Fail to get table constrains of Hive tables with Spark API

Now we can not get table constrains like `not null` with Spark API. It may be the bug of Spark.

Here is the process to re-produce the issue.

We can create the table with constrains with hive.

```
create table notnull_col_table2 (col1 int not null, col2 int);
```

But we may not get the not null constrains with Spark API.

```
def main(argv: Array[String]): Unit = {

val spark = SparkSession.builder()
.master("local")
.config("spark.hadoop.hive.metastore.uris", "thrift://localhost:9083")
.config("spark.sql.catalogImplementation", "hive")
.getOrCreate()

val df = spark.table("db1.notnull_col_table2")

val schema = df.schema
println(schema) # StructType(StructField(col1,IntegerType,true),StructField(col2,IntegerType,true))

val col1_schema = schema.fields(0)
println(col1_schema.nullable) # true
}
```

The issue is about Spark API not retrieving NOT NULL constraints from Hive tables. Start by examining how Spark's Hive integration fetches table metadata, particularly in the Spark SQL Hive client code. Look for where column nullability is determined. The test involves creating a Hive table with a NOT NULL column and checking the schema via spark.table. 'Done' means the nullable field in the schema correctly shows false for the constrained column.

Escrito por el modelo de indexación a partir del texto del issue.