Support reading list of CSVs with `read_csv`

[ehariri]

Pandas should support reading lists of CSV files which contain the same schema. Currently, the filepath_or_buffer argument of read_csv in Pandas must direct to a single CSV file. This is a common use case that other systems like Bodo, Spark, Dask, ... all support.

Assuming ~/path/to/csv contains part1.csv and part2.csv with the same scheme, we wish to do

pd.read_csv(['~/path/to/csv/part1.csv', '~/path/to/csv/part2.csv'])

rather than read the files individually.

[datapythonista]

I think the details need to be discussed, but the general idea shouldn't be controversial. I guess besides providing a list of files, it'd be nice to be able to provide wildcards (e.g. path/to/csv/part*.csv).

[ehsantn]

Yes, providing wildcards is actually in our wishlist too.