[SPARK-12904][SQL] Strength reduction for integral and decimal literal comparisons

[rxin]

This pull request implements strength reduction for comparing integral expressions and decimal literals, which is more common now because we switch to parsing fractional literals as decimal types (rather than doubles). I added the rules to the existing DecimalPrecision rule with some refactoring to simplify the control flow. I also moved DecimalPrecision rule into its own file due to the growing size.

[rxin]

i broke the previous monolithic decimal precision rule into 2 parts, and then added integralAndDecimalLiteral

[rxin]

This should subsume #10845

cc @viirya - I refactored the code so it is a lot cleaner, and looked at overflow cases.

[viirya]

LGTM

[SparkQA]

Test build #2447 has finished for PR 10882 at commit d5acd17.

• This patch fails Spark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[SparkQA]

Test build #49932 has finished for PR 10882 at commit d5acd17.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[rxin]

Merging this in master.