Updated validation/BIEN2_Analytical_DB_overview.docx with Brad's revision

schemas/vegbien.sql: taxonlabel_set_canon_label_id(): Fixed bug where always need to set canon_label_id on insert (because it must be a self-reference rather than NULL if there is no matched_label_id), but wasn't being set when no matched_label_id because the IS DISTINCT FROM check returned false

mappings/VegCore-VegBIEN.csv: decimalLatitude/Longitude->geoscrub input coordinates: Also set to NULL if 0 here, not just for the coordinates linked to the datasource's place instance

sql_io.py: put_table(): Switched back to using run_query_into()'s add_pkey_ option now that it uses sql.add_pkey_or_index() instead of sql.add_pkey()

sql.py: run_query_into(): add_pkey_: Use add_pkey_or_index() instead of add_pkey(). This should prevent errors for input rows which match multiple output rows in VegBIEN.

README.TXT: Data import: After import, removed steps to reinstall tnrs_accepted, tnrs_other because this would overwrite the versions of these staging tables that were used to build the current version of the database. (The backup should store a snapshot of the database's input data as it was used to build the database.) Note that these tables will still be reinstalled when the next import starts.

Added validation/BIEN2_Analytical_DB_overview.docx

inputs/import.stats.xls: Updated import times

inputs/.TNRS/tnrs_*/create.sql: Fixed bug where need to COALESCE the WHERE condition with false to ensure that a boolean value, not NULL, is negated by NOT. Otherwise, tnrs_other's query would exclude rows for which the condition returned NULL rather than false.

schemas/vegbien.sql: make_analytical_db(): Truncate tables before (re-)populating them in case make_analytical_db() has already been run