Swift - 帮助解密调用堆栈

Question

我想知道崩溃堆栈中是否有任何东西可以帮助我找到“Collection <__ NSCFSet：0x600000986940>在被枚举时发生变异的根”。错误。从我到目前为止所学到的，在枚举时调用save上下文可能会导致这个问题。虽然，我认为使用.main.async会修复它。它不是。我想我需要弄清楚如何实现并发，但首先我想知道问题发生在哪里。

0   CoreFoundation                      0x0000000106c4f6fb __exceptionPreprocess + 331
    1   libobjc.A.dylib                     0x0000000105c72ac5 objc_exception_throw + 48
    2   CoreFoundation                      0x0000000106c4c5cc __NSFastEnumerationMutationHandler + 124
    3   CoreData                            0x00000001065c688f -[NSManagedObjectContext executeFetchRequest:error:] + 2463
    4   ncdrainage                          0x0000000104f18e64 $s10ncdrainage15ApplicationDataC18deleteLocalRecords04listF7DeletedySDyS2SG_tF + 1572
    5   ncdrainage                          0x0000000104f106db $s10ncdrainage15ApplicationDataC15downloadUpdates13finishClosureyySo23UIBackgroundFetchResultVc_tFySo19CKServerChangeTokenCSg_Sbs5Error_pSgtcfU1_ySo14CKRecordZoneIDC_AJ10Foundation0C0VSgSbALtcfU2_ + 1115
    6   ncdrainage                          0x0000000104f2f09b $s10ncdrainage15ApplicationDataC15downloadUpdates13finishClosureyySo23UIBackgroundFetchResultVc_tFySo19CKServerChangeTokenCSg_Sbs5Error_pSgtcfU1_ySo14CKRecordZoneIDC_AJ10Foundation0C0VSgSbALtcfU2_TA + 59
    7   ncdrainage                          0x0000000104f109a4 $sSo14CKRecordZoneIDCSo19CKServerChangeTokenCSg10Foundation4DataVSgSbs5Error_pSgIeggggyg_AbESo6NSDataCSgSbSo7NSErrorCSgIeyByyyyy_TR + 372
    8   CloudKit                            0x00000001054f26a7 -[CKFetchRecordZoneChangesOperation _handleProgressCallback:] + 4864
    9   CloudKit                            0x00000001054ba54f -[CKOperation _handleProgressCallback:completion:] + 61
    10  CloudKit                            0x00000001054b6d3c __21-[CKOperation _start]_block_invoke.370 + 216
    11  CloudKit                            0x00000001055306c5 __68-[CKOperationCallbackManager _performCallbackForOperation:callback:]_block_invoke + 278
    12  libdispatch.dylib                   0x000000010922d725 _dispatch_block_async_invoke2 + 83
    13  libdispatch.dylib                   0x000000010921fdb5 _dispatch_client_callout + 8
    14  libdispatch.dylib                   0x0000000109227225 _dispatch_lane_serial_drain + 778
    15  libdispatch.dylib                   0x0000000109227ed0 _dispatch_lane_invoke + 477
    16  libdispatch.dylib                   0x0000000109231ea3 _dispatch_workloop_worker_thread + 733
    17  libsystem_pthread.dylib             0x0000000109608611 _pthread_wqthread + 421
    18  libsystem_pthread.dylib             0x00000001096083fd start_wqthread + 13

Answer 1

它不像崩溃堆栈那样多，因为它没有。由于没有像NSManagedObjectContextPerform这样的东西，似乎你不能在正确的线程中使用你的上下文。您应该使用提交给NSManagedObjectContext :: perform或NSManagedObjectContext :: performAndWait的块的上下文调用代码。

所以代码应该是这样的：

func deleteObjects(context: NSManagedObjectContext) {
    context.perform {
        let objects = context.fetch(fetchRequest)
        for object in objects {
            context.delete(object)
        }

        context.save()
    }
}

您可以通过启用core data concurrency debugging来在调试中捕获此类问题。