Enable to browse into a directory that contains a broken symlink

[romaingz]

Description

Using the file browser on the left, I cannot browse to a directory that contains a broken symlink.

Reproduce

1. In your home create a directory that contains a broken symlink
2. Go to jupyterlab
3. On the left pane (file browser) double click on that directory
4. The browser does not open the directory

Expected behavior

I expect the directory to open and exclude the broken symlink from the view.

Context

• Operating System and version: CentOS 7
• Browser and version: Chrome 80
• JupyterLab version: 1.2.1

Troubleshoot Output

Here's what my browser shows as a response for the http request (it's actually on a single line, I've replaced \n chars to make it more readable):

{"message": "Unhandled error", "reason": null, "traceback": "Traceback (most recent call last):
  File \"/opt/conda/lib/python3.7/site-packages/tornado/web.py\", line 1699, in _execute
    result = await result
  File \"/opt/conda/lib/python3.7/site-packages/tornado/gen.py\", line 209, in wrapper
    yielded = next(result)
  File \"/opt/conda/lib/python3.7/site-packages/notebook/services/contents/handlers.py\", line 112, in get
    path=path, type=type, format=format, content=content,
  File \"/opt/conda/lib/python3.7/site-packages/notebook/services/contents/filemanager.py\", line 431, in get
    model = self._dir_model(path, content=content)
  File \"/opt/conda/lib/python3.7/site-packages/notebook/services/contents/filemanager.py\", line 338, in _dir_model
    if self.allow_hidden or not is_file_hidden(os_path, stat_res=st):
  File \"/opt/conda/lib/python3.7/site-packages/notebook/utils.py\", line 154, in is_file_hidden_posix
    stat_res = os.stat(abs_path)
PermissionError: [Errno 13] Permission denied: '/path/to/broken_link'
"}

Note that path/to/broken_link is the actual broken link and my issue is about being enable to browse into /path/to.

[vidartf]

Given that the exception is in the content manager (/notebook/services/contents/filemanager.py in trace), I think this issue might be better posted on the jupyter/notebook repo.