Error handling
Вернуться к: Примеры
This section describes how errors are handled. There are two types of errors:
-
SCA runtime exceptions are those that signal problems in the management of the execution of components, and in the interaction with remote services. These might occur due to network or configuration problems.
-
Business exceptions are those that are defined by the programmer. They extend the PHP Exception class, and are thrown and caught deliberately as part of the business logic.
Handling of Runtime exceptions
There are two types of SCA runtime exception:
-
SCA_RuntimeException - signals a problem found by or perhaps occurring within the SCA runtime. This can be thrown for a variety of reasons, many of which can occur regardless of whether a connection is being made to a local or a remote service: an error in one of the annotations within a component, a missing WSDL or php file, and so on. In the case of Web services, an SCA_RuntimeException can also be thrown if a SoapFault is received from a remote Web service and the fault code in the SoapFault indicates that a retry is unlikely to be successful.
-
SCA_ServiceUnavailableException - this is a subclass of SCA_RuntimeException and signals a problem in connecting to or using a remote service, but one which might succeed if retried. In the case of Web services, this exception is thrown if a SoapFault is received with a fault code that indicates that a retry might be successful.
Handling of Business exceptions
Business exceptions may be defined and thrown by a component in the normal way, regardless of whether the component has been called locally or remotely. The SCA runtime does not catch business exceptions that have been thrown by a component called locally, so they will be returned to a caller in the normal way. If a component has been called via a Web service, on the other hand, the SCA runtime on the service providing end does catch business exceptions, and will ensure these are passed back to the calling end and re-thrown. Assuming that the calling end has a definition of the exception (that is, is able to include a file containing the PHP class defining the exception) the re-thrown exception will contain the same details as the original, so that the getLine() and getFile() methods for example will contain the location where the exception was thrown within the business logic. The exception will be passed in the detail field of a soap fault with a fault code of "Client".
Вернуться к: Примеры