[Jsonpath] On @index and similar

Daniel P <danielaparker@gmail.com> Tue, 26 April 2022 13:33 UTC

MIME-Version: 1.0
References: <mailman.110.1650927448.28781.jsonpath@ietf.org>
In-Reply-To: <mailman.110.1650927448.28781.jsonpath@ietf.org>
From: Daniel P <danielaparker@gmail.com>
Date: Tue, 26 Apr 2022 09:33:35 -0400
Message-ID: <CA+mwktLr2SpcAPiMa9GXjb=AhJ=H3f65aULCxzeNm+U6i_qD9w@mail.gmail.com>
To: jsonpath@ietf.org
Content-Type: text/plain; charset="UTF-8"
Archived-At: <https://mailarchive.ietf.org/arch/msg/jsonpath/KvZH7KMOuhXj_xGoNWLEWXdwsiE>
Subject: [Jsonpath] On @index and similar
Precedence: list

> From: Greg Dennis <gregsdennis@yahoo.com>

> I think @foo is too similar to @.foo and will cause problems for human readers.
>
> Regarding functions, I think this syntax will back us into a corner where we can't support a function that takes a parameter, should someone want that sort of thing.
>
> Additionally, how would you get the length of a nested array, e.g. @.foo.length where @.foo is an array?  You really need the traditional function syntax to do this: length(@.foo).
>

Agreed on the unsightliness (and unnecessity, and ad hockery) of the
@foo notation.  Prior experience suggests just going with function
syntax. @index sounds a lot like the XPath fn:position() function, why
not do something similar?

Daniel

[Jsonpath] On @index and similar Daniel P