Providing metadata for ODBC-compliant data

You can use the command line or table tree to supply metadata for ODBC-compliant data.

Supplying metadata for ODBC-compliant data is somewhat different from supplying metadata for flat files. TenUp cannot use spec trees for ODBC-compliant data. Additionally, ODBC-compliant data sources do not have some of the kinds of table metadata that the 1010data Insights Platform provides. For example, while ODBC-compliant data sources are familiar with the idea of a column name, they do not have an equivalent to Insights Platform column labels. Use the options in this section to more precisely control the target table's metadata in the Insights Platform.

To specify a table tree with an ODBC-compliant database as the source:

$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] @[TABLE_TREE_FILE].xml [QUERY]

Table information

By default, if no additional information is provided to TenUp via a table tree, the table name of the target table is the path to the table that was provided in the TenUp command. However, you can provide a table title with the -t command line option:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -t "My Table"
You can also provide both a short description and a long description for the table with the -n and -N options respectively:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -n [SHORT_DESCRIPTION] -N [LONG_DESCRIPTION]
In the 1010data Insights Platform, you can see the information for a table, such as name, title, short description, and long description, by selecting the table and clicking the Information () icon in the Object Manager:

Column information

Much like the table metadata, you can also provide column metadata for each column of a table. By default, column names in the newly-created Insights Platform table will be the same as the column names in the original data source (however, TenUp will force them into an all lower-case format). If you want to change the column names from those of the original data source, use the -h option:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -h [COL1_NAME],[COL2_NAME],...[COLn_NAME]
Additionally, TenUp provides an option for creating column labels. By default, the column label is the same as the column name. Specify new column labels with the -H option:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -H [COL1_HEADING],[COL2_HEADING],...[COLn_HEADING]
You can also provide a description for each column with the -E option:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -E [COL1_DESC],[COL2_DESC],...[COLn_DESC]

Column types and formats

Column data types are simplified in the Insights Platform compared to the ODBC specification. During a normal extract and load process, TenUp makes its best effort to translate both the column types and formats from the data source to Insights Platform equivalents. If you find that this translation isn't meeting the needs of the job, or would simply like to alter the way the system interprets the types and formats, TenUp provides several options.

You can change the type of the data in a given column using the -i option. After -i, provide a comma-separated list of types, as follows:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -i [COL1_TYPE],[COL2_TYPE],...[COLn_TYPE]

See Data types in this guide for a reference of valid column data types in TenUp.

When loading a SQL_BIGINT column, any provided type for -i will take priority over the --use-long-ints switch. For example, if a user specifies -i i (integer), the column will be defined as an integer column, even if --use-long-ints is specified.

The --use-doubles switch loads SQL_BIGINT columns as double precision floating point columns. This is the current default behavior if -i is not used.

For more information about data types in the 1010data Insights Platform, see Data types in the 1010data Reference Manual.

Column display formats in the Insights Platform are controlled with the -f option. You may want to pay special attention to this setting if you have unusually wide or variable-width columns in the original data source:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] -f [COL1_FORMAT],[COL2_FORMAT],...[COLn_FORMAT]

To see the available options for column display formats, see Display formats in the 1010data Reference Manual.

You can override the default column display width with --default-column-width. The width is the number of characters to display in the column (1-100). The default width is 12.

If the contents of a cell exceeds the specified width, the value is truncated. In this case, the number of visible characters is reduced by two, and " >" is appended to the value. For example, for width:4, the string Example would be displayed as Ex >. Clicking on the > displays the full contents of the cell in a separate window.

To set a new column width, use the --default-column-width switch followed by the new [COLUMN_WIDTH], as follows:
$ tenup64 -u [USERNAME] -p [PASSWORD] -C [CONNSTR] [NEW_TABLE] [QUERY] --default-column-width [COLUMN_WIDTH]