/**
* Alphanumeric postcodes post-processing script ensures that both the expanded
* and contracted version of alphanumeric postcodes are indexed.
*
* Without this script a postcode such as '1383GN' would not be matched to the
* query '1383'.
*
* The script is intended to detect these alphanumeric postcodes and index both
* permutations, ie. '1383GN' = ['1383GN', '1383 GN'].
*
* The inverse case should also be covered. ie. '1383 GN' = ['1383 GN', '1383GN'].
*
* Note: the regex is currently restrictive by design, the UK for instance uses
* alphanumeric postcodes in the format 'E81DN' which could cause error when splitting
* with this method, they are currently ignored. Future work should consider global
* postcode formats.
*
* Note: this script is intended to run *before* the 'deduplication' post processing
* script so that prior aliases don't generate duplicate terms.
*/