A CSV connector configured to import a binary value fails with:

Change detection engine import all items for connector CSV Connector Example failed with reason Line 1 column 1 value of 010011010110000101101110 is not valid. Duration: 00:00:00.1406250

This was tried with a byte value, byte array value, binary value and base64 encoded string; all resulting with a message similar to the one provided above.

• Either the validator should be updated to handle string values
• Or an intermediate step is included where the CSV connector tries to parse the values retrieved based on the field ValueType, before creating with the Validator
• Otherwise, a more descriptive exception should be provided denoting that this connector doesn't support that value type.

The field "Telphone" is multivalue, and the following line is being imported:

13,Matthew Clark,1337,"0413371337,0412123123",another@address.com

However, the Telephone value is still being sent to FIM as a single valued field:

dn: CN=859cbd1d-fe77-4cd7-b88d-998f12aae491
objectClass: csvPerson
IdBID: 859cbd1d-fe77-4cd7-b88d-998f12aae491
Key: 13
Name: Matthew Clark
EmployeeID: 1337
Telephone: 0413371337,0412123123
Email: another@address.com

The current message thrown when the target CSV file is empty is the following:

Save entities to connector failed.
Save entities [Count:40] to connector Test CSV - C:\log\AurionGeneric.csv failed with reason Non-negative number required.
Parameter name: value. Duration: 00:00:00.0040002
Error details:
System.ArgumentOutOfRangeException: Non-negative number required.
Parameter name: value
at System.IO.FileStream.set_Position(Int64 value)
at Unify.Product.IdentityBroker.CsvConnector.<>c__DisplayClass2a.<AddEntities>b__28(Stream old, Stream new)
at Unify.Product.IdentityBroker.FileStreamCommunicator.Shift(UpdateFile fileUpdate)
at Unify.Product.IdentityBroker.CsvConnector.AddEntities(IEnumerable`1 entities)
at Unify.Product.IdentityBroker.EventNotifierAddingConnectorDecorator.AddEntities(IEnumerable`1 entities)
at Unify.TestHarness.Connector.ConnectorTestHarness.<>c__DisplayClass6.<Add>b__5(IEntitySchema schema, IEnumerable`1 entities)
at Unify.TestHarness.Connector.ConnectorTestHarness.Shift(Guid sourceId, Action`2 entityHandler)
at Unify.TestHarness.Connector.ConnectorTestHarness.Add(Guid sourceId, Guid targetId)
at Unify.TestHarness.Connector.ConnectorTestHarnessServiceController.Add(StandardPostData postData)

Instead, the schema field names should be written as the header, with an informational event logged stating that the file was updated with the correct heading.

When specifying the schema for a CSV file, you cannot change what is the required / readonly fields to generate keys. The CSV file format is not always controllable as it can be system produced. Can you please allow the modification of the schema settings.

Scrolling to the right will rubber-band back to the left for the CSV file path.

There are some display issues with the CSV connector.

• Some label names are still garbled
• Some FTP settings display regardless of the type of communicator (File and FTP)
• A validation message does not display if the file path is cleared
• There is a ")" character after "Ignore Certificate Errors"

The default schema provider for CSVs should mark the key as required.

An additional new line is added for each exported entry.

To allow for the order of the schema to not be important (as it can't be changed), and so that the schema can be manually set up, the CSV connector should look at the header to determine the order of attributes as entries are exported.

The "Path" attribute of the CSV Connector is not a required field, and it is possible to create the connector without specifying one. The connector screen reflects the error "The schema for the CSV Test was not updated for the following reason: The path is not of a legal form. "