Can be any of the following values:

• AUTO_TYPE_COERCION (default)
• ALLOW_TYPE_COERCION
• DISALLOW_TYPE_COERCION

See below for details.

Controls whether sequence bounds are taken into consideration for type assignability.

If false, a DataWriter’s type containing a sequence with a larger maximum length will not be assigned to a DataReader’s type containing a sequence with a smaller maximum length. Since the types are not assignable, the DataReader will not match when type information is available.

If true, a sequence in a DataReader's type can have a maximum length smaller than that of a sequence in a DataWriter's type. The types will be assignable, and the DataReader will match; however, when the length of the sequence in a particular DataWriter's sample is larger than the DataReader's maximum length, that sample is discarded. See "Verifying Sample Consistency: Sample Assignability" in the Core Libraries Extensible Types Guide.

Default: true

Controls whether string bounds are taken into consideration for type assignability.

If false, then a DataWriter’s type containing a string with a larger maximum length will not be assigned to a DataReader’s type containing a string with a smaller maximum length. Since the types are not assignable, the DataReader will not match when type information is available.

If true, then a string in a DataReader’s type can have a maximum length smaller than that of a string in a DataWriter’s type. They are assignable, and the DataReader will match; however, when the length of the string in a particular DataWriter’s sample is larger than the DataReader’s maximum length, that sample is discarded. See "Verifying Sample Consistency: Sample Assignability" in the Core Libraries Extensible Types Guide.

Default: true

Controls whether member names are taken into consideration for type assignability.

If false, types containing members with the same ID and different names are not assignable to each other. Since the types are not assignable, the DataReader will not match when type information is available.

If true, members of a type can change their name while keeping their member ID. For example, MyType and MyTypeSpanish are only assignable if ignore_member_names is true:

struct MyType {

    @id(10) int32 x;

    @id(20) int32 angle;

};

struct MyTypeSpanish {

    @id(10) int32 x;

    @id(20) int32 angulo;

};

Since the types are assignable, the DataReader will match.

Default: false

Controls whether type widening is allowed. A type T2 widens a type T1 when T2 contains required members that are not present in T1.

If a DataReader of T2 sets prevent_type_widening to true, then the DataReader will not be matched with a DataWriter of T1 with fewer members because T1 is not assignable to T2.

If a DataReader of T2 sets prevent_type_widening to false, then the DataReader will match with the DataWriter of T1. The DataReader will assume a value for members in T2 that are not in T1. See "Prevent Type Widening" below.

Default: false

Controls whether type information must be available in order to complete matching between a DataWriter and this DataReader.

If false, matching may occur as long as the type names match. Note that if the types have the same name, but the types are not assignable, DataReaders may fail to deserialize incoming data samples. If force_type_validation is true and no type information is available, then the DataReader will not match.

Default: false

This property can take the following values:

• 0 (default value): Received samples containing a union discriminator value that selects a union member on the DataWriter but not on the DataReader are dropped.
• 1: Received samples containing a union discriminator value that selects a union member on the DataWriter but not on the DataReader are set to the default union value on the DataReader.
  
  Note: The default union value is a union with the discriminator set to select the default member, if one is defined, or otherwise to the lowest value associated with any member. In addition, the value of that member is set to the default value for its corresponding type.
• 2: Received samples containing a union discriminator value that selects a union member on the DataWriter but not on the DataReader preserve the discriminator value and do not select any union member on the DataReader.

Received samples containing a union discriminator value that do not select a union member on the DataWriter always preserve the discriminator value on the DataReader with accept_unknown_union_discriminator set to 1 or 2, unless the union discriminator value is an enumerator which is not valid on the DataReader’s type. In this case, the union is set to its default value.

Assume the following two types:

Publisher Type:

@mutable

union MyUnion switch(int32) {
    case 0:
	int32 m1;
    case 1:
	int16 m2;
    case 2:
	double m3;
}; 

Subscriber Type:

@mutable

union MyUnion switch(int32) {
    case 0: 
	int32 m1;
    case 1:
	int16 m2;
}; 

When this property, accept_unknown_union_discriminator, is set to 0, if the DataWriter sends a union with the discriminator set to 2, the DataReader will fail deserializing the sample and will drop it.

When this property is set to 1, the DataReader application will the set the sample to its default value, which has the discriminator set to 0 and member m1 set to the default value for an int32, which is 0.

When this property is set to 2, the DataReader application will deserialize a sample where the discriminator is set to 2 and where there is no member of the union (m1 or m2) selected.

You can set this property as part of the Property QoS for either the DomainParticipant or the DataReader. If it is set in both the DomainParticipant and DataReader, the value in the DataReader's QoS will be applied.

This functionality is supported both in generated code as well as when using the DynamicData API. However, if you use the value 2, there is no API in DynamicData to obtain the discriminator value if it does not select a union member.

When set to 1, samples containing an unknown enumerator can be successfully deserialized to the default enumeration value. For example, given the following two types:

Publisher Type:

enum MyEnum {
    ONE = 1,
    TWO = 2,
    THREE = 3
};
struct MyType {
    MyEnum m1;
};

Subscriber Type:

enum MyEnum {
    ONE = 1,
    TWO = 2
};
struct MyType {
    MyEnum m1;
};

By default, if the DataWriter sends m1 = THREE, the DataReader cannot deserialize the sample. However if this property is set to 1 then the Subscribing application will receive a sample with m1 = ONE. The default enumeration value is defined as the first declared member of the enumeration.

You can set this property as part of the Property QoS for either the DomainParticipant or the DataReader. If it is set in both the DomainParticipant and DataReader, the value in the DataReader's QoS will be applied.

This functionality is supported both in generated code as well as when using the DynamicData API.

This property has been replaced with ignore_member_names and ignore_enum_literal_names in the TypeConsistencyEnforcementQosPolicy (see 2.3 Type-Consistency Enforcement), but is still supported for compatibility with previous releases. If this property is set, its value supersedes the values in the QosPolicy.