Closes#740
`CompactBlockDownloading` protocol is renamed to `BlockDownloaderService`. `CompactBlockDownloader` is renamed to `BlockDownloaderServiceImpl`.
What is the motivation for this rename?
Goal of this change is to extract code from `CompactBlockProcessor` which is related to downloading. So naturaly it should be called "downloader".
But `CompactBlockDownloader` was already taken. After analysis of `CompactBlockDownloading` and `CompactBlockDownloader` I found out that
`CompactBlockDownloader` doesn't download blocks exactly. It's more like abstraction over `LightWalletService`. And it does very similar job to
`LightWalletService`. That is why I decided to rename it to "service".
What is the motivation for protoco/impl naming scheme?
First `BlockDownloaderService` protocol doesn't describe capability in my eyes. It describe what something is. Also `BlockDownloaderServicing` sounds
strange. I agree that add suffix "Impl" to class name isn't super nice. But it is acceptable. And it makes sense to use nicer type name
(e.g. `BlockDownloaderService`) for protocol because it's used on much more places in the code.
- Added `BlockDownloader` protocol and `BlockDownloaderImpl` class. The code related to downloading was exctracted to this class.
- `id` property of transcations is no longer Optional. `id` column in DB
is primary key and it can't be nil. This was oversight in previous
changes.
- Sorting order of transactions is switched from asc to desc. asc is
oversight in previous changes. Android is using desc.
- Using `IFNULL` when sorting transactions by minedHeight.
- Update tests accordingly. And revert few changes on darkside tests.
These changes are no longer needed because using `desc` sorting order
on transactions.
- Fix all the tests. In few tests code had to be changes because order
of transactions is newly defined. It wasn't before.
- Change properties of Transaction.* objects. Some had to be made
Optional to make all tests work.
- Add new tests for TransactionDAO.
- Removed methods from TransactionDAO that weren't used.
- Removed `Transaction` struct and `TransactionEntity` protocol.
- Removed some other protocols from `TransactionEntity.swift`.
Use correct default fee and memo count
Closes#677
- Previously when keychain wanted to do wipe wallet had to do it on it's
own. Which isn't good. Now wallet can simply call
`synchronizer.wipe()` and it's done.
- This change brings support for wipe. `Synchronizer` has new method
`wipe()` which takes care of everything. All the database connections
are closed. Databases are removed. `InternalSyncProgress` is reset.
Change condition when wipe can run
Closes#657
- Now SDK downloads N downloads and scan those. And repeat until sync is
done. This safes lot of space of disk used to store downloaded blocks.
- Most changes are done in `CompactBlockProcessor.processNewBlocks`.
- `SyncRanges` structure is changed a bit.
- `SyncStatus` is changed. SDK now report that it's syncing with
progress. It doesn't report each phase of the sync process separately.
Also appropriate notifications were updated.
Address review comments
- Small changes
- Add tests to check computation of processing ranges for single loop
- Previously we had one range for each sync which was used for each
phase of sync process. Newly there is separate range for each phase of
the sync process.
- I added `InternalSyncProgress` utility. This utility tracks progress
of some phases. And it is able to compute right ranges which should be
used for syncing.
- Some unused download code from `CompactBlockProcessor` is removed.
Fix tests
Address review comments
- Rebase to master
- Update how range for `processBatchFinished()` is computed.
- Refactor `InternalSyncProgress`
- Add tests for `InternalSyncProgress`
Address review comments
Change how latest downloaded block is tracked
- Cache DB is no longer used to track which block was downloaded as
latest. `InternalSyncProgress` is used to do that.
- Thanks to that #660 is fixed. And cache DB can be completely removed
after sync process.
- Added sleep(1) to some darkside tests when latest block height is set.
Otherwise lightwalletd in darkside mode doesn't have time to setup and
tests are flaky.
Fix TransactionEnhancementTests.testBasicEnhancement test
There's a problem withstanding:
````
▿ RustWeldingError
▿ genericError : 1 element
- message : "Error while fetching transparent balances for AccountId(0): Invalid column type Null at index: 0, name: address"
````
Which is fixed by commit `d0297bff0cdaaff42ad26f3be5bc261f6c828e8f` of "https://github.com/zcash/librustzcash".
nil-coalesce thrown errors on test so it reaches the end.
Make balance calls throwing to show underlying errors but nil-coalesce in non-throwing APIs
Add missing fulfill() calls
Add changelog entry
document and handle errors on getting shielded balance before
attempting to download sapling parameters.
Closes#488
- When sync process is finished cache DB is removed from disk.
- Latest downloaded block is preserved becuase it's required to computes
in which phase to resume sync process.
- `SDKSynchronizer.latestDownloadedHeight` public API is removed
it makes no longer sense. This API was reading from cache DB.
state for this synchronizer.
also reports the SychronizerState on synchonizerStarted notification.
Add tests that verify the state being reported properly
Fix `BlockBatchValidationTests`
`XCAsyncTestCase` does not work as intended but `wait {}` does what
we need.
restored `needsToStartScanningWhenStopped` as a Bool var on actor
fixed Enhancement test
- Test was failling with:
generalError(message: "block not found in cache, should always be in cache in darkside mode")
- There was no wait so lightwalletd in darkside mode didn't have enough time to accept mocked state.
- Sample app refactored for the processor being an actor
- tests refactored as well
- dark side tests fixed
- utilities separated to new file
- synchronizer's start and stop are no longer in async context
- updating the UI for the scan fixed
- blocking API removed, only latestBlockHeight() stayed
- non-blocking closure based API removed
- unit tests updated to use async API
[#492] Get rid of blocking API (#551)
- forgotten commented code cleaned up
- some comments were still mentioning result (completion closure), removed
[#534] Use UnifiedSpendingKey for shielding and Spending
This commit implements the use of Unified Spending Keys for shielding
and spending as well as rolling Unified Addresses.
Users should obtain addresses by rolling them from the SDK.
USKs replace Sapling Extended Spending keys and TransparentAccountPrivKeys
when shielding or spending
Closes#534
Co-authored-by: Kris Nuttycombe <kris@nutty.land>
* Fix rebase issues
* PR Suggestion. Make `lastError` an Optional
* Fix test `testReOrgRemovesOutboundTxAndIsNeverMined`
Co-authored-by: Kris Nuttycombe <kris@nutty.land>
- sendToAddress and SaplingParameterDownloader refactored to async
- unit tests WIP
[486] sendToAddress async/await
- testing code cleanup
[#486] sendToAddress async/await
- reducing number of Tasks in Sample app
[#486] sendToAddress async/await
- I finally figure out how to make the dark side tests async, using new sendToAddress APIs and pass, POC done, 18 more tests to refactor
[#486] sendToAddress async/await
- async sendToAddress unit test refactored except NetworkUpgradeTests
[#486] sendToAddress async/await
- NetworkUpgradeTests refactored to async API
[#486] sendToAddress async/await
- cleanup
[#486] sendToAddress async/await (#536)
- PR comments resolved
Issue Reported:
When the synchronizer is stopped, the processor does not cancel
the download correctly. Then when attempting to resume sync, the
synchronizer is not on .stopped and can't be resumed
this doesn't appear to happen in master branch that uses
structured concurrency for operations.
Fix:
This commit makes sure that the download streamer checks cancelation
before processing any block, or getting called back to report progress
Michal Fousek
Lukas Korba
Dimitris Apostolou
Francisco Gindre
Francisco Gindre
Lukas Korba