`maxDepth` on `jsonValue` not working

I’m running v 2.23.0

When running this:

fc.assert(
    fc.property(fc.json({maxDepth: 10, depthFactor: 100}), (json) => {
        console.log(json);
        testPoolifyUnpoolify(json as JsonValue);
    }),
    {verbose: true, seed: 1},
);

I always get this output no matter what maxDepth or depthFactor is. Same when I use fc.jsonValue or fc.jsonObject(different output, but same thing, the output is the same)

When I leave out seed, I get different outputs everytime, but the maxDepth of the object never changes (even when I set maxDepth and depthFactor to 1000

[1.1605637550189104e+65,-1.8423363825974171e-155]
[null,null]
false
"zi"
["jx<gAFMtP",null,false,"*-evzweV-*",-2.9e-322]
{"*+s)%%":-3.0615445798526363e-137,">}D5Pk<=(":true,"":-8.916095377805842e-269,"<_":false}
{"Rc>gV%j&N~":true,"$$\"#$  / ":null,"":null,"#":null,"x]@I":null}
-4.1525241057849684e-106
null
["ei#5F3"]
[null,null,null,null,null]
false
[null]
"5N| G@VQ.."
{"":true,"D\\d.d":false,"&'":true,"+l3 )<$*":-2.85409194801039e-191,"zAtfZ&":"hQln"}
[-1.0653066974430237e+248,9.454767484148118e+268,-1.7653965860088798e-214,-3.607621003530778e-129,-4.852957352547658e-136]
["?E:cR",false,"c4}-hv%yW",false,-2.238093053332995e-192]
null
[false,true]
{"5y":null,"y":null}
true
[null,-8.4137512129510665e+245,"<h|xw"]
[true,-3.6772501579408783e-138,false]
[null]
false
{"vM{^^":" ImdH","2fx*":-1.6e-322,"":-1.7976931348623143e+308,"qHH6Gk":7.157927070759342e+43,"@Y0xs":null}
{}
{"e*f&z<4":7.037047740509628e-132}
[")+~","zd&q%","ALFk\\#c*KV"]
["rC","Gq)`('Hs_",3.025804617711035e+178,"k&n",true]
1.7976931348623113e+308
null
{"X)j":2.1629205585163314e-231,"N{o4a ":"yB:","u/9`":null,"{0":-1.9217347146390265e+113,"oFf*Gus":null}
"}"
[false,true,true,true]
true
["|u%a}8!z0#","{K","q"]
[]
false
{}
null
null
[true,false]
-3.603629137412196e-265
{"_zK":null,"AIo<4$:A":"SFq^+i7","#%&9eK!wO":false}
true
["5@#n0|","O1$",""]
{"":-2.9782890422455327e+255," j!0":-1.0479956480311591e-162,"fH(.a":2.758176663043067e+92,"}(jKG":-1.902033828458498e-126,"8w'E":1.2732807500211005e+179}
{"#|eA|9}h!>":true}
"$-(F)"
3e-323
[-1.8746904761410538e-298]
false
[null,null,null,null,null]
[".T%wwfpg:y"]
[null]
{"Nf&5":2.8873533988267525e-170,"":true,"Xc\"":null,"_ >vSeh":"*G?","=OawKi7":null}
[]
{"OC[MOJw%":null}
"ges(dk("
{">Eg&Gdv\"":2.2881157822979433e+226,"{]>Pj":133020.52617428786,"`knOi{ ":1.2632862668499098e-71,"":5.3318996347400214e+72}
-3.411537840694293e-11
"`^O7pV"
"ttx;X@N"
3.0691830158402806e-155
{"%*EH{. -":true}
[null,null]
1.1454945138091173e-46
null
true
{"\"":"leM r~*"}
{"~!D@L@gR":"le","0yO":")?&|Rl","FNsb~*\\7;N":"1{\\+Zi"}
null
["w<-%Ra"]
[]
-3.098772027733863e-58
{}
[]
false
null
null
["+Mb^[OOwf"]
{"bk":-4.0175676216709717e-243,"z`F2":-7.159008907172276e+93,"":null,"w0fr\",":null,"2KcAl=":null}
[false,true,true,true,false]
"5f"
true
[" jRS[B","8@{Y@Q0AV#","|+B"]
false
{"a{":"A\\/hhKc","FwmbjTo]1>":")N%"}
-1.5e-323
[]
null
"t~h"
true
{"v":true,"n&E@9XQqS":true,"":true}
"b_#=E$N\\a"
{"WJ!":-2.646684589725734e-273,"ls1}`c":"t'iT2b}2","#xOVY<+\"":null,"":null}
{"Jv0WdIGF":" )!3\\lT","!/Yms:":"`jx.X","":"nz","h}t":"#","<":"u^"}
{"q(;=:a)":7.423270140880227e+284,"!":"[WWg?","uU<n":null,"":"y3"}
["ur","U[1&#$SSC_","qjc4Y6Z","y=","UZ>% "]

[dubzzz]

MaxDepth and depthFactor seems to hit an asymptote when I increase the former and decrease the latter- is this a known issue? https://runkit.com/stevenqzhang/fast-check-maxdepth-testing

This is a known issue due to:

• 33% chance to have basic values
• 33% chance to have arrays with half of them with basic values
• 33% chance to have objects with half of them with basic values

Which is definitely something that must change. But not sure I can really change such thing in a minot release as it would have potential terrible impacts on some existing test suites on user side. But there is definitely something to do for v3 (coming soon).

depthFactor? — default: 0 — factor to change the probability to generate leaf values as we go deeper in the structure, numeric value >=0 (eg.: 0.1). The closer the number is to 0, the higher the probability of generating leaf values

Definitely, like it. Feel free to open a PR with the suggested update. By the way, there probably other references to this depthFactor in the md file, they can all be adapted with the same idea.

[dubzzz]

Closing the issue as there is no activity on it. Feel free to re-open it or open another one referencing this one if you believe there is still something to check.