Upstream proxy double-encodes GET query params. url.ParseQuery only
decodes once, leaving values like %E5%BF%AB instead of actual UTF-8.
Add fullyDecodeURL to repeatedly decode until stable before signing.
RawQuery from GET callbacks can be double-URL-encoded by upstream
proxies. Use Query().Encode() to rebuild from fully decoded params,
ensuring VerifyNotification computes the signature on correct values.
- After opening pay_url in new window, navigate to order status page
with countdown timer and order polling (reuses QRCode view)
- QRCode view now shows "pay in new window" UI when no QR code present
- Add verbose logging for webhook signature verification failures
- EasyPay redirect passes all configured CIDs (cidAlipay, cidWxpay) comma-separated
- Payment redirect opens in new window instead of current window redirect
- Order action column uses styled icon buttons with tooltip
- Pending order limit error returns i18n-friendly message with metadata
- Cancel order rate limit feature (ported from sub2apipay) with rolling/fixed window modes
- Admin settings UI for cancel rate limit configuration
- Fix provider dialog triggering unintended settings save on open
- Provider save now correctly calls saveSettings after provider update
- API client interceptor forwards reason and metadata fields for structured error handling
Backend request/response structs used camelCase JSON tags while the
rest of the codebase uses snake_case (from Ent ORM entities). This
caused field name mismatches — notably the admin refund deduct_balance
field was silently ignored.
Backend: changed all camelCase JSON tags in payment handler/service
structs to snake_case (CreateOrderRequest, CreateOrderResponse,
CreatePlanRequest, UpdatePlanRequest, CreateProviderInstanceRequest,
UpdateProviderInstanceRequest, AdminProcessRefundRequest, RefundResult).
Frontend: updated all API call payloads and response field accesses
to use snake_case, removed manual camelCase conversion in
buildPlanPayload, toggleForSale, provider dialog emit, etc.
- EasyPay webhook: add GET route + read params from URL query (fix 404)
- Stripe: expose as single "stripe" method to users, sub-types (card/link/
alipay/wxpay) passed to PaymentIntent internally via instance config
- QR code page: use order expiresAt for countdown, add cancel button
- InstanceSelection carries SupportedTypes for provider-specific routing
- Fix CreateOrderRequest field name mismatch (payment_type → paymentType)
- Payment page now discovers available types from providers (not global config)
- Backend GetLimits returns map keyed by payment type
- EasyPay shows 3 buttons (跳转/支付宝/微信), Stripe shows card/alipay/wxpay/link
- Auto-select first method after limits loaded
- Skip UpdatePaymentConfig when no payment fields provided (prevents wipe)
- Remove unused defaultMinRechargeAmount/defaultMaxRechargeAmount constants
- Fix mergeConfig comment to match actual behavior
- Backend: payment fields added to GET/PUT /admin/settings (full replace)
- Frontend: single API call for all settings (no separate payment config API)
- Payment page: show "充值未开放" when no payment methods available
- Pending order check when disabling provider
- Backend: add ListProviderInstancesWithConfig that decrypts config
and masks sensitive fields (keys, secrets) as "••••••••"
- Frontend: pre-fill non-sensitive config values when editing provider
(PID, API base URL, notify URL, return URL, channel IDs etc.)
- Sensitive fields left empty for user to re-enter if needed
- Add PurchaseRouter.vue to dynamically render PaymentView or
PurchaseSubscriptionView based on payment_enabled vs
purchase_subscription_enabled settings
- Fix route guard to allow access when either payment system is enabled
- Split sidebar menu: /purchase shows for both systems, /orders only
for built-in payment
- Remove auto-migration that forcibly disabled legacy purchase system
on first startup (MigrateLegacyPurchaseURL)
- Add payment_enabled to PublicSettings API response so frontend can
check both systems from a single endpoint
- Restore legacy purchase link config UI controls in admin Payment tab
- Split payment settings save to use dedicated PUT /admin/payment/config
API instead of mixing into general settings endpoint
- Add admin payment config API functions (getConfig/updateConfig)
- payment_service: split GetDashboardStats (131→35 lines) into 4 sub-functions,
extract applyPagination helper, replace bubble sort with sort.Slice,
use sync.Once for EnsureProviders, define magic number constants
- providers: merge duplicate PC/Mobile functions in alipay/wxpay,
extract loadKeyPair/queryOrderTotalFen in wxpay, add centsToYuan in stripe,
define constants for success codes, currencies, event types
- handlers: extract requireAuth and parseIDParam helpers to eliminate
repeated auth/ID-parsing boilerplate
- registry: remove dead seen map in GetProviderByKey
- types: centralize order status constants in payment package
- load_balancer+config_service: unify duplicated containsType into
exported InstanceSupportsType function
Critical:
- Refund idempotency keys now include UnixNano timestamp (alipay/wxpay)
- PaymentView passes qr/pay_url params to QR code page
- Fix dead code branch in order result handling
Quality:
- Currency symbol $ → ¥ in admin stats and refund dialog
- StripePaymentView setTimeout cleanup on unmount
- Webhook body size limited to 1MB (io.LimitReader)
- SubscriptionPlan features type documented
- Run gofmt on user schema, config test, group handler
- Remove unused mergeGroupIDs function
- Restore shared test helpers (newJSONResponse, queuedHTTPUpstream)
that were in deleted Sora test file
- Add int64(0) param to SelectAccountWithLoadAwareness callers (signature change from channel scheduling refactor)
- Add UsageMapHook type and struct field to StreamingProcessor
- Revert Claude Max cache billing code to upstream/main (not part of channel feature)
- Revert credits overages logic to upstream/main (non-channel change)
- Remove Instructions field reference (non-channel OpenAI feature)
- Restore sora_client_handler_test.go from upstream + add channel service nil params
- Fix 7 stale comments still mentioning "限制检查" in handlers/services
- Make billingModelForRestriction explicitly list channel_mapped case
- Add slog.Warn for error swallowing in ResolveChannelMapping and
needsUpstreamChannelRestrictionCheck
- Document sticky session upstream check exemption
Move the model pricing restriction check from 8 handler entry points
to the account scheduling phase (SelectAccountForModelWithExclusions /
SelectAccountWithLoadAwareness), aligning restriction with billing:
- requested: check original request model against pricing list
- channel_mapped: check channel-mapped model against pricing list
- upstream: per-account check using account-mapped model
Handler layer now only resolves channel mapping (no restriction).
Scheduling layer performs pre-check for requested/channel_mapped,
and per-account filtering for upstream billing source.
Replace response.BadRequest with response.ErrorFrom + infraerrors.BadRequest
to provide machine-readable reason codes (VALIDATION_ERROR, INVALID_CHANNEL_ID,
MISSING_PARAMETER) for frontend i18n support.
- PricingSourceChannel/LiteLLM/Fallback for resolver source
- MediaTypeImage/Video/Prompt for result.MediaType
- Reuse BillingModeToken/BillingModeImage for billing mode
- Reuse BillingModelSourceChannelMapped/PlatformAnthropic in handler
- Backend: reject intervals with all-null price fields on save
- Backend: filterValidIntervals skips empty intervals in pricing resolver
- Frontend: red border + asterisk on empty interval rows
- Backend: antigravity groups now match anthropic/gemini channel pricing
- Fix errcheck: defer rows.Close() with nolint
- Fix errcheck: type assertion with ok check in channel cache
- Fix staticcheck ST1005: lowercase error string
- Fix staticcheck SA5011: nil check cost before use in openai gateway
- Fix gofmt: format chatcompletions_to_responses.go
- Parse candidatesTokensDetails from Gemini API to separate image/text output tokens
- Add image_output_tokens and image_output_cost to usage_log (migration 089)
- Support per-image-token pricing via output_cost_per_image_token from model pricing data
- Channel pricing ImageOutputPrice override works in token billing mode
- Auto-fill image_output_price in channel pricing form from model defaults
- Add "channel_mapped" billing model source as new default (migration 088)
- Bills by model name after channel mapping, before account mapping
- Fix channel cache error TTL sign error (115s → 5s)
- Fix Update channel only invalidating new groups, not removed groups
- Fix frontend model_mapping clearing sending undefined instead of {}
- Credits balance precheck via shared AccountUsageService cache before injection
- Skip credits injection for accounts with insufficient balance
- Don't mark credits exhausted for "exhausted your capacity on this model" 429s
erio