Unfinished quotes (also in skipped lines) result in an empty CSV with read_csv()

[Rmulet]

Checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of Polars.

Reproducible example

import polars as pl
pl.read_csv(b'#"Comment\nColA\tColB\n1\t2',separator='\t',comment_char='#')

Log output

Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/research/modules/installs/python/python-3.10.2/lib/python3.10/site-packages/polars/io/csv/functions.py", line 366, in read_csv
    df = pl.DataFrame._read_csv(
  File "/research/modules/installs/python/python-3.10.2/lib/python3.10/site-packages/polars/dataframe/frame.py", line 775, in _read_csv
    self._df = PyDataFrame.read_csv(
polars.exceptions.NoDataError: empty CSV

Issue description

Whenever there are unfinished quotes in a file, polars.read_csv() raises an exception and reports that the CSV is empty. This error is not very informative, since the file is not, actually, empty.

This can be particularly misleading when the unfinished quotes are in a commented/skipped line (see example above), since one would expect those to be completely ignored.

I managed to find a workaround by setting quote_char to another character that was not used anywhere else (%), but that's probably not the optimal way to deal with this. Setting quote_char to None did not work.

Expected behavior

The expected behaviour would be that:

a) Unfinished quotes are completely ignored in commented or skipped rows
b) The error message triggered by unfinished quotes elsewhere is informative

Installed versions

--------Version info---------
Polars:              0.19.13
Index type:          UInt32
Platform:            Linux-5.4.0-163-generic-x86_64-with-glibc2.31
Python:              3.10.2 (main, Feb 23 2022, 22:06:35) [GCC 9.3.0]

----Optional dependencies----
adbc_driver_sqlite:  <not installed>
cloudpickle:         2.2.0
connectorx:          <not installed>
deltalake:           <not installed>
fsspec:              2023.10.0
gevent:              <not installed>
matplotlib:          1.5.3
numpy:               1.23.1
openpyxl:            <not installed>
pandas:              2.1.3
pyarrow:             14.0.1
pydantic:            <not installed>
pyiceberg:           <not installed>
pyxlsb:              <not installed>
sqlalchemy:          1.4.32
xlsx2csv:            <not installed>
xlsxwriter:          3.0.3```

</details>

[ritchie46]

It should provide a better error. We should not read the file as this is an ill defined csv file.

[ritchie46]

Ah, I now see the quote is in the comment line. Comment line should not escape fields.