REGR: prefer user-provided mode

[twoertwein]

follow-up to #39253

• closes BUG: V1.2 DataFrame.to_csv() fails to write a file with codecs #39247
• tests added / passed
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

Make sure that the user-provided mode is always preferred (if it contains 't' or 'b') and extend list of text-only classes.

[jreback]

why is t here? is this tested?

[twoertwein]

"t" is used to explicitly request text-mode but it usually omitted. One of read_json or to_json is using it. I will add a test with a handle that needs that.

[jreback]

kk

[jreback]

@twoertwein ping when ready here (merge master as well)

[twoertwein]

I assume that the occasional errors in test_server_and_default_headers can be fixed with pytest --dist=loadfile?

[jreback]

I assume that the occasional errors in test_server_and_default_headers can be fixed with pytest --dist=loadfile?

yeah that makes sense. separate PR though

[jreback]

thanks @twoertwein

[jreback]

@meeseeksdev backport 1.2.x

[lumberbot-app]

Something went wrong ... Please have a look at my logs.

[roeps]

Perhaps I'm misusing the API, but this seems to have broken the ability to write gzipped JSON. I haven't fully walked the code path, but here is what I'm seeing:

Reproduction:

from io import BytesIO
import pandas as pd

dataframe = pd.DataFrame([1, 2, 3], columns=['a'])
object_stream = BytesIO()
dataframe.to_json(object_stream, compression='gzip', orient='records', lines=True)

Result:

        if path_or_buf is not None:
            # apply compression and byte/text conversion
            with get_handle(
                path_or_buf, "wt", compression=compression, storage_options=storage_options
            ) as handles:
>               handles.handle.write(s)
E               TypeError: a bytes-like object is required, not 'str'

../../.env/lib/python3.7/site-packages/pandas/io/json/_json.py:105: TypeError

---

From what I can tell, _json's get_handle uses a static 'wt' mode which now, with the latest changes to common's _is_binary_mode, will always return false rather than check the path_or_buf instance type, thereby omitting the b flag on the mode passed to the rest of get_handle.

Is this expected behavior/am I misusing the Pandas API? Not the reproduction code works with Pandas 1.2.1