Built by Maven

CSV

CsvDataSet reads and writes CSV dataset documents. Each CSV file represents a table — one CSV file per table.

  • CSV base filename must match table name.
  • Requires a file named table-ordering.txt containing all table names (same as the base file name; case insensitive but best to match case), one per line, sorted in insert order.
  • All data CSV files reside in the same directory as table-ordering.txt.
  • Numeric CSV data file fields must not have trailing spaces before the comma-separator, as it will be interpreted as a String instead of a number.
  • Specify null values as null without quotes.

Sample directory layout for two tables, TEST_TABLE and SECOND_TABLE:

data/
  table-ordering.txt
  TEST_TABLE.csv
  SECOND_TABLE.csv

table-ordering.txt:

TEST_TABLE
SECOND_TABLE

TEST_TABLE.csv — the second row’s COL0 and COL2 are null:

COL0,COL1,COL2
row 0 col 0,row 0 col 1,row 0 col 2
null,row 1 col 1,null

SECOND_TABLE.csv:

COL0,COL1
row 0 col 0,row 0 col 1

Load the directory as a dataset:

IDataSet dataSet = new CsvDataSet(new File("data"));