Spark Doris Connector 26.1.0

## Improvement

1. Support exposing Doris ARRAY columns as Spark Array[String] (#362
(https://github.com/apache/doris-spark-connector/pull/362))

2. Support Spark 4.1 (#365 (https://github.com/apache/doris-spark-connector/pull/365))
3. Support one-way TLS connections to Doris (#369
(https://github.com/apache/doris-spark-connector/pull/369))

4. Add Doris regression test programs (#370 (https://github.com/apache/doris-spark-connector/pull/370))
5. Support S3 TVF batch sink mode (#371 (https://github.com/apache/doris-spark-connector/pull/371))
6. Use Arrow Flight SQL as the default read mode, with fallback to Thrift when unavailable (#367
(https://github.com/apache/doris-spark-connector/pull/367))

## BugFix

1. Fix backend discovery for the specified compute group in cloud mode (#361
(https://github.com/apache/doris-spark-connector/pull/361))

2. Fix Arrow timestamp conversion for Doris DATETIME and TIMESTAMPTZ types (#366
(https://github.com/apache/doris-spark-connector/pull/366))

# Thanks

@addu390
@gnehil
@JNSimba
@LuciferYang

Review the linked pull requests #362, #365, #369, #370, #371, #367, #361, and #366 to understand the changes included in version 26.1.0. Done means the listed improvements and bug fix are incorporated into the release.

Written by the indexing model from the issue text.