved it globally but only reviewed one part.

[kaikreuzer]

Many thanks - you're the best 😎!

[lolodomo]

@cweitkamp : just after installing version 2.5.5, I discovered the authentication to the Freebox is no more working. I am afraid you have broken the binding with your change in the method hmacSha1 !

[lolodomo]

If I restore the previous code in the method hmacSha1, the binding is working again.

[lolodomo]

Here is the comparison with the old code and your new code:

20:03:08.989 [DEBUG] [reebox.internal.api.FreeboxApiManager] - hmacSha1 new code: *0F7480528064D7F06AA363F4E9646C83B9CFCCD1*
20:03:08.995 [DEBUG] [reebox.internal.api.FreeboxApiManager] - hmacSha1 old code: *0f7480528064d7f06aa363f4e9646c83b9cfccd1*
20:03:09.001 [DEBUG] [reebox.internal.api.FreeboxApiManager] - hmacSha1 old == new: false

The difference is the case. I will apply a lowercase to the result with the new code to see if it works.

[cpmeister]

If you find a fix, it would be a good idea to make a unit test out of it.

[lolodomo]

@cweitkamp : where is your unit test, I don't find it ?

[cweitkamp]

@lolodomo I am sorry to break it. See #7460 (comment) for the unit test. I did not commit it anywhere. Just executed it locally.

IIRC I had to transform the result to lower case in another case too.

[lolodomo]

In case it broke several bindings, it could be necessary to produce a new distribution 2.5.6 very quickly