mappings/Makefile: Don't set $(SHELL) to /bin/bash because this is no longer needed

mappings/Makefile: Removed empty VegCSV section. mappings/Makefile's only functionality is now to clean up (sort) the core maps whenever they change and create human-readable maps from them.

mappings/Makefile: Removed no longer used self maps, because the new automapping mechanism does not use them

input.Makefile: Existing maps discovery: Substituted Veg+ for $(via) because it's now only used once

mappings/VegCore-VegBIEN.csv: Changed input column header from VegCore[Veg+] to VegCore because this is more accurate. This is possible now that we're using new automapping scripts that do not require a particular column header.

inputs/*/*/map.csv: Changed _merge to _join everywhere because _merge's (slower) duplicate elimination functionality is not needed (the combined columns do not both contain the same value, so they can simply be concatenated)

inputs/*/*/map.csv: Changed _merge to _join everywhere because _merge's (slower) duplicate elimination functionality is not needed (the combined columns do not both contain the same value, so they can simply be concatenated)

schemas/functions.sql: _label(): Accept params of any type, in order to support types other than text (which come from staging tables that are imported directly from a SQL export). This fixes a bug in SALVIAS.plotMetadata's column-based import.

schemas/functions.sql: _label(): Support NULL labels by not prepending a label

mappings/Veg+-VegCore.csv: Changed output column header from Veg+ to VegCore because this is more accurate. This is possible now that we're using new automapping scripts that do not require a particular column header. Note that this change now requires the map.csvs to use VegCore as their output column header, because otherwise the Veg+ header will get automapped to VegCore. (The header replacing is a feature to support changing the header when the schema of the column's terms changes.)