[ 2374938 ] Naming convention violation (3)

[HolQue]

How to handle leading and trailing blanks inside key names?

JSON:

" param " : "value"

Result:

DotDict({' param ': 'value'})

I have doubts that it makes sense to allow this.

But if leading blanks are part of a key name, than this is a naming convention violation. Because the naming convention tells: Key names have to start with a letter, digit or underscore. And a blank is nothing of them.

Cross check (forbidden character added):

" #param " : "value"

Result:

Invalid key name: " #param ". Key names have to start with a letter, digit or underscore.

Conclusion: The leading blanks are part of the key name, but not involved in the naming convention check. In my opinion this is inconsistent.

[HolQue]

How about blanks inside a key name?

This is accepted:

"par am" : "value"

Shall this be allowed or not?

[namsonx]

Hello Holger,

The fix for the first case " param " : "value" was pushed to stabi branch.

For the second case "par am" : "value", we should discuss with Thomas before making a decision.

Thank you,
Son

[HolQue]

Hi Son,

The error message of

"  param  " : 1,
"p21"       : ${param}

changed from

'The parameter '${param}' is not available!'

to

'Invalid key name: " param ". Key names have to start with a letter, digit or underscore.'

Hmm, both is not nice. My current understanding of the naming convention I have posted here:

#356 (comment)

And I added this content also to the component description. I assume now that this is valid.

The last point of the naming convention is:

But leading and trailing blanks will be removed (and therefore do not cause errors).

Therefore my expectation of the JSONP content above is:

"p21" = 1

As per

#356 (comment)

blanks inside key names are not allowed. Also the naming convention tells that only letters, digits and some assorted special characters are allowed. Blanks are not part of this list of allowed special characters. Therefore this

"par am" : "value"

should not be allowed.

[namsonx]

Hello Holger,

The change for this ticket was pushed to stabi branch.

Thank you,
Son

[HolQue]

Retest successful. Issue can be closed.

Remaining aspects are already covered by #356

[test-fullautomation]

part of 0.13.2