ZaStoGram_desktop/Telegram/SourceFiles/test/docs/helpers/README.md

11 KiB

Harness helper index

Harness guide · Documentation format

Search this index by module or need, then read only the selected reference and its source header. These are lookup pages, not a required reading bundle. Self-tests exercise the harness; they are not routines every product overlay should call.

Module Kind Use
test_agent API Runtime gate, startup overrides and fixture access
test_runner API Stages, gated groups, bounded readiness and exact-object actions
test_gated_stage Self-test Self-tests for explicit and inherited gated skips
test_session_gate API Test-account gate, switching, restoration and explicit production legs
test_controller_ready API Exact account/session window controllers from the window registry, observation-only readiness with named refusals and one guarded action
test_log API Evidence logs, assertions and telemetry formatting
test_log_lines Self-test Self-test for one physical record per log call
test_text_reads API Space-normalized text assertions
test_lang_pack API Temporary language overrides and restoration
test_l10n_seams Internal dependency Phrase-read observer, app-config overrides and the request holder for the localization capture
test_probe API Windowed observations and correlated requests
test_expiry_window API Served-expiry observations bound to a live owner, generation and link; frozen cutoff windows and boundary polls with named refusals
test_widgets API Typed discovery, a button's own disabled state, live publications and input events
test_wheel Self-test Self-test for wheel delivery
test_queued_drag API Deliver queued drag starts before the synthetic release
test_qpa_mouse API Mouse press, move and release through Qt's platform layer and window routing, for widgets that grab the mouse; a drag stage that waits for queued work before its release and leaves no button, grab or hover
test_qpa_mouse_self_test Self-test Self-test for QPA grab routing, the drag stage's wait and refusals, coordinates, double clicks and clean end state
test_touch API Registered touchscreen and touchpad input through Qt's platform layer
test_text_drop Self-test Self-test for text-drop routing
test_clipboard API Clipboard round-trip and classified read-back
test_console_lock API Current console-lock classification
test_console_shield API Opt-in shield that vetoes and counts console input outside harness delivery, passes product closes and quits, and releases on finish
test_capture API Prepared grabs, paint owners, mapped targets with typed coordinate owners, strict comparisons, checked full-owner crops and saved-image markers
test_mapped_target Self-test Self-test for full mapped-target visibility and typed coordinate owners
test_post_paint API Samples after real product paints, shared capture scopes and paint receipts
test_full_owner_capture Self-test Self-test for coordinated samplers, strict renders, checked full-owner crops and panel sizing
test_animation_clock API Deterministic elapsed-time animation frames
test_paint_stream API Whole-owner QPicture streams and manifest for optimized replay timing
test_corner_patch API Rounded-corner comparisons and controls
test_marking_read API Prevent primary-window read marking
test_window_exposure API Keep covered windows painting without activation
test_layer_root Self-test Self-test for the painting-layer resolver
test_via_window Self-test Self-test for window-mapped captures
test_toast_capture API Isolated toast-subtree captures, shown/hiding/gone toast state with a bounded gone wait, and painted-content captures
test_hover Self-test Self-test for click/drag hover cleanup
test_press_origin Self-test Self-test for press-point ripple origins
test_activation Self-test Self-test for injected window activation
test_box_button API Resolve and click shell footer buttons
test_ink API Painted bands, ink, colour and contrast, glyph cores on flat and gradient grounds, isolated glyph bounds
test_style API Settled palette samples and baseline checks
test_palette API Switch and restore palettes while respecting system following
test_panel API Separate-panel discovery, show-cache readiness and typed fixture sizing
test_menu API Same-turn menu readings, actions, keyboard activation of accessible custom rows, and captures
test_chat_menu API Attribute chat-list context menus to the exact message
test_chat_show API Show a chat-list message wholly in view, painted, ready and hovered before input
test_chat_show_self_test Self-test Self-test for chat-list show acceptance, re-correction and named refusals
test_chat_show_place API Place several chat-list messages wholly in or out of view together, gated by geometry and a fresh visibility reading, and prove they stayed for a window
test_widget_ops API Tap a widget message's validated ops: the --attacks leak scan, its renderer's limit reports, the latest conformance result, every data-stress value with its receipt, the newest host's batch counts, scheme and theme message
test_widget_paint_policy API Run widget-sandbox's paint-policy.json against the heavy-box rule and limit glue Telegram compiles, one row per entry, without the app's state
test_widget_host API Override a widget message's host measurement options and record what its host got and reported: applied options, stats, nodes, load, verbatim death with its time, diag counts and error logs, one record per host; and its renderer core's work, timed by a disposable overlay's scopes
test_widget_attacks API Run widget-sandbox's attack corpus through the chat list with tapped ops, one click each and loopback network controls
test_widget_bot API Answer an armed widget message's bot requests from a table at the runtime's request seam, with a record per message
test_widget_conformance API Run widget-sandbox's conformance sets through the chat list in a one-column window on its own screen, with real input, 2x stage crops and requested whole-card pictures, into the run file report.mjs reads
test_widget_stress API Run widget-sandbox's stress kit through the chat list, one widget alive at a time in a one-column window on the fastest screen, with frame, renderer, host and widget readings per level, into the run file of one pass that report.mjs reads
test_widget_fixture Internal dependency Shared live-chat fixture of the widget stage packs: gates, local widget messages with their media fields, cards, a watched core's stop the desktop decided itself, runtime log, controls, the primary window's one-column placement and release
test_messages API Observe matching sent server messages
test_history_fixtures API Lifetime-owned history service-message fixtures
test_notify_override API Local notify-setting override and restoration
test_custom_emoji API Explicitly registered ready emoji fixture
test_launch_fuse API Assert intended OS launches while blocking them; report a credential-bearing one by label
test_open_handoff API Observe document-open handoff and launch classification without OS launch
test_transfer API Observe document transfer success/failure
test_web_cache API Clear the exact web-document cache entry
test_rpc_retry API Observe server-error resends
test_rpc_fixture API Boxed controlled RPC replies
test_fixture_fence API Refuse and record requests naming registered local fixture message or synthetic peer ids
test_fixture_transaction API Read-before-write server fixture changes with rate deadlines and a resumable restoration record
test_wallet_ready API Judge served wallet freshness
test_wallet_intercept API Controlled wallet request interception
test_gram_reconcile API Reconcile golden-wallet custody and words
test_secrecy_scan API Certify fixture-secret absence from logs
test_publication_identity_scan API Audit supplied public identities and launch-log coverage
test_publication_identity_scan_self_test Self-test Self-test for identity, coverage and claim separation
test_log_scan_internal Internal dependency Shared reader for secrecy, identity and coverage scans
test_scenario Scenario slot Disposable task scenario entry point

Read the selected module's header before using it; the contracts there are more precise than the summary above. Search this directory before writing a new local helper: an overlay that reimplements a shared facility is a test flaw like any other, because the local copy carries none of the refusals the shared one accumulated. One overlay rebuilt grab-check-save by hand after CaptureWidget had already refused blank and hidden targets for months.

For new or changed modules, follow Adding a harness helper.