Upload CSV to Database (Trino) is broken after #28192

[devlatte]

Bug description

The "Upload CSV to Database" feature for Trino is not working properly after the changes introduced in #28192.

This issue did not exist in version 4.0.2, where the feature was functioning correctly.

Screenshots/recordings

Go to the "Upload CSV to Database" feature in the Superset UI.
Select Trino as the target database.
Attempt to upload a CSV file.
Observe the error or malfunction.

If Table Already Exists: Fail

Traceback (most recent call last):
  File "/app/superset/commands/database/uploaders/base.py", line 114, in _dataframe_to_database
    database.db_engine_spec.df_to_sql(
  File "/app/superset/db_engine_specs/trino.py", line 515, in df_to_sql
    if database.has_table_by_name(table.table, table.schema):
AttributeError: 'Database' object has no attribute 'has_table_by_name'


If Table Already Exists: Replace

Traceback (most recent call last):
  File "/app/superset/commands/database/uploaders/base.py", line 114, in _dataframe_to_database
    database.db_engine_spec.df_to_sql(
  File "/app/superset/db_engine_specs/trino.py", line 546, in df_to_sql
    location=upload_to_s3(
  File "/app/superset/db_engine_specs/hive.py", line 85, in upload_to_s3
    bucket = s3.Bucket(bucket_path)
  File "/usr/local/lib/python3.10/site-packages/botocore/client.py", line 922, in __getattr__
    raise AttributeError(
AttributeError: 'S3' object has no attribute 'Bucket'```



### Superset version

4.1.1

### Python version

3.10

### Node version

16

### Browser

Chrome

### Additional context

This functionality was working correctly in version 4.0.2.
The regression seems to have been introduced with the changes in [#28192](https://github.com/apache/superset/pull/28192).

I attempted to use the "Replace" setting as a workaround, but it also fails. 
The issue appears to be caused by hive.py relying on an outdated version of boto3.

### Checklist

- [X] I have searched Superset docs and Slack and didn't find a solution to my problem.
- [X] I have searched the GitHub issue tracker and didn't find a similar bug report.
- [X] I have checked Superset's logs for errors and if I found a relevant Python stacktrace, I included it here as text in the "additional context" section.

[dosubot]

Hey @devlatte! I'm a bot designed to help you with bug fixes, troubleshooting, and contributions. I'm here to assist you while waiting for a human maintainer. Let's squash some bugs together!

I found a similar unsolved discussion regarding Trino connection issues for CSV uploads in Superset version 4.1.0rc2. You can view the discussion here ^[1].

Regarding the errors you're encountering, it seems there are two main issues:

1. Missing has_table_by_name Method: The error indicates that the Database object lacks the has_table_by_name method. This might be due to changes introduced in pull request feat: new Columnar upload form and API #28192. You may need to check if this method was removed or renamed in the recent updates.

2. S3 Bucket Attribute Error: The error 'S3' object has no attribute 'Bucket' suggests a potential issue with the boto3 library version or its usage in hive.py. The upload_to_s3 function uses boto3 to interact with S3, and there might be compatibility issues with the version of boto3 you are using. Ensure that your boto3 library is up-to-date and compatible with Python 3.10. Additionally, verify that the AWS credentials and permissions are correctly configured ^[2].

You might want to investigate these areas further to resolve the issues.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[rusackas]

Pinging @dpgaspar since he's the author of #28192