Once one or more mutate operations have been processed, the bulk mutate job will eventually be completed. This means that the job will have a result corresponding to each operation. The types of results for an operation are:
The results of a job are structured so that the each operation can be easily matched to its corresponding result. So for every {@link BulkMutateRequest} in a job, there is a corresponding{@code BulkMutateResult}; For every {@link OperationStream}in a request part, there is a corresponding {@link OperationStreamResult}; And, for every {@link Operation} in an operation stream, thereis an {@link OperationResult}.
Within an operation stream, the operations are processed in atomic batches. So even one faulty operation may cause its neighboring non-faulty operations to fail just because they happened to be processed in the same batch. It is important to keep this batch failure reporting scheme in mind when handling operation failures.
There will be a single {@link ApiException} returnedfor each failing batch of operations, and it will have a similar structure and interpretation to the exceptions returned by the synchronous AdWords API services. This exception is returned in a {@link FailureResult}that corresponds to the first operation of the failing batch.
The exception may contain multiple errors, and each error may have {@code OGNL path}s that identify the faulty operation(s) by index (or position within the batch). The first operation in the batch has an implicit index of 0. The subsequent operations in the batch have a corresponding {@link BatchFailureResult} which explicitlyspecifies the index of the operation within the batch.
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |