Reserved words

[rob-ross]

The JSON spec explicitly defines its literal values as:

3. Values

A JSON value MUST be an object, array, number, or string, or one of
the following three literal names:
   false
   null
   true
The literal names MUST be lowercase.  No other literal names are
allowed.

I'm curious then why you are allowing False, Null, and True. What is the use-case for this?

Also:
nil, Nil, none, None

What is the use-case for allowing these synonyms for null, and with multiple initial capitalizationd? I don't have insight into what "most people" are doing with JSON Path; are these common in other implementations in other languages?

Finally,

undefined, missing

How are these being used in queries?

Thanks!

[jg-rp]

I'm curious then why you are allowing False, Null, and True. What is the use-case for this?

We allow True, False and None because that's what Python uses and Python developers might expect them to work. Maybe not the best choice of default configuration, but we're stuck with it until version 2.

Also: nil, Nil, none, None

nil is a Ruby thing. Ruby idioms have leaked into the template engine I mentioned in another thread, so it's important that I have the option of supporting nil as an alias for null.

Finally, undefined, missing

Explicit comparison to undefined (a bit like JavaScript's undefined, aliased as missing) is there for historical reasons. It predates RFC 9535 existence/non-existence for test-expr.

$.some[?(@.thing == undefined)]

is equivalent to

$.some[?(@.thing)]

[rob-ross]

re : $.some[?(@.thing == undefined)] is equivalent to $.some[?(@.thing)]

Did you mean @.thing != undefined? Assuming you want to select the value for the "thing" property in the some object?

[jg-rp]

Yes, sorry.

$.some[?(@.thing == undefined)] would be the same as $.some[?([email protected])]

$.some[?(@.thing != undefined)] would be the same as $.some[?(@.thing)]