#678 Add the ability to generate Spark schema based on strict integral precision

[yruslan]

New option:

.option("strict_integral_precision", "true")

An example copybook:

      01  R.
                03 SEG-ID        PIC X(1).
                03 SEG1.
                  05 NUM1        PIC 9(2).
                03 SEG2 REDEFINES SEG1.
                  05 NUM2        PIC S9(9).
                03 SEG3 REDEFINES SEG1.
                  05 NUM3        PIC S9(15).

A schema without the option:

root
 |-- SEG_ID: string (nullable = true)
 |-- SEG1: struct (nullable = true)
 |    |-- NUM1: integer (nullable = true)
 |-- SEG2: struct (nullable = true)
 |    |-- NUM2: integer (nullable = true)
 |-- SEG3: struct (nullable = true)
 |    |-- NUM3: long (nullable = true)

A schema with the option:

root
 |-- SEG_ID: string (nullable = true)
 |-- SEG1: struct (nullable = true)
 |    |-- NUM1: decimal(2,0) (nullable = true)
 |-- SEG2: struct (nullable = true)
 |    |-- NUM2: decimal(9,0) (nullable = true)
 |-- SEG3: struct (nullable = true)
 |    |-- NUM3: decimal(15,0) (nullable = true)

(note integer and long are replaced with decimals with scale=0)

[github-actions]

JaCoCo code coverage report - 'cobol-parser'

File	Coverage [85.72%]	🍏
ParserVisitor.scala	89.47%	🍏
ANTLRParser.scala	85.79%	🍏
CopybookParser.scala	83.75%	🍏
CobolSchema.scala	81.56%	🍏
RecordExtractors.scala	78.7%	🍏
DecoderSelector.scala	68.7%	🍏

Total Project Coverage	86.99%	🍏

[github-actions]

JaCoCo code coverage report - 'spark-cobol'

File	Coverage [86.16%]	🍏
CobolSchema.scala	89.79%	🍏
CobolParametersParser.scala	85.2%	🍏

Total Project Coverage	80.77%	🍏