What does it do? The statement tt:switch can be used to execute a maximum of one conditional transformations from a set of conditional transformations. A list of cases is specified in tt:switch, where the syntax of each case is formulated by a conditional transformation, namely a subelement tt:$[s-$|d-$]cond. In this case, tt:$[s-$]cond defines a serialization-relevant case, and tt:$[d-$]cond a deserialization-relevant case. Other direct subelements are not possible in tt:switch. In contrast to conditional transformations positioned outside of tt:switch, for a case specifying at least one attribute using, data, or check is not necessary, provided that the content of tt:$[s-$|d-$]cond is not a pattern . In the list of cases, the following must be noted:
Only one serialization-relevant case may be specified in which no attribute using, data, or check is specified. This case is called standard serialization.
A maximum of one deserialization-relevant case may be specified that does not contain a pattern. This case is called standard deserialization.
Serialization Serializations follow these rules: The first serialization-relevant case tt:$[s-$]cond whose explicitly specified prerequisites are met is serialized and the element tt:switch is exited. If the prerequisite is not met for any of the serialization-relevant cases with explicitly specified attributes using, data , or check, the standard serialization is executed (if it exists) and the element tt:switch is exited. If no standard serialization exists, the exception CX_ST_SWITCH_NO_CASE is raised. In successful serializations, exactly one case is executed.
Deserialization Deserializations follow these rules: The first deserialization-relevant case tt:$[d-$]cond that contains a suitable pattern and whose prerequisites are met is executed (the assertions are deserialized and met) and the element tt:switch is exited. If no deserialization-relevant case that contains a pattern can be executed, the system tries to execute a standard deserialization, if it exists. After the execution, the element tt:switch is exited. If the possible prerequisites using, data, or check of the standard deserialization are not met, the exception CX_ST_REF_ACCESS is raised. If no standard deserialization exists, the exception CX_ST_SWITCH_NO_CASE is raised. In successful deserializations, exactly one case is executed.
Latest notes: When case distinctions are programmed, serializations and deserializations should be formulated as far as possible in common cases tt:cond, whereas the directional cases tt:s-cond and tt:d-cond should be used only if no common formulation is possible. NON_V5_HINTS ABAP_HINT_END
ABAP_EXAMPLE_VX5 In serializations, the transformation DEMO_ST_SWITCH1 below sets the attribute size of element Paragraph depending on the value of the data node SIZE to exactly one value of Small, Big, or Medium, where Medium is created in standard serializations. If, for example, the input values for SIZE lie between 10 and 20, the result looks as follows: < Paragraph size='Medium'>Text < /Paragraph> In deserializations, data is used to assign to the data node SIZE a value 8, 16, or 28 for each of the three valid attribute values Small, Medium, and Big. If a different attribute value is specified, the value 12 is assigned to SIZE in standard deserializations and the invalid attribute is skipped with tt:skip. TRNS DEMO_ST_SWITCH1 In the above transformation, serializations and deserializations are handled completely separately and are divided within a single tt:switch element. Since there is no common case for serializations and deserializations, tt:serialize and tt:deserialize and two tt:switch elements can also be used to achieve the same result. In DEMO_ST_SWITCH2, the identifiers s- and d- can be omitted in front of cond and the query about whether SIZE is bound to an ABAP data object must be queried only once: TRNS DEMO_ST_SWITCH2 ABAP_EXAMPLE_END
