TypeError: the element must be a dictionary or a list but is of type '<class 'str'>'

[hagfelsh]

So this is quite a puzzle. I've followed the code through scalpl.py's setdefault() function and I'm still not sure why this is happening.

The short of it is "why is this happening and what am I missing"...

I hope this reproduces for you as easily as it does for me!

Some environment details in case that's relevant somehow:

• Linux Mint 20
• Pycharm Pro 2021.1.2 (latest)
• venv python version 3.8.10
• scalpl version 0.4.2

Here's the traceback:

I've also attached a screenshot of the pycharm debug window describing exception.

Traceback (most recent call last):
  File "/home/keiser/PycharmProjects/vmagic_web/venv/lib/python3.8/site-packages/scalpl/scalpl.py", line 252, in setdefault
    return item[last_key]
TypeError: string indices must be integers

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/keiser/.config/JetBrains/PyCharm2021.1/scratches/scalpl_problem.py", line 62, in parsevmx
    vmx_dict.setdefault(keystr, value)
  File "/home/keiser/PycharmProjects/vmagic_web/venv/lib/python3.8/site-packages/scalpl/scalpl.py", line 259, in setdefault
    raise type_error(last_key, path, item)
  File "/home/keiser/PycharmProjects/vmagic_web/venv/lib/python3.8/site-packages/scalpl/scalpl.py", line 37, in type_error
    raise TypeError(
TypeError: Cannot access key 'amd' in path 'cpuid.80000001.eax.amd': the element must be a dictionary or a list but is of type '<class 'str'>'.
python-BaseException

Here's the script that's generating this problem.

The idea is to parse the source file (pasted below) and return a dictionary of its contents.
https://pastebin.com/Qm94aWLA

Here's the source file the script is designed to work against:

It's worth noting that this problem doesn't happen on every file like this. Many of them work just fine.
Problem example:
https://pastebin.com/fLSrcAyA

Working example:
https://pastebin.com/E9m4NS27

I created two vimdiff html files to compare these two. One examines the files as-is and the other examines a sorted version of them. In order to upload them to github, I've appended .txt to the filename--just trim that off and you can view them in a browser easily enough. Or diff them with your own tools of course!

sorted_vmx_differences.html.txt
vmx_differences_unsorted.html.txt

The most significant difference between working and non-working appears to involve this block of text. Perhaps something here is too dirty to handle as-is? I tried replacing the '-' characters in right side of the = with 'X' but that made no difference.

cpuid.80000001.eax = "--------------------------------"
cpuid.80000001.ebx = "--------------------------------"
cpuid.80000001.ecx = "--------------------------------"
cpuid.80000001.edx = "-----------H--------------------"
cpuid.80000001.eax.amd = "--------------------------------"
cpuid.80000001.ebx.amd = "--------------------------------"
cpuid.80000001.ecx.amd = "--------------------------------"
cpuid.80000001.edx.amd = "-----------H--------------------"

By the way--scalpl is a delight to use, thank you for sharing it. I use it all the time :)