feat: address mocking failures

[goosewobbler]

No description provided.

[github-actions]

Release Preview — @wdio/native-spy 1.0.10

Scope: @wdio/native-spy

This PR is labeled for a patch release.

This PR will trigger the following release when merged:

Packages

Package	Version
@wdio/native-spy	1.0.10

Changelog

@wdio/native-spy wdio-native-spy@v1.0.9 → 1.0.10

Added

• enhance mock function state management and last call tracking (native-spy)
• enhance mock function with original implementation support (native-spy)
• add test for initial implementation retrieval in mock function (native-spy)
• add getMockImplementation method to retrieve current mock implementation (native-spy)

Changed

• remove redundant implementation reset in mock functions (native-spy)
• standardize package.json files array formatting across multiple packages

Fixed

• correct mock return value handling (native-spy)
• improve mock function behavior and state management (native-spy)

Tags

• wdio-native-spy@v1.0.10

---

Updated automatically by ReleaseKit

[greptile-apps]

Greptile Summary

This PR addresses mocking failures by introducing a new @wdio/native-spy package (a lightweight vitest-compatible mock implementation), refactoring the electron-service to use a shared mockFactory.ts with proper Error-serialisation across CDP, adding class mocking support for Electron, and implementing a JavaScript-only Tauri mocking layer via window.__wdio_mocks__ and invoke interception. CI workflows are also updated with a new vscode-server-action version and revised archive actions.

Confidence Score: 4/5

Safe to merge after addressing the debug console logs and clarifying mock-reuse reset behaviour in the Tauri service.

The core P0/P1 issues from previous review rounds (Error serialisation, mockReturnThis chaining, implementationFn dispatch precedence) appear addressed. Remaining findings are all P2: debug TEST log noise in production init, double-round-trip inefficiency in resetAllMocks, and a mock-reuse semantic divergence from electron.

packages/tauri-service/src/commands/mock.ts (double-reset + reuse semantics), packages/tauri-plugin/guest-js/index.ts (debug log cleanup)

Important Files Changed

Filename	Overview
packages/native-spy/src/mock.ts	New standalone vitest-compatible mock implementation. Dispatch-order issue (defaultRejectedValue before implementationFn) persists from prior threads. mockReturnValueOnce/mockResolvedValueOnce correctly no longer clear implementationFn. mockRestored branch correctly sits after all value branches.
packages/tauri-service/src/mock.ts	New Tauri mock factory. mockRejectedValue/mockRejectedValueOnce now correctly serialize Error objects. mockReturnThis correctly returns mock for chaining. Core logic sound but shares double-reset issue with resetAllMocks.
packages/tauri-service/src/commands/mock.ts	New command handlers for mock, clearAllMocks, resetAllMocks, restoreAllMocks. Two issues: resetAllMocks/restoreAllMocks double-reset inner mocks (N+1 browser round-trips), and mock() resets existing mock state on reuse, diverging from electron-service behaviour.
packages/electron-service/src/mockFactory.ts	New shared mock method factory for electron. mockRejectedValue/mockRejectedValueOnce now correctly handle Error objects for both api and prototype accessor kinds. Clean refactor from duplicated per-mock code.
packages/tauri-plugin/guest-js/index.ts	Plugin frontend API with invoke interception and console forwarding. Two debug TEST: console calls left in production init(). Init messages are intentionally emitted twice (before and after forwarding is active).
e2e/test/electron/mocking.spec.ts	Comprehensive electron mocking e2e tests. The beforeEach clipboard setup in restoreAllMocks suite has been correctly re-added, addressing the previous flakiness concern.
packages/native-spy/test/spy.spec.ts	Good test coverage for fn(). Tests for mockRestore, mockReturnValue/mockResolvedValue/mockRejectedValue after restore, serialization, and cross-mock invocation order are all present.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    subgraph Electron["Electron Service"]
        EM[electron.mock / mockAll] --> MF[mockFactory.ts buildMockMethods]
        MF --> CDP[CDP bridge browser.electron.execute]
        CDP --> IM_E[Inner Mock native-spy fn]
        MF --> OM_E[Outer Mock native-spy fn]
    end
    subgraph Tauri["Tauri Service"]
        TM[tauri.mock] --> CM[createMock tauri mock.ts]
        CM --> TE[tauriExecute browser.execute]
        TE --> WM[window.__wdio_mocks__ native-spy fn]
        CM --> OM_T[Outer Mock native-spy fn]
    end
    subgraph TauriPlugin["Tauri Plugin guest-js"]
        INIT[init] --> II[setupInvokeInterception Wraps core.invoke]
        II --> WM
        INIT --> SPY[window.__wdio_spy__ native-spy]
        SPY --> WM
    end
    subgraph NativeSpy["@wdio/native-spy"]
        FN[fn] --> DISPATCH{Dispatch}
        DISPATCH -->|queue| IQ[implementationQueue]
        DISPATCH -->|rejected| DRJ[defaultRejectedValue]
        DISPATCH -->|impl| IFN[implementationFn]
        DISPATCH -->|resolved| DRS[defaultResolvedValue]
        DISPATCH -->|this| RT[returnThis]
        DISPATCH -->|return| DRV[defaultReturnValue]
        DISPATCH -->|restored| OF[originalFn]
    end
    IM_E --> FN
    WM --> FN
    OM_E --> FN
    OM_T --> FN

Loading

_{Reviews (13): Last reviewed commit: "fix(tauri): clear stale mocks at session..." | Re-trigger Greptile}