Use new sort_map instead of manually specifying the sort order

Added sort_map to sort a map spreadsheet in the standard order

Removed no longer needed join_passthru, because join_union_sort now serves its purpose

Don't generate mappings/for_review/DwC-VegBIEN.specimens.csv because it's a derived map with lots of duplicated mappings for the various DwC versions

mappings/Makefile: Generate DwC-VegBIEN.specimens.csv directly from DwC1-DwC2 and DwC2-VegBIEN mappings by using join_union_sort with header_num=1, rather than via intermediate DwC1-VegBIEN.specimens.csv

union: Added header_num option to select which map's header to use as the output header

Rename join_sort to join_union_sort and have it run union in ignore mode. This will automatically append the joined map when the input map is a derivative of the joined map, such as for NYBG-DwC.

union: Pass through map 0, so that if ignore is set, the input map will still be output. Allow either map's input label to contain the other's input label to enable e.g. appending mappings for an older input version to those for a newer input version.

DwC1-DwC2 mapping: Changed input label to DwC1, which is allowed by the now relaxed label constraints imposed by union

union: Check if two maps can be combined based on whether map 0 column 0 label contains map 1 column 0 label instead of being equal. This allows map 0's input 0 root to contain the datasource name as well as a format that allows it to be combined with a more general map. Added ignore flag to not print an error if column labels don't match.