Decoder stops parsing a file if a nested list has a length mismatch

[lcarsos]

Description

I have a toon file that looks something like this:

groups[2]:
  - a: foo
    b[1]{w,x}:
      wince,0
  - a: bar
    b[0]{w,x}:
      wise,1
...

It's odd to me that I get the right error if it's the first entry in the list.

For the second "object" I just get the string "a: bar". Looking through the code it looks like that's because any ToonDecodeError that gets raised in any list gets swallowed on line 783 of decode_list_array that seems to just assume that this must just be some unquoted string.

I have a branch where I changed the type thrown in the array parser to at least not swallow length mismatches (https://github.com/lcarsos/toon-python/tree/fix/nested-array-obj) since otherwise you'd have to reraise through several levels but that doesn't feel like the right fix to me.

I also noticed another case where a nested error was swallowed like this, wanted to get your opinion on how you'd prefer this get tackled before I go off writing code that will never merge.

Reproduction Steps

decode file:

groups[2]:
  - a: foo
    b[1]{w,x}:
      wince,0
  - a: bar
    b[0]{w,x}:
      wise,1

Expected Behavior

Expected thrown error: "Expected 0 rows, but got 1"

Actual Behavior

No errors, but incorrectly parsed:

{
  "groups": [
    {
      "a": "foo",
      "b": [
        { "w": "wince", "x": 0 }
      ]
    },
    "a: bar"
  ]
}

Environment

python 3.14
both 0.9.0 and main

Additional Context

No response