Mishandling of wildcard characters in read_parquet

[Malkard]

Describe the bug
When trying to read a single file containing wildcard characters (in our case, '['), the read_parquet function fails to find the file:

wr.s3.read_parquet(path="s3://bucket/some_file[something].parquet")

results in a NoFilesFound exception

To Reproduce
Version 2.4.0
Probably exists since the introduction of wildcards with #322

To reproduce: Put a parquet file containing square brackets in a bucket and try to read it.

The implementation should accept a preceding '' to indicate the literalness of a special character.

Workaround
Replace wildcards with '?' before passing to read_parquet.

[maxispeicher]

As a quickfix you could just use glob.escape() before passing the path:

import glob
import awswrangler as wr

s3_path = "s3://bucket/some_file[something].parquet"
escaped_s3_path = glob.escape(s3_path)
wr.s3.read_parquet(path=escaped_s3_path)

[Malkard]

Thanks, this works too, and is somewhat cleaner.

Since the intent of the "path" argument of read_parquet is to handle wildcards, this could be considered as the proper way to deal with this use case, instead of my suggestion. In which case a simple mention in the documentation might be valuable.

[igorborgest]

Released on version 2.5.0!

[jtlz2]

This still doesn't work for me (REF: https://stackoverflow.com/questions/75614048/how-do-i-merge-several-parquet-files-into-one-using-awswrangler) - what am I doing wrong? Thanks!