swift - 使用 Swift 2.0 处理 CoreData 中的顺序错误的最佳实践

Question

假设我做了这样的事情:

do {
  try context.executeFetchRequest(..)
  // tweak the results
  // build a relation
  try context.save()
} catch { ... }

最好是为 context.save() 的调用执行第二个 do block ，还是按照上面所示的顺序执行所有 try 调用？

不过，如果我想以不同的方式处理任何潜在的错误，最好将它们作为嵌套的 do block 来执行。同样，如果您想以相同的方式处理任何错误，则应该连续执行它们。

Answer 1

我的观点是，获取、关系构建和保存 - 是完全不同的事情，彼此是不同的，因此，必须是以不同的方法执行，每种方法都有自己适当的错误处理代码。

像(伪代码):

...
entity = ...
// need to "build some relation" with entity
try {
    call db_update_operation ((context) -> {
        entities = fetch_from_context(context, ... fetch parameters ... )

        connecting = entity.inContext(context)
        for (relationed in entities)
            relationed.entity = connecting
    })
} catch () {
    // dealing with errors, if any
    // can get here within 3 situations:
    //   1. there was error in fetch conditions
    //   2. there was error in relations establishing
    //   3. there was error in context saving
    // each of that must be treated differently, like
    // fetch/save errors should be passed to higher
    // levels of abstraction, while relation-establishment 
    // errors can be processed here
}


func fetch_from_context( context, ... )

    try {
        fetched = context.fetch(..., &error)
    } catch () {
        // dealing with errors
        // only fetch errors should be processed here
    }

end

func db_update_operation( callback )

    context = detauch_context (get_parent_context)

    try {
        callback(context)

        try {
            context.save(&error)
        } catch () {
            // dealing with save (and only save ) errors, if any
        }

    } finally {
        // make sure, that we've relesed created context
        // other errors is not our deal at this point
        ...
    }
end