What additional rules should be supported?

[ross-spencer]

Out of the examples investigated so far, ImageMagick Convert output suggests the need for combining INDEX and GOTO, e.g. here you would ideally access more complex values in the map at position zero, i.e. [0]["image"]["version"] STARTSWITH "ImageMagick":

[
    {
        "version": "1.0",
        "image": {
            "name": "static/images/JSON_logo-crockford.png",
            "permissions": 664,
            "format": "PNG",
            "formatDescription": "Portable Network Graphics",
            "mimeType": "image/png",
            "class": "DirectClass",
            "geometry": {
                "width": 1076,
                "height": 1052,
                "x": 0,
                "y": 0
            },
            "resolution": {
                "x": 118.11,
                "y": 118.11
            },
            "printSize": {
                "x": 9.11015,
                "y": 8.90695
            },
            "units": "PixelsPerCentimeter",
            "type": "GrayscaleAlpha",
            "endianness": "Undefined",
            "colorspace": "Gray",
            "depth": 8,
            "baseDepth": 8,
            "channelDepth": {
                "alpha": 1,
                "gray": 8
            },
            "pixels": 1131952,
            "channelStatistics": {
                "alpha": {
                    "min": 255,
                    "max": 0,
                    "mean": 89.7469,
                    "standardDeviation": 121.782,
                    "kurtosis": -1.61559,
                    "skewness": -0.620008,
                    "entropy": 0.935796
                },
                "gray": {
                    "min": 0,
                    "max": 255,
                    "mean": 138.901,
                    "standardDeviation": 80.5287,
                    "kurtosis": -1.58598,
                    "skewness": -0.120706,
                    "entropy": 0.709949
                }
            },
            "alpha": "#DCDCDC00",
            "renderingIntent": "Undefined",
            "gamma": 0.454545,
            "backgroundColor": "#1A1A1AFF",
            "borderColor": "#DFDFDFFF",
            "matteColor": "#BDBDBDFF",
            "transparentColor": "#00000000",
            "interlace": "None",
            "intensity": "Undefined",
            "compose": "Over",
            "pageGeometry": {
                "width": 1076,
                "height": 1052,
                "x": 0,
                "y": 0
            },
            "dispose": "Undefined",
            "iterations": 0,
            "compression": "Zip",
            "orientation": "Undefined",
            "properties": {
                "Comment": "Created with GIMP",
                "date:create": "2025-04-13T21:22:56+00:00",
                "date:modify": "2025-04-13T21:22:56+00:00",
                "date:timestamp": "2025-05-04T12:52:55+00:00",
                "icc:copyright": "Public Domain",
                "icc:description": "GIMP built-in D65 Grayscale with sRGB TRC",
                "icc:manufacturer": "GIMP",
                "icc:model": "D65 Grayscale with sRGB TRC",
                "png:bKGD": "chunk was found (see Background color, above)",
                "png:iCCP": "chunk was found",
                "png:IHDR.bit-depth-orig": "8",
                "png:IHDR.bit_depth": "8",
                "png:IHDR.color-type-orig": "4",
                "png:IHDR.color_type": "4 (GrayAlpha)",
                "png:IHDR.interlace_method": "0 (Not interlaced)",
                "png:IHDR.width,height": "1076, 1052",
                "png:pHYs": "x_res=11811, y_res=11811, units=1",
                "png:text": "1 tEXt/zTXt/iTXt chunks were found",
                "png:tIME": "2025-04-13T21:22:55Z",
                "signature": "354520c0c68f2980688fb4420c898110fd66484258d6e6775cdc022f5e8b868b"
            },
            "profiles": {
                "icc": {
                    "length": 544
                }
            },
            "artifacts": {
                "filename": "static/images/JSON_logo-crockford.png"
            },
            "tainted": false,
            "filesize": "60174B",
            "numberPixels": "1.13195M",
            "pixelsPerSecond": "75.1801MB",
            "userTime": "0.010u",
            "elapsedTime": "0:01.015",
            "version": "ImageMagick 6.9.12-98 Q16 x86_64 18038 https://legacy.imagemagick.org"
        }
    }
]

What other rules might be useful?

[ross-spencer]

Additionally, a fix in e099577 suggests that nested GOTO is also easily possible, but needs to be well documented, i.e. GOTO statements would always need to be last in marker listings. Alternatively, a more explicit syntax for nested GOTO could be used as well.

NB. nested goto would enable traversing maps to a fairly arbitrary depth.

[ross-spencer]

Sublime Text Project files might benefit from supporting index, if we take the most basic version of that object:

{
	"folders":
	[
		{
			"path": "/tmp/sublimetext-workspace"
		}
	]
}

We might want to say, INDEX[0] ISTYPE dict. Or some other interesting variant where we access the path key.

[ross-spencer]

CryEngine DBATable looks as follows:

{
  "entries": [
    {
      "path": "Animations/animations.dba",
      "condition": {
        "inFolder": {
          "path": "animations/motusAnims/"
        }
      }
    }
  ]
}

{
  "entries": [
    {
      "path": "Animations/human/human_animations.dba",
      "condition": {
        "inFolder": {
          "path": "animations/human"
        }
      }
    }
  ]
}

We can find examples in GitHub: https://github.com/search?q=DBATable.json&type=code

Examples:

• https://github.com/tec-sx/DoxD/blob/c2b94294f0838261c0dfa14e53c062008c8b3cc2/Assets/Animations/DBATable.json#L3
• https://github.com/Battledrake/CryShooter/blob/42184444606e1dfd41c963c87f8d16bfef8656a1/Assets/Animations/DBATable.json

Just searching for "entries" and istype "list" seems like it might be a bit loose, ideally we would look for multiple path objects in the list to make the ruleset more robust.

[ross-spencer]

CSL-JSON uses just a map with an array of "items" and it is only the item's objects that have required fields according to the schema:

NB. example from Zotero.org.

{
  "items": [
    {
      "id": "20895184/FVJZPTDB",
      "type": "book",
      "title": "test",
      "number-of-pages": "10",
      "URL": "https://example.com",
      "author": [
        {
          "family": "test",
          "given": " test"
        }
      ],
      "issued": {
        "date-parts": [
          [
            1970,
            1,
            1
          ]
        ]
      }
    }
  ]
}

The CSL-JSON documentation lists other sources for csl-json we might want to compare with:

[ross-spencer]

Asciicast v3 has an embedded structure: https://docs.asciinema.org/manual/asciicast/v3/

E.g. a minimum header: {"version": 3, "term": {"cols": 80, "rows": 24}}

Ideally, we read a dict at "term" and then check for "colls" and "rows" with arbitrary values there (or indeed, type == int.

NB. there's a way to unlock this already using a form of chaining rule-sets but it's not tested well enough and may lead to undefined behaviors without some ability to step into, and step out of an existing structure.