ok    1  _integer = the rows written
ok    2  the current row is restored after the write
ok    3  CSV round trip: commas and quotes inside strings survive
ok    4  the first line is the header
ok    5  headers 0: no header line, read back with headers 0
ok    6  TSV by the tab keyword
ok    7  TSV by a .tsv name
ok    8  include 1,3 writes two columns
ok    9  cluster output #channel
ok    10  JSON by a .json name, read back by cluster input
ok    11  JSON by the json keyword
ok    12  field | delimiter both ways
ok    13  a second write replaces the file
ok    14  empty cluster: one header line, _integer = the JSON lines read, JSON reads back as 0 rows
ok    15  cluster output through a cluster parameter
ok    16  cluster output: ... declares a cluster named output
ok    17  the statement writes the cluster named output

passed:  17   failed:  0 
