Improve performance of getting values by field name

[yruslan]

Background

EBCDIC in-place processing relies on accessing individual fields by name using the parsed copybook. But each such access requires traversal of AST and search for the field name. This is inefficient. The performance can be improved by caching/memoization of the mapping between field names and its reference to the field definition.

Feature

Improve performance of getting values by field name.

Example [Optional]

--

Proposed Solution [Optional]

--