Loading ...
Sorry, an error occurred while loading the content.

JSON.org grammar

Expand Messages
  • Peter Michaux
    I think the grammar on the right of json.org would be more complete if it started with json_text object array I remember having trouble finding out whether or
    Message 1 of 4 , Aug 26, 2008
    • 0 Attachment
      I think the grammar on the right of json.org would be more complete if
      it started with

      json_text
      object
      array

      I remember having trouble finding out whether or not just a number,
      for example, counted as valid JSON when I started learning about JSON.

      Peter
    • John Cowan
      ... RFC 4627 is definitive. That said, I think it s unfortunate that a a bare number, string, truth value, or null doesn t count as a JSON text. Why was this
      Message 2 of 4 , Aug 26, 2008
      • 0 Attachment
        Peter Michaux scripsit:

        > I remember having trouble finding out whether or not just a number,
        > for example, counted as valid JSON when I started learning about JSON.

        RFC 4627 is definitive. That said, I think it's unfortunate that a
        a bare number, string, truth value, or null doesn't count as a JSON text.
        Why was this limitation imposed?

        --
        No, John. I want formats that are actually John Cowan
        useful, rather than over-featured megaliths that http://www.ccil.org/~cowan
        address all questions by piling on ridiculous cowan@...
        internal links in forms which are hideously
        over-complex. --Simon St. Laurent on xml-dev
      • Tatu Saloranta
        On a related (?) note: I have heard that comments were allowed at some point (i.e. were proposed as part of formal json grammar). If so, why were they dropped
        Message 3 of 4 , Aug 26, 2008
        • 0 Attachment
          On a related (?) note: I have heard that comments were allowed at some
          point (i.e. were proposed as part of formal json grammar). If so, why
          were they dropped (simplicity?).
          Worse: json.org's default parser implementation seems to support
          multiple extensions, which leads developers to assume these are part
          of json, not just that extra features implementation has. :-/

          The reason I am asking this is that I have gotten multiple user
          requests to support comments, and while ideally I would want to stick
          with the standard format, it is hard to argue against supporting
          majority of content found in t he wild (significant portion of which
          allegedly contains such comments -- comments are good places to stick
          debug-info about generator/app that created content, when it was
          created etc).

          -+ Tatu +-

          On Tue, Aug 26, 2008 at 7:45 AM, John Cowan <cowan@...> wrote:
          > Peter Michaux scripsit:
          >
          >> I remember having trouble finding out whether or not just a number,
          >> for example, counted as valid JSON when I started learning about JSON.
          >
          > RFC 4627 is definitive. That said, I think it's unfortunate that a
          > a bare number, string, truth value, or null doesn't count as a JSON text.
          > Why was this limitation imposed?
          >
          > --
          > No, John. I want formats that are actually John Cowan
          > useful, rather than over-featured megaliths that http://www.ccil.org/~cowan
          > address all questions by piling on ridiculous cowan@...
          > internal links in forms which are hideously
          > over-complex. --Simon St. Laurent on xml-dev
          >
          > ------------------------------------
          >
          > Yahoo! Groups Links
          >
          >
          >
          >
        • Douglas Crockford
          ... I have removed comments from the reference implementation.
          Message 4 of 4 , Aug 26, 2008
          • 0 Attachment
            --- In json@yahoogroups.com, "Tatu Saloranta" <tsaloranta@...> wrote:
            >
            > On a related (?) note: I have heard that comments were allowed at some
            > point (i.e. were proposed as part of formal json grammar). If so, why
            > were they dropped (simplicity?).
            > Worse: json.org's default parser implementation seems to support
            > multiple extensions, which leads developers to assume these are part
            > of json, not just that extra features implementation has. :-/
            >
            > The reason I am asking this is that I have gotten multiple user
            > requests to support comments, and while ideally I would want to stick
            > with the standard format, it is hard to argue against supporting
            > majority of content found in t he wild (significant portion of which
            > allegedly contains such comments -- comments are good places to stick
            > debug-info about generator/app that created content, when it was
            > created etc).


            I have removed comments from the reference implementation.
          Your message has been successfully submitted and would be delivered to recipients shortly.