db_xml.py: put(): put_(): Removed no longer needed in_row_ct_ref param, which is only used by put_table(). Rewrapped function body.

sql_io.py: put_table(): ignore(): literals: Only replace invalid literal with NULL or remove row if that column actually contains the invalid value in question. This handles the case where all columns are being ignore()d because the specific column couldn't be identified, and this was not the invalid column.

mappings/VegX-VegBIEN.stems.csv: plot: Mapped note

mappings/VegX-VegBIEN.stems.csv: plot: Added landform mapping

schemas/vegbank.ERD.pdf: Auto-repaired with Adobe Reader so that the repair message doesn't pop up whenever it's opened

schemas: Added vegbank.ERD.pdf so the VegBank ERD is easily accessible when mapping

mappings/VegX-VegBIEN.stems.csv: project: Mapped sourceaccessioncode. This entailed adding a distinguishing suffix to the projectname input mapping.

mappings/DwC2-VegBIEN.specimens.csv, VegX-VegBIEN.stems.csv: Removed all manual mappings to datasource_id now that datasource_id is auto-populated, both on the VegBIEN output side and the DwC/VegX input side. This should greatly simplify many of the mappings!

db_xml.py: put(): Don't suppress exceptions thrown by sql_io.put_table() by passing them to on_error(), because some exceptions indicate unrecoverable database connection problems such as a broken connection, which should abort the import

db_xml.py: put(): Support datasets with no rows, where root.firstChild == None. Documented that to use an entire XML document, you need to pass root.firstChild rather than root.