Conformance Rules - Tooltips

[asheem]

Background

Conformance Rules can be difficult to understand.

Feature

We can add tooltips to the rules which provide a concise explanation.

Proposed Solution [Optional]

Conformance Rules	ToolTips
CastingConformanceRule	Conversion of one input data type to another output data type.
ConcatenationConformanceRule	Combine data contents from two or more input columns together in one output column.
DropConformanceRule	Drop (or remove) a column from the input dataset so that it doesn't appear in the output.
LiteralConformanceRule	Place a literal value in a specified output column.
MappingConformanceRule	Lookup and join values from input and output columns to target columns.
NegationConformanceRule	Convert numeric values from positive to negative, and vice versa.
SingleColumnConformanceRule	Place a value from an input column in a struct, where the output column contains the name of the input field as an alias.
SparkSessionConformanceRule	Extract values from the spark session conf_info file for use in the dataset (e.g. version).
UppercaseConformanceRule	Transform input text values to uppercase.

[Zejnilovic]

I like it, but I am still unsure about LiteralConformanceRule and SingleColumnConformanceRule

[GeorgiChochov]

Here's an explanation of what these two rules actually do. @asheem I trust you can improve on my hasty wording:

• LiteralConformanceRule is more about placing a literal value in a specific column, it has no Input column that it relies on.
• SingleColumnConformanceRule just wraps a value in a struct with a single field. The output column specifies where the struct will be placed and the alias specifies what the name of the filed holding the input will be.

For example, the following configuration:

Input Column: single_value
Output Column: conformed_struct
Input Column Alias: value

would result in the following output:

single_value	conformed_struct
"asd"	{ value : "asd"}

[asheem]

Thanks guys, now updated.

[GeorgiChochov]

Place a value from and input column or alias in a struct to an output column.

It's actually not "input column or alias", it is the value of the input column placed as a field in a struct, where the name of the filed is the alias provided:

input_column	output_column
"asd"	{ input_column_alias : "asd"}

[asheem]

Thanks Georgi, now updated.