Mirrors prod's TK2 wire flow: the first arriver (parked, picks up cached
pair on a later poll) gets matching_state 3007 (SUCCEEDED_OWNER); the
second arriver (whose poll triggered the pair) gets 3004 (SUCCEEDED).
Observationally inert in the public matching code path today — the
client's Matching class writes isOwner from the response into a field
that nothing in TK2/ranked reads. Matching_Room (private rooms) DOES
read it but from a separate code path that doesn't consult our response.
We send the split anyway for prod fidelity and to leave room for future
flows (rematch UI, etc.) that might start consuming it.
TryPairAsync now returns PairUpResult(Match, IsOwner) instead of bare
PendingMatch?, so the controller can decide owner vs joiner without
re-deriving it.
Also documents on DoMatchingResponseDto why we omit prod's `room_id`
field (not in the client's DoMatchingDetail model; private-room flows
get their room id from a different API and don't consult this response).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Two client-crash bugs in the do_matching response when no partner is
waiting:
1. matching_state was 3001 (RC_BATTLE_MATCHING_ILLEGAL); the client's
Matching.OnFinishedDoMatching switch maps that to an error dialog,
not a retry. The retry state is 3002 (RC_BATTLE_MATCHING_RETRY).
2. node_server_url was omitted entirely. The client's
DoMatchingBase.SettingDoMatchingData reads it via
data["node_server_url"].ToString() with no Keys.Contains guard, so
absence throws KeyNotFoundException out of NetworkManager.Connect
before the matching_state switch is even reached. Prod RETRY
captures send "" while waiting and the real URL only on SUCCEEDED;
match that.
battle_id stays absent; its accessor IS guarded.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Solo pollers park (3001 RETRY); two concurrent pollers pair and both
receive 3004 + same BattleId. Cache hits on the first arriver's next
poll. ?scripted=1 retains today's solo Scripted path for dev work.
Response DTO's BattleId/NodeServerUrl become nullable so 3001 omits
them on the wire (WhenWritingNull policy drops them).
ASP.NET's default bool binder rejects "1" as a value, so the scripted
opt-in is bound as string? and parsed permissively (accepts "1" and
"true"/"True"/etc.) rather than relying on built-in bool binding.
IMatchingBridge.RegisterPendingBattle now takes a MatchContext; PendingBattle
carries it; BattleSession stores it. ArenaTwoPickBattleController builds ctx
from IMatchContextBuilder. ScriptedLifecycle still uses ScriptedProfiles for
the player half — Tasks 5/6 migrate the lifecycle.
Existing tests updated: MatchingBridgeTests, BattleNodeFlowTests,
InMemoryBattleSessionStoreTests, BattleSessionDispatchTests, BattleSession
PumpTests, ArenaTwoPickBattleControllerTests (which now seeds a TK2 run +
adds a no-active-run 400 case).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
GameStart already detects the Steam-vs-UDID mismatch produced by
wipe-and-resignup; it now also reclaims the orphan. New
ViewerRepository.MergeAnonymousViewerInto transfers the fresh UDID
from V_new onto V_old in one save (freeing the unique-index slot),
then deletes V_new in a second save. Partial-failure mode is a
benign null-UDID viewer; two rows never contend for the same UDID.
Side benefit: future GetViewerByUdid lookups now short-circuit to
V_old without going through the Steam handler.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Wipe-and-resignup left the client stuck with the blank V_new's id in
Certification.ViewerId. /tool/signup is anonymous, so it can't see the
Steam ticket and creates a fresh anonymous viewer keyed on the new UDID;
the Steam handler on the next request resolves to V_old and serves its
data, but no normal-response hook overwrites Certification.ViewerId.
GameStart now compares the UDID-keyed viewer to the auth-resolved one
and emits rewrite_viewer_id when they differ, which Cute/GameStartCheckTask
writes back into Certification.ViewerId.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Prod do_matching captures (data_dumps/captures/traffic_prod_tk2_*) send
the node URL as host:port/socket.io/ with no scheme prefix —
e.g. "node06.shadowverse.jp:13560/socket.io/". BestHTTP's SocketManager
expects this exact shape; the leading ws:// we were sending plus the
missing /socket.io/ path was preventing the client from completing the
post-do_matching connect (eventually times out with "connection timed
out").
Update BattleNodeOptions default, Program.cs override, and both
controller and bridge tests to use "localhost:5148/socket.io/".
Discovered during v1 smoke walkthrough.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The decompiled client's DoMatchingBase.SettingCardMasterId calls
jsonData["card_master_id"].ToInt() with no Keys.Contains guard when
matching_state ∈ {3004, 3007, 3011}. Omitting the field crashes the
client with KeyNotFoundException at Cute.NetworkManager+Connect.
Add CardMasterId to DoMatchingResponseDto with a default value of 1
(matching the /load/index response and prod captures). Extend the
controller test to assert the field is present.
Caught during the v1 smoke walk-through; full client log line:
[Error: Unity Log] KeyNotFoundException: The given key was not
present in the dictionary.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
BattlePassRepository._curveCache and MissionCatalogRepository._maxLevelCache
were private-static fields populated lazily on first read from whatever
DbContext happened to be in scope. In production "one DbContext lineage
per process" makes that fine. Under parallel test execution each
SVSimTestFactory owns its own SQLite :memory: DB, so the first reader's
DB (often empty, in tests that don't seed BP) poisoned the cache for
concurrent readers from a seeded DB — assertions like "BP level info
must be present after seeding" failed because the process-static cache
returned an empty list populated by the other test's empty DB.
The first patch attempted a `BypassCacheForTests` static flag, which is
exactly the kind of test-only seam that rots the production code: future
caches get the same flag, repos accumulate hidden knobs, and the
underlying invariant ("a cache populated from arbitrary scope serves
arbitrary scope") goes unaddressed.
Instead, move both caches into the DI-registered IMemoryCache.
AddMemoryCache() registers it as singleton-per-service-provider:
production has one provider → one IMemoryCache → identical caching
semantics to before. Each WebApplicationFactory builds its own
provider → its own IMemoryCache → cache is naturally scoped per fixture,
no cross-test bleed possible.
The ResetLevelCurveCache() method and its three call sites
(SVSimTestFactory.SeedGlobalsAsync, BattlePassServiceTests,
LoadControllerTests) are deleted — a fresh factory owns a fresh empty
cache, no manual invalidation needed.
With this and the previous StoryService fixture-instance fix in place,
ParallelScope.All works: 776/776 in 57s wall clock (down from 59s on
Fixtures, 2m13s pre-parallelism).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The full-catalog regression test hardcoded "35 active packs as of
2026-05-23" but the controller filters by DateTime.UtcNow against each
pack's commence/complete dates. When two packs (99047, 80047) crossed
their complete_date of 2026-06-01 01:59:59 UTC, the test started
failing with Expected: 35 / But was: 33 — which had been masked all
along by NUnit's trx serializer OOMing on a different test.
The hardcoded count conflated three things that happened to be equal
on the day the test was written: packs in the seed file, packs active
right now, and 35. The test's real intent (per its class docstring) is
"every pack the importer ingests round-trips through /pack/info";
pinning the clock with TimeProvider would solve today's drift but
re-break the moment someone regenerates the seed or retires a pack.
Expected count now derives from the seed file at test time, filtered
by the same predicate the controller uses (PackRepository
.GetActivePacks: IsEnabled && commence <= now <= complete) via the
shared ImporterBase.ParseWireDateTime parser so any date-string quirk
parses identically on both sides. Spot-check on pack 99047 swapped for
"any pack with non-default pack_category" — same schema-fidelity
coverage (non-zero category survives JSON round trip) without pinning
to an id that rotates.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
MessagePack [Key("...")]-keyed contracts reject unknown fields, so request
DTOs that omit BaseRequest's envelope (viewer_id, steam_id,
steam_session_ticket) fail deserialization on the real msgpack wire path.
Routing smoke + JSON-direct tests didn't catch this because S.T.J. tolerates
extra keys and the routing smoke uses ValidBaseRequestJson, but anything
sent via the actual client encrypted=True path threw
MessagePackSerializationException.
Fix: every Arena*Request now inherits BaseRequest. Also updates the JSON
controller tests + e2e to include the envelope so the [ApiController]
auto-400 validation passes.
Discovered via /arena_colosseum/get_fee_info crash on the in-game arena
screen.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Mirror of the outer-repo data_dumps/ reorganization (commit e1e595d in
the SVSim outer repo): updates all data_dumps/extract/ → data_dumps/scripts/,
data_dumps/client_master_csv → data_dumps/client-assets, data_dumps/traffic
→ data_dumps/captures/traffic in XML doc-comments and inline comments
across importers, controllers, middlewares, DTOs, and tests. Doc-only;
no logic changes; build green.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Prod's /pack/get_gacha_point_rewards offers leader cards from packs
where the leader sits in a non-Legendary tier — UCL pack 16015 has
Kyoka (711531010, Runecraft) and Miyako (711331010, Dragoncraft) as
Gold-tier rows with is_leader=1 in the drawrates. The old filter
(Tier == Legendary && !IsAltArt) excluded them, so the in-game
exchange UI was empty despite the banner advertising leader-card draw
rates.
Fix: filter on (Tier == Legendary || IsLeader) && !IsAltArt. Captures
every legendary plus any leader card regardless of page tier. Verified
against the captured 16015 response in
traffic_prod_all_gacha_exchange.ndjson (28 entries: 26 legendaries +
2 Gold-tier leaders).
Across the seeded data this surfaces 6 additional cards: 3 Bronze-tier
leaders + 3 Gold-tier leaders. The 68 Legendary-tier and 81 Special-
tier leaders were already included.
Renames legendaryCardIds -> exchangeableCardIds for clarity.
Regression test seeds a Gold-tier IsLeader=true card with a Skin row
and asserts the exchange catalog returns it with the Skin reward
entry.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The two wire fields differ for seasonal packs (verified against
traffic_prod_all_gacha_exchange.ndjson — every captured request pairs
odds_gacha_id=16xxx with parent_gacha_id=10xxx). The OLD DTO docstring
assumed they were always equal; today's controller used
ParentGachaId, which lands on the base/family pack id (often a
synthesized disabled stub with no GachaPointConfig) and returns [].
Fix:
- GetGachaPointRewards and ExchangeGachaPoint now consume OddsGachaId.
- Update both DTO docstrings to document the seasonal-pack pattern.
- Regression test seeds (16015 enabled w/ GachaPointConfig, 10015
disabled stub w/o config) and asserts the response uses 16015's
catalog.
Symptom: opening pack 16015 (parent_gacha_id=16015 in /pack/open)
accrued gacha points correctly, but /pack/get_gacha_point_rewards with
{odds_gacha_id:16015, parent_gacha_id:10015} returned an empty list.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Extends /pack/open beyond the v1 CRYSTAL_MULTI=2 / DAILY=3 / RUPY_MULTI=7
trio to cover every type_detail whose payment primitive already exists:
1 CRYSTAL -> ICurrencySpendService crystal debit
6 RUPY -> ICurrencySpendService rupee debit
4 TICKET / 5 TICKET_MULTI -> debit child.ItemId from OwnedItemEntry
(ticketsNeeded = cost * packNumber), 400 on
missing/short balance; reward_list gets a
RewardType=4 post-state Item entry to mirror
project_wire_reward_list_post_state
Skin-overload type_details (8/9/13) and free-pack overlays (10/11/12)
stay 501 — they need selection / banner plumbing the current code
doesn't have.
Tutorial alias unchanged: it still consumes the gating ticket post-draw
and stamps tutorial_step=100. The two ticket flows diverge by intent
(tutorial = free server-grant; normal = paid by ticket inventory).
Removed Open_rejects_ticket_type_detail (asserted the old 501 path);
covered by Open_rejects_insufficient_tickets. Updated
NonTutorial_pack_open_does_not_emit_tutorial_step to assert the new
200-on-ticket-success behavior — same invariant under test.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bootstrap Program.cs now calls PackDrawTableImporter after PackImporter.
Delete DbCardPoolProvider, ICardPoolProvider, and the DbCardPoolProvider
tests — the new IPackDrawTableRepository covers what GachaPointService
needed (legendary-tier card_ids per pack) and PackOpenService takes the
draw table directly.
GachaPointService now resolves the legendary catalog from
PackDrawTable.CardWeights filtered by Tier==Legendary, instead of
ICardPoolProvider.GetPool then a rarity filter. Same end set, no DB pool
walk.
Test fallout: tests that fabricate custom card sets for gacha-point
tests now call factory.SeedPackDrawTableFromSetAsync(packId, setId)
to install a matching legendary-tier stub. Full suite: 647/647 green.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Sampler is now driven by PackDrawTable: roll DrawTier per slot by
cumulative slot-rate weights, then pick a card within tier by per-card
weights renormalized within the tier. Rate-less Guaranteed-Leader-Card
rows draw uniform over (pool minus owned), falling back to the full
pool when all are owned. Bonus slot fires once at the end of a 10-pack
open when HasBonusSlot is set.
Slot 8 falls back to the general slot's per-card weights for the rolled
tier when slot-8 has only a rarity-level rate quoted (the common shape
on normal packs).
PackController.Open loads the draw table + viewer owned card ids and
passes them to the sampler; the category-based forced-Legendary slot-8
override is gone. ICardFoilLookup replaces ICardPoolProvider for the
foil-twin heuristic.
Drops the test-fixture pack-draw seed overlay so the production seed
flows through the importer tests; controller tests that fabricate their
own card sets now call factory.SeedPackDrawTableAsync(...) to install a
matching stub draw table.
WeightedPick helper handles the cumulative-band roll for both stages.
Five sampler tests + four WeightedPick tests + five importer/repo
tests; full suite is 653/653 green.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
A real /load/index dump emits my_rotation_id as a bare number (0) for
unset MyRotation slots, which 400'd against the string? DTO field
(AllowReadingFromString only covers string->number). FlexibleStringConverter
accepts either form. Also skip empty deck slots (no cards) on import — a
dump carries every slot, mostly empty placeholders the client manages
itself.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Extends POST /admin/import_viewer to accept owned_cards (snapshot full-replace).
Unknown card_ids are skipped, counted in skipped_card_count on the response, and
logged server-side. Count is clamped to OwnedCardEntry.MaxCopies (3). Also injects
ILogger into AdminController and adds Cards/Items/Decks to the viewer reload graph.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Route /load/index currency, owned-card list, and cosmetic id-lists through
IViewerEntitlements so freeplay mode inflates all three without touching
the viewer's DB state. Adds LoadControllerFreeplayTests (2 tests).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
practice/deck_list returns the same wire shape as /deck/info (the client parses
both via DeckGroupListData), but only ever sent user decks — so a fresh account
saw no default decks and couldn't start a practice match.
Extract the /deck/info hydration into a shared IDeckListBuilder used by
/deck/info, /deck/my_list, and /practice/deck_list. Practice passes
padEmptySlots:false (deck *select*, not builder) — matches the prod practice
capture, which returns real decks unpadded plus the 8 per-class default decks
and per-class leader-skin settings. Retire the near-duplicate
PracticeDeckListResponse DTO in favor of the shared DeckListResponse.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Wires IGachaPointService.TryExchangeAsync into a controller endpoint.
Loads the viewer with the full cosmetic-grant graph (Cards, Sleeves,
Emblems, Degrees, LeaderSkins, MyPageBackgrounds) plus the gacha-point
balance + received marker, with AsSplitQuery to avoid the cartesian
explosion documented in project_ef_split_query. Returns BadRequest with
the outcome's error code on validation failure, or the post-state
reward_list on success.
Two integration tests: happy-path verifies card grant + balance debit +
received-marker persistence + post-state reward_list shape; insufficient-
balance path returns 400.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
PackController.Info's ownedItemsByItemId projection used `i.Item.Id` to
key the dict — EF translates that to the FK column today, but any future
model change that breaks the nav→column mapping would fall back to client
eval and collapse every key to 0 (the default Item constructor's Id),
silently hiding every tutorial pack via item_number=0. EF.Property<int>
reads the shadow FK directly and is robust to nav changes.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Reject empty / whitespace / explicit-null / over-cap names with 400
instead of NREing on null assignment or storing arbitrarily-long
strings the DB column has no cap on. 24-char limit is a conservative
backstop against direct API abuse; the client UI enforces its own
keyboard limit.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Three coupled correctness fixes to /tutorial/gift_receive's response:
- received_ids / total_receive_count_list / reward_list are now built
from `toClaim` (the gifts THIS call granted), not from `requestedIds`.
Echoing the client's request meant idempotent re-claims re-fired the
"+N received" popup and direct-assigned the same post-state totals
again, breaking the documented idempotency contract.
- is_unreceived_present is now `unclaimedPresents.Count > 0`. The
hardcoded false hid the inbox badge after partial claims even when
present_list still carried unclaimed gifts.
- tutorial_step echoes the persisted (max-preserved) state instead of
a hardcoded 41. A replay against a state>=41 viewer used to surface
41 on the wire and regress the client's tutorial state machine.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Same project_ef_nav_include_pitfall as 27ebb51's tutorial pack_open fix
but in the gift path: without .ThenInclude(i => i.Item), the existing
OwnedItemEntry's Item nav defaults to a new ItemEntry() (Id=0), so
RewardGrantService.ApplyAsync's `FirstOrDefault(i => i.Item.Id == detailId)`
misses pre-existing rows. It falls through to add a new entry, and the
(ViewerId, ItemId) unique index added 2026-05-25 throws on SaveChanges →
500 to the client, no tutorial advancement, no currency grant.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The tutorial alias bypassed the currency / type_detail / open-count guards
unconditionally. Combined with the unconditional TutorialState=100 write, any
authenticated viewer could send /tutorial/pack_open with any parent_gacha_id
to draw a pack for free and clobber their state down to 100.
Two gates: parent_gacha_id MUST be 99047 (the legendary starter), and the
viewer's TutorialState MUST be below 100. The state write is also max-preserved
as a belt-and-braces backstop. Mirrors the 31→41 guard in GiftController.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The endpoint used to write the client-supplied step verbatim, so a stale or
replayed request with tutorial_step=0 against any later-stage viewer would
regress the persisted state to 0. NextSceneSwitcher routes step==0 to
AreaSelect section 0, which has no chapter data — the client LINQ-Single()
crashes on next /load/index, bricking the viewer. Math.Max-preserve matches
the 31→41 pattern in GiftController.TutorialGiftReceive.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Without .ThenInclude(i => i.Item), the OwnedItemEntry.Item nav defaults to a
new ItemEntry() with Id=0 (project_ef_nav_include_pitfall), so the
FirstOrDefault(i => i.Item.Id == ticketItemId) lookup never matched. The
ticket was never decremented and reward_list omitted the post-state entry —
on the next /tutorial/pack_info the pack stayed visible and the client
re-clicked into plain /pack/open, which 501s on type_detail=5.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Spec at docs/api-spec/endpoints/post-login/download_time-{start,end}.md
already documented both endpoints fully against the decompiled
Wizard/DownloadStartTask.cs and Wizard/DownloadFinishTask.cs — the
controller side was the gap.
The client fires /download_time/start before kicking off an Akamai
asset bundle download and /download_time/end on completion. Both are
pure telemetry from our perspective. When NukeIdentityOnStartup wipes
PlayerPrefs broadly (the pre-narrow loader behaviour), the client
decides it needs to download tutorial assets, calls /download_time/start,
and a 404 there surfaces as an HTTP error popup before the download
proceeds. Stubbing with empty data:{} bodies plus result_code:1 is the
documented minimum-viable response.
Acts as belt-and-suspenders against the narrow IdentityWipe (which
preserves the cache index so downloads shouldn't trigger) ever being
bypassed by a different code path.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The client's PlayerStaticData.UpdateHaveUserGoodsNumByJsonData does direct
assignment on each reward_list entry's reward_num, so currency/item totals
must be the new viewer balance — not the gift delta. Fresh accounts were
seeing their cached crystal/rupy balances clobbered down to the gift counts
until the next /load/index. Matches the project_wire_reward_list_post_state
memory and the prod capture (which shows 120 rupy = baseline 20 + gift 100).
Stack [HttpPost("/tutorial/pack_open")] alias on PackController.Open. Detect
isTutorialPath via HttpContext.Request.Path; gate the type_detail rejection,
currency switch, open-count tracking, and currency reward_list entries behind
!isTutorialPath so the starter legendary pack (99047/990047, type_detail=5)
bypasses the purchasable-pack code path. After grant, set MissionData.TutorialState=100
and emit tutorial_step=100 in PackOpenResponse — this is the sole END transition,
per live-traffic capture. Add pack 99047 to test-fixture packs.json.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Stacks a second [HttpPost("/tutorial/pack_info")] absolute route on
PackController.Info so the tutorial flow resolves to the same action
and returns the same pack_config_list as /pack/info (no filtering in v1).
Adds PackControllerTests.cs with TutorialPackInfo_returns_same_list_as_pack_info
verifying byte-identical responses from both URLs.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Implements POST /account/update_name — writes Viewer.DisplayName and
returns an empty array per the prod capture. Includes TDD test covering
the persist side-effect.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
POST /tutorial/update echoes tutorial_step back and saves it to
Viewer.MissionData.TutorialState. is_skip=1 is handled server-side
by honoring whatever tutorial_step value the client sends (client
already sends 100 when skipping). Adds TutorialUpdateRequest DTO,
TutorialUpdateResponse DTO, injects SVSimDbContext into
TutorialController, and adds GetViewerTutorialStateAsync helper to
SVSimTestFactory.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
