Use empty keys instead of _empty_ in json decoding#1926
Conversation
|
Should be mentioned as a BC break in the changelog/NEWS |
|
@staabm I don't usually do that in the PR as there might be changes and I would have to sync it. So I almost always update news and upgrading after I merge the PR ;) |
| ["E"]=> | ||
| float(1.23456789E+34) | ||
| ["_empty_"]=> | ||
| [""]=> |
There was a problem hiding this comment.
Not really sure which one do you mean as github is showing this comment on two different places (github is being wierd though :) ). One in the commit view (File changed)
> - ["_empty_"]=>
> + [""]=>
which is to replace _empty_ and one in Conservation which shows
- ["e"]=>
+ ["e"]=>
which I'm not really sure. It might be some new line but if I look from the command line git diff, I don't see anything so might also be some github diff issue.
Anyway there isn't anything special that we should worry about :) It's just about not using _empty_ and make decoding array and object a bit more consistent!
There was a problem hiding this comment.
If you check out the code the diffs is quite a bit different around these lines so github seems to be broken. I'd review in your favorite git tool instead.
|
What's the state here? |
|
I will merge it this or next week |
6 participants
This follows #1836 and uses empty properties instead of empty. That improves symmetry of the
json_encodeandjson_decode.