Decode Value from database to prevent double enconding of umlauts

[wanze]

Hi @tiefenb,

Thanks for the pull request. I would like to introduce test cases to make sure that everything also works for other languages before merging this. From my understanding, $sanitizer::entities1 should prevent double encoding issues, but this might not work correctly with the introduced ENT_XML1 flag. 🙈

We should add a test explicitly covering the following cases:

• Umlauts öüä are not encoded
• Turkish characters should not get encoded, see https://github.com/wanze/SeoMaestro/issues/15
• Scandic letters should not get encoded, see https://processwire.com/talk/topic/20817-seomaestro/?do=findComment&comment=194190

As everything worked fine before I added support to not encode german umlaut characters (see https://github.com/wanze/SeoMaestro/commit/057d8119f37f21bfbf8f4b3aad9e2447b4c90596), I'm thinking of simply revert the change introducing the ENT_XML1 flag and then manually replace encoded german umlauts.

Cheers

[jnessier]

Any updates in here?

[wanze]

Thanks @tiefenb, finally found some time to extend the tests and merge this. New release coming soon.