178 commits

Author	SHA1	Message	Date
BruebachL	ad0fd2d4b0	[CrashReporting] Sentry send path with consent gating (#7522) ... * [CrashReporting] Real send path with consent gating (DSN still placeholder) require_user_consent + give/revoke around the Ask dialog, message event with in-memory crash-log.txt bytes attachment, before_send scrubs message/logentry/breadcrumbs/data of cockatrice:// secrets. Failure box offers Save (shows path) / Discard (revokes + rejects) / Cancel (back to dialog); reject() override so X/Escape cannot bypass revocation; silent AlwaysSend path falls back to a local file. Empty DSN keeps the previous mock behavior byte-identical. Tests: 528/528 ctest green, format clean, code + UI gates PASS. * [CrashReporting] Attach crash log and user note to the pending crash event The redacted log used to ride a separate message event from the native crash (minidump + stack), forcing manual correlation in Sentry. Fold it into the pending crash envelope(s) instead by appending well-formed attachment items before consent releases them: sentry-native replays cached envelopes byte-for-byte (raw read, raw serialize in 0.17.3), so the appended crash-log.txt and user-note.txt items upload with the very event that already carries the dump and stack. Covers both storage locations: Ask mode's consent-held files in <database>/cache and AlwaysSend's transport-dumped files in <database>/<run>.run, which must be attached before sentry_init replays them. When no pending crash event exists (unclean exit without a native crash) the standalone message event remains as fallback, now also carrying the optional note. The dialog gains an optional free-text field ('Here's what I was doing...') that travels as user-note.txt; the mock JSON records it too. * [CrashReporting] Send PROJECT_VERSION as the Sentry release main() reported VERSION_STRING (PROJECT_VERSION_FRIENDLY) as the Sentry release — that carries a build date, so every build of the same version landed in its own release, and it never matched the installer, which reports CPACK/PROJECT_VERSION. Bake COCKATRICE_SENTRY_RELEASE= PROJECT_VERSION instead: date-free, identical between client and installer, and derived from the git tag on tag builds, so events, symbol uploads and the installer log all group under one release. * [CrashReporting] Friendlier crash dialog copy Reword the report dialog along a single friendly arc: headline ('Oops, we crashed — and now we need your help to make Cockatrice better!'), why helping is easy, exactly what gets collected (crash dump with stack trace, app version, OS info, redacted log excerpt), the value of each report ('better and less crashy for everyone'), and the consent choice ('Send and remember makes this automatic without interruption in the future'). The checkbox itself stays direction-neutral ('Remember my choice') because it also persists a 'Don't send' decision. * [CrashReporting] Log where crash reports go and warn when they cannot The client resolved COCKATRICE_SENTRY_DSN (environment over the baked-in value) without ever saying which one won, and sentry-native refuses to build a request for a DSN it considers invalid without reporting anything, so a tester pointing the client at a scratch project had nothing to go on when a crash never appeared. Say which DSN source won, and log the destination (host and project id, never the key) once the SDK is armed. Warn instead of staying quiet when no DSN is configured at all, when a DSN is malformed - disabling reporting rather than silently falling back to the project baked into the build, which is the rule the installer already follows in cmake/CrashReporting.cmake - and when the build was compiled without Sentry support. A deliberate Never-send stays at info level: that is a choice, not a misconfiguration. Also warn when a pending crash envelope was captured for a different destination than the session that will upload it. sentry-native posts every envelope with the DSN held in the transport state of the session doing the sending, and envelopes restored from disk are raw, so the "dsn" header of a cached envelope is never read: a crash captured for one project can be uploaded to another with nothing in the log to show it. Tests cover the DSN shape check (same unanchored pattern the build matches, so a DSN this build accepted is never rejected here), destination description without the key, and the mismatch detection, which stays quiet for a different key on the same project, unreadable files, and headers without a dsn. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-19 14:57:03 +02:00
BruebachL	6385f3fab3	[CrashReporting] Mock crash-report dialog and reporting preference (#7521) ... * [CrashReporting] Mock dialog, sentry skeleton and CI prerequisites Mock: DlgCrashReport preview dialog (no account needed, redacted log preview), CrashReportingMode Ask/Always/Never setting in DebugSettings, --simulate-crash flag and Help-menu preview entry. Send writes a local mock JSON only. Vendor sentry-native 0.17.3 via FetchContent behind WITH_CRASH_REPORTING (breakpad everywhere, backend none under sanitizers) plus a CrashReporter facade with early init and post-loop shutdown. DSN is a placeholder: empty CMake default with COCKATRICE_SENTRY_DSN env override, so SDK init is skipped and mock behavior is unchanged until a real DSN is configured. CI prerequisites riding here so every layer below builds green: libcurl dev plus CA bundles in the Linux images and CodeQL env, sentry Qt integration off (its own find_package cannot see the macOS thin Qt layout), and the StrFunc Using construction makensis requires. Tests: ctest green, format clean, code plus UI gates PASS. * [CrashReporting] Forward mode changes to SDK and fix Linux CI opt-in Connects PersonalSettings::crashReportingModeChanged to giveConsent()/revokeConsent() so switching from AlwaysSend to NeverSend (or Ask) mid-session reaches the SDK instead of leaving consent given until restart. Forwards WITH_CRASH_REPORTING into the docker container the way CCACHE_DIR is, so the Linux package leg actually builds with crash reporting ON instead of silently building OFF. * [CrashReporting] Give crash-report logging its own category per review * [CrashReporting] Drop unused DebugSettings constructor visibility per review * [CrashReporting] Document reporting modes on the enum per review * [CrashReporting] Match the dump scan to the writer's temp fallback Review nit: when the handler's SHGetKnownFolderPath lookup fails it writes to <temp>/../.., but the startup scan looked in QDir::tempPath(), so a dump produced through the fallback would never be found and its crash would never offer to be reported. Walk up the same two levels in the scan. The path is Windows-only (#ifdef Q_OS_WIN); the statements were compiled and run standalone against Qt 6 to confirm the API use. * [CrashReporting] Complete the dump-scan fallback path The previous commit walked the scan's fallback up two levels but dropped the CrashDumps/Cockatrice suffix that the writer appends unconditionally, so a dump written through the fallback was still invisible to the scan. Evaluate the writer's expression verbatim rather than mirroring it with QDir: cdUp() is existence-checked and silently leaves the path alone when the directory is missing, parent_path() is not, and QDir::absoluteFilePath() joins the shared suffix without risking a doubled separator when the walk lands on a root. Verified by compiling the block verbatim against Qt 6 with -Wextra -Werror and comparing it with the writer's join for both the normal and the fallback case. Also record that %LOCALAPPDATA% is the scan's stand-in for FOLDERID_LocalAppData and why the two normally agree. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-19 14:57:03 +02:00
Zach H	ea6af0048c	[Server] Parse the rightmost address from forwarded header lists (#7586) ... X-Forwarded-For carries a comma-separated list, and constructing a QHostAddress from the raw header value returns a null address for it, so a configured X-Forwarded-For or a multi-hop header never resolved and websocket clients stayed on the proxy's address. - Add parseForwardedAddress, returning the rightmost usable entry of a forwarded header list - Use it for the websocket handshake instead of parsing the raw value - Exempt loopback from per-address auth rate limits so a same-host proxy with a customized trusted_sources list cannot lock every user out - Add forwarded_address_test Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-18 16:53:55 -04:00
Zach H	46cdbac830	[Servatrice] Match trusted_sources by address, not substring (#7582) ... trusted_sources was checked with QString::contains. A dual-stack listener reports a local proxy as ::ffff:127.0.0.1, which never matches the default "127.0.0.1,::1". Behind nginx the forwarded client IP header was then ignored, every websocket user shared the proxy's address, and logins failed with "too many concurrent connections" once max_users_per_address was reached. Add isTrustedSource(), which parses each entry as a QHostAddress and compares exact addresses with IPv4-mapped addresses normalized, and use it at all four call sites. The ban command also read server/trusted_sources instead of the documented security/trusted_sources; it now uses the same getter. Entries that are not IP addresses are logged at startup since they no longer match anything. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>	2026-10-18 16:50:04 -04:00
BruebachL	a139084ad6	[Security] Add challenge-response auth with scrypt verifiers and stop storing plaintext passwords (#7193) ... * [Security] Add challenge-response auth with scrypt verifiers and stop storing plaintext passwords Challenge-response authentication: the client derives a scrypt verifier (RFC 7914, EVP_PBE_scrypt, N=32768, r=8, p=1) and authenticates with HMAC-SHA256(key, nonce), so neither the password nor its hash is transmitted. The stored format becomes "$scrypt$<n>$<r>$<p>$<salt>$<verifier>" and Response_PasswordSalt now carries the cost parameters. Strict servers only accept scrypt verifiers; legacy accounts are migrated after a successful login. Fix #344 for challenge-response servers: a saved profile stores the derived verifier under the password key instead of the plaintext password. The connect dialog loads it without revealing it, autoconnect passes it through, the change-password dialog no longer prefills the old password field with it, and the client only persists the verifier when "Save password" is checked. Took 3 minutes Took 1 minute Took 10 seconds Took 7 minutes * [Security] Harden credential format acceptance and challenge nonce validation * [Security] Refuse KDF downgrade; trim profile name and warn on silent credential drop * [Security] Add SupportsAll ServerOptions enumerator to fix enum validation assert The server advertised both SupportsPasswordHash (1) and SupportsChallengeResponseAuth (2) via a bitwise-OR that produced the value 3, which no enumerator declared. Protobuf's proto2 enum validation rejects the value in _internal_set_server_options, crashing servatrice on the debug assertion. Add a SupportsAll = 3 enumerator with an identical wire value so the bitmask reads unchanged while the value is now a documented, valid member of the enum. * [Security] Normalize username when issuing challenge nonce cmdRequestPasswordSalt bound the challenge nonce to the raw, un-simplified username while cmdLogin verifies it against the simplified name. A user who typed leading or trailing whitespace in the username box would be silently rejected during challenge-response login even with the correct password. Simplify at the nonce-issuing site so both commands agree on the normalized name. Also drive-by formatting: collapse the now single-expression serverOptions initializer onto one line. * [Security] Keep an encrypted legacy password backup for old-client logins Migrating an account to scrypt stranded pre-challenge clients, which can only present the legacy salt+hash: after migration the server served them a scrypt salt they cannot use, so the only recovery was a password reset. When security/legacy_backup_key is configured (64 hex chars, openssl rand -hex 32), submitPasswordVerifier now seals the legacy credential with AES-256-GCM under that key before overwriting the row, storing it in password_legacy_backup with the time it was last used. Old clients keep working: cmdRequestPasswordSalt serves the backup's 16-char salt in password_salt so they can derive the legacy hash they were always able to compute, while new clients derive the verifier from the new scrypt_salt field, which always carries the verifier salt. The non-challenge login path accepts the plaintext or legacy hash against the decrypted backup and refreshes its window; a migrated row with no usable backup fails closed. Backups age out so a database dump taken after the window closes never carries a crackable legacy hash: getUsableLegacyBackup purges on first use past security/legacy_backup_ttl_days (default 90, clamped 1..3650), and a six-hourly sweep drops backups idle accounts never touch. changeUserPassword clears the backup since a new credential invalidates it. Without a configured key nothing is sealed or purged, so a temporarily missing key cannot destroy existing backups. The AAD is the simplified user name on both sides, sealing the blob to its row so it cannot be moved between accounts. Sealing failures degrade to the previous lockout behaviour rather than blocking migration. The columns extend the unmerged 0036->0037 migration, so the schema version stays 37. * [Security] Address review: proto compat, plaintext verify, AAD case, strict scope Fix the two blockers and follow-ups from the 2026-10-16 review: - Advertise challenge-response in a new supports_challenge_response field instead of the ServerOptions enum and drop SupportsAll: proto2 enums are closed, so already-released clients parsed server_options = 3 as NoOptions and fell back to plaintext passwords. server_options stays exactly SupportsPasswordHash on the wire; old clients ignore the new field. - Verify a plaintext password directly against a migrated scrypt row with verifyPassword() so email changes (cmdAccountEdit) and old clients sending password keep working without a legacy backup key. Pre-hashed values are still checked against the backup, and its TTL window is refreshed only when the backup is what actually matched. - Use the stored row name as the AES-GCM AAD on both seal and open so a login whose case differs from the stored name (utf8mb4_unicode_ci) still opens its backup. - Require challenge-response in strict mode only for registered SQL accounts (userExists gate), leaving guest/unregistered logins and none/password auth methods unaffected. - Document in servatrice.ini.example that a captured legacy hash can take over a not-yet-migrated account and that strict only constrains registered SQL accounts. Schema version is 38. * [Security] Raise legacy backup TTL default to 180 days Gives operators and slow-to-upgrade users a longer downgrade window for pre-challenge clients before unused encrypted legacy password backups age out. * [Security] Fix servatrice.cpp formatting per clang-format-18 * [Security] Add missing forward declaration for Command_SubmitPasswordVerifier The cmdSubmitPasswordVerifier declaration references the proto-generated type, which otherwise does not name a type in this header and breaks every C++ build. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> Co-authored-by: BruebachL <bruebachl@cockatrice.example>	2026-10-16 22:52:26 +02:00
BruebachL	ce51e7147d	[Servatrice] Log room channel messages in audit log (#7479) ... * [Servatrice] Log room channel messages and bump schema to 38 * Fixup: show room_channel audit rows and keep migration nullable The log history query returns room_channel rows, but the log tab dropped them. Render them in the room table with an extra Channel column. Also drop the NOT NULL on target_type from the 37->38 migration so migrated schemas match the nullable servatrice.sql definition (and strict-mode ALTERs cannot fail on legacy NULL rows). Update the channel tests to the flag-based privilege model: the admin fixture now carries IsModerator alongside IsAdmin, as the database layer assigns. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-16 22:46:39 +02:00
BruebachL	c99113ab0e	[Client] Tabbed room channels with unread states and collapsible chat (#7478) ... * [Client] Add room channel selector with per-channel chat views * [Client] Tabbed room channels with unread states, collapse, read-only, banners Reworks the room chat channel display: channels move out of the combobox into a tab bar below the message input showing every channel available to the user. - Each tab shows its unread message count and is tinted when it holds unreads or a mention; both colors are user-configurable chat settings (channelUnreadColor, channelMentionColor) with a settings-page section. - Clicking the active tab collapses the chat: message view and input hide and the splitter section minimizes to the tab strip; expanding restores the previous splitter sizes. Viewing a channel clears its unread state. - Channels the user may read but not write (per write_access_level) show with the message input disabled and a read-only placeholder; sends are guarded locally, the server still enforces. - Channel banner_message strings render at the top of their channel view on join, like the room join message on Main. - Replayed channel history no longer inflates unread counts; moderation redaction now applies across all channel views. * Fixup: keep chat tab bar visible when chat is collapsed QSplitter children are collapsible by default, so setSizes({1, 0}) hid the whole chat group box including the channel tab bar. Mark the chat widget non-collapsible so collapsing clamps to its minimum size (title + tab bar) instead. * [Room] Explicitly initialize MAIN_CHANNEL per review * [Room] Address channel tab review follow-ups - gate @all mentions on moderator/admin sender to match ChatView, whole-word matching for own name and highlight words - hide single-channel tab bar so old servers do not show a lone Main tab that collapses chat on misclick - capture target ChatView in sendMessage so errors land in the channel the message was sent from - share ignore/unregistered filtering between RoomSay and RoomChannelSay via isRoomMessageIgnored * [Client] Fix tab_room.cpp formatting per clang-format-18 --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> Co-authored-by: BruebachL <bruebachl@cockatrice.example>	2026-10-16 22:46:39 +02:00
BruebachL	6c660d6b21	[Installer][Client] Back up user data before old uninstallers can delete it (#7503) ... * [Installer][Client] Back up user data before old uninstallers can delete it * [Installer][Client] Fix stale backup resurrection, batch session backups, extend coverage NSIS: clear stale %TEMP% backup and only arm restore when something was copied, directory-existence checks, back up loose .dec decks, drop downloaded pics but keep pics\CUSTOM. Helper: single session dir per run (ms timestamps), same-basename disambiguation, partial copies still report their location. * [Installer][Client] Fix review findings: throttle backups, NSIS old dir, portable messaging * [Installer][Client] Fix review findings: backup markers, NSIS relocate, size cap * [Installer][Client] Fix review findings: sounds, session coverage, NSIS wording Client: back up data/sounds in portable mode and mention it in the warning; reuse a recent session only when complete and covering the current sources (source manifest), estimate size only for fresh copies, normalize dedup via isInsideDir. Helper: source manifest per session with sessionCoversSources(), clear out-params on all failure paths. NSIS: document defensive portable branch, merge backup over surviving data/ to repair partial deletes, copy (not move) wording in relocate, note .txt exclusion in deck notice. * [Installer][Client] Fix review findings: stale backup, empty reuse, relocate, portable warning NSIS: wipe TEMP backup once per run to avoid resurrecting deleted files, check data dir existence in relocate so dir-only trees are moved. Client: never reuse a recent session for an empty source set, only fall back to a covering session on copy failure, warn fresh portable installs instead of staying silent. * [Installer][Client] Fix review findings: usable session reuse, loose decks, NSIS pics, symlinks * [Installer][Client] Fix portable backup exclusions in offender merge * [Client] Split at-risk data warning into focused helpers Ricky's nit on the 200+ line warnAboutDataPathsInProgramDir: extract collection (portable/installed), backup, message building and dialog into named helpers. No behavior change, all tr() strings untouched. * [Client] Read path settings inside isPortableBackupExcluded * [Client] Make stateless at-risk helpers static * [Client] Extract at-risk data warning out of MainWindow * [Installer][Client] Fix review findings: subdir backup, portable warning, backup API NSIS: - Back up unknown top-level install subdirs (e.g. <installdir>\decks), not just loose deck files, so the 3.1.1 -> new upgrade no longer loses data folders the old RMDir /r deletes before the client ever runs. Payload dirs (plugins/translations/sounds/themes) stay in place to avoid restoring stale DLLs; restore + relocate handle the new backup set and the one-time notice covers both decks and folders. - Restart the %TEMP% backup when the 32/64-bit passes point at different dirs instead of merging (which restored one dir's files into the other); document that only the last pass relocates and the %TEMP% copy is not a durable second copy. Client: - Drop the portable at-risk warning: portable installs never shipped an uninstaller and portable updates preserve data/, so every portable user got a spurious dialog plus daily host-machine copies. - AtRiskDataWarning is now a plain class with Q_DECLARE_TR_FUNCTIONS instead of QObject+Q_OBJECT; logging category renamed to at_risk_data_warning; MainWindow wrapper inlined into startupConfigCheck. - Lower the synchronous startup backup cap 500 -> 200 MB. Utility/tests: - Remove test-only backupAtRiskPath()/findRecentBackupSession() and the no-marker/no-manifest legacy paths; isBackupSessionComplete() now requires the complete marker; tests updated to the plural API. * [Client] Restore collectInstalledAtRiskPaths per review a23e02b5 removed the definition along with the portable helpers, breaking the build (undefined reference at link time, and isInsideAnyDir unused under -Werror on Arch/Debian 13). Restores the 202c0d65 version (pathsInsideDir plus loose top-level deck files) as suggested. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-16 22:05:42 +02:00
BruebachL	456d9d9e70	[Server] Inform client of login/registration rate limit timeout (#7468) ... * [Security] Add per-address rate limiting for auth endpoints Introduce a thread-safe RateLimiter that tracks attempts per key (IP address) within a sliding time window, and wire it into the authentication endpoints: - Login: failed login attempts from an address are counted; once the configured maximum is exceeded within the window, further logins from that address are rejected with RespTooManyRequests. A successful login clears the failed attempts for that address. - Registration: implement the previously stubbed tooManyRegistrationAttempts, limiting how many accounts can be created per address per window. - Forgot-password: throttle both the email-request and the email-challenge paths per address. New [security] settings with defaults: max_login_attempts_per_ip=5 / login_attempt_window_seconds=900 max_registrations_per_ip=2 / registration_window_seconds=3600 max_forgot_password_requests_per_ip=3 / forgot_password_window_seconds=3600 Adds unit tests for the RateLimiter (window limit, over-limit blocking, clearing, per-key independence). Took 3 minutes * [Server] Fix rate limiter: disable semantics, draining window, bounded keys, pre-work login gate - recordAttempt()/isBlocked() return false when maxAttempts <= 0 or windowSeconds <= 0, so the documented 'set to 0 to disable' actually disables instead of blocking every address permanently - Keys at their limit are rejected without recording, so repeated attempts at the attacker's pace cannot keep sliding the window and holding the lockout open; a block now clears once the recorded attempts age out of windowSeconds - Attempts map is bounded: an opportunistic prune (every 60s) erases keys whose attempts have fully aged out - Login is throttled before the work: isLoginRateLimited() (a new virtual on Server, backed by isBlocked()) gates loginUser(), so a locked-out address no longer burns a database round trip and password verification per attempt - recordAttemptAt()/isBlockedAt() time seams make the sliding-window behaviour deterministically testable; tests rewritten for the real semantics (no reliance on the maxAttempts=0 bug, no sleeps) * [Server] Inform client of login/registration rate limit timeout - Add retryAfterSeconds/retryAfterSecondsAt to RateLimiter to compute seconds until oldest attempt exits the sliding window - Expose loginRateLimitRemainingSeconds via Server/Servatrice - Include denied_end_time (absolute epoch) in RespTooManyRequests for login pre-auth and when transitioning to lockout after failed attempt - Include denied_end_time in registration rate limit response - Client: show retry time for login RespTooManyRequests - Client: simplify registration RespTooManyRequests message to show retry time when provided - Add test for retryAfterSecondsAt * [Client] Restore registration guidance and drop duplicated rate limit reason The registration RespTooManyRequests case now only shows the retry-after copy when the server actually sent a denied_end_time; the per-email account cap path (no end time) falls back to the previous guidance about already having an account and contacting the server operator. Drop the canned denied_reason_str from the new login rate limit branches so the client no longer prints the same sentence twice. * [Server] Exempt trusted_sources from auth rate limits Addresses in security/trusted_sources bypass per-address limits, matching the existing max_users_per_address bypass in initSession. Without this, Webatrice behind local nginx (all clients as 127.0.0.1) would lock out every web user after 5 mistyped passwords, and cap the whole server at 2 registrations / 3 forgot-password requests per window. Adds Servatrice::isRateLimitExempt helper; recordFailedLogin, isLoginRateLimited, recordRegistrationAttempt and recordForgotPasswordAttempt all check it. # Conflicts: # servatrice/src/servatrice.cpp * [Server] Key login rate limit by (address, username) A successful login cleared the whole per-IP failure history, so 4 bad guesses plus 1 self-login (or guest UnknownUser login) from the same address bypassed the limiter. Failures are now bucketed per (address, lowercase username) and a success clears only its own pair. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-16 05:47:43 +02:00
BruebachL	fec693ca2a	[Tests] Scope loader matcher settings to the process to fix parallel-run flakes (#7504) ... * [Tests] Scope loader matcher settings to the process to fix parallel-run flakes * [Tests] Clean up per-process settings sandbox after loader matcher run * [Tests] Skip loader matcher sandbox cleanup during gtest discovery --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-15 12:21:40 +02:00
BruebachL	cce606764c	[Server] Resolve a zone dump request to an in-range index span (#7401) ... Command_DumpZone.number_cards is an unbounded sint32 supplied by the client, and cmdDumpZone used it as an index without ever putting it in range. for (int i = 0; (i < cards.size()) && (i < numberCards || numberCards == -1); ++i) { const auto &findId = cmd.is_reversed() ? cards.size() - numberCards + i : i; Server_Card *card = cards[findId]; A forward request happened to be safe, because the loop stopped at the zone's own size. A reversed one was not. number_cards is documented by the client as -1 for "the whole zone", and cards.size() - numberCards is therefore cardCount + 1, one past the last card. A request larger than the zone puts the offset at a negative index instead. Both read out of bounds, then dereference the result as a Server_Card, so either adjacent heap memory or a crash. Any player can reach it: the zone only has to be a public one, or the requester's own, and both number_cards and is_reversed come straight off the wire. A reversed dump of -1 is enough on any public zone with cards in it. Resolve the request to a half-open span up front. -1 and any positive count are clamped to what the zone actually holds, any other negative count resolves to nothing, and a reversed request takes the span from the end rather than computing an offset from a value that may not be in range. cards.at() then carries the bound as a second line of defence, and qsizetype keeps the loop and the list's own indexing type aligned. Also report card_count as the zone's real size. It was numberCards < cards.size() ? cards.size() : numberCards, which agrees with the zone size for every request that was in range and reported the client's own oversized number otherwise. The client uses card_count to draw placeholders when a zone lists no cards, so overstating it invents cards that were never there. Put the resolution in its own translation unit. server_abstract_player.cpp refers to the RNG global that only servatrice's main.cpp defines, so anything linking that object needs the whole server binary behind it, which is why no test could reach this code before. server_dump_zone_test covers each request shape and then asserts the span is contained in [0, cardCount) across the edges of the request domain, INT_MIN and INT_MAX included. Reverting the arithmetic makes three of them fail, the two out-of-bounds shapes and the containment sweep. The client's own copy of this convention at view_zone.cpp:93 already clamps the way this now does. Leaving that duplication for a follow-up, since the client reaching the same answer by a different route is not a defect. Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-13 11:23:13 +02:00
BruebachL	24874b10cf	[GameSelector] Persist quick-filter toolbar edits and repair filter settings migration (#7500) ... * [GameSelector] Persist quick-filter toolbar edits and repair filter settings migration - Save game-filter parameters on toolbar changes, debounced with flush on close - Repair pre-camelCase gamefilters.ini keys in place on startup - Map the legacy creator-name filter to hostNameFilter - Balance begin/endFilterChange in GamesProxyModel::refresh - Align the hideIgnoredUserGames default between model and settings - Cover the migration with regression tests * [Review] Address RickyRister comments on filter persistence - Rename game-type prefix constants to SCREAMING_SNAKE to match MIGRATION_SENTINEL_KEY and project style - Document why GameSelector destructor flushes a pending debounced filter save (toolbar path has no other writer) --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-13 07:45:08 +02:00
BruebachL	a712553c81	[CardDB] Reload after the first run enables all sets without freezing startup (#7470) ... * [CardDB] Reload when the first run enables all sets checkUnknownSets() enabled every set when none was enabled, but the database had already been parsed under that empty enablement, so the parsers had stripped every printing. It only notified the models and never reloaded, leaving the deck editor's card database display empty until the next launch. Route the all-sets-enabled case through the same reload the new-sets path uses, and skip caching a snapshot whose sets are all disabled. * [CardDB] Own checkUnknownSets after startup and fix the cache skip The loader's initialLoadComplete guard skipped checkUnknownSets() on the first successful load of the session, so a download that landed after startupConfigCheck() and the onboarding wizard had already run (e.g. skipping the wizard download and fetching later from the menu) left every set disabled and the deck editor blank until the next launch. The guard also let the check run on the worker before the queued snapshot swap, and alongside reloadCardDatabasesAndNotify()'s own check, double-emitting cardDatabaseNewSetsFound on manual reloads. Move the check to the callers that load after startup: the post-update, custom-import, and spoiler paths run it via a QFutureWatcher once the finished snapshot has been swapped in, and the loader's synchronous path-change reload runs it directly. Judge the cache skip by the parsed sets' own enablement, what the parsers actually saw, instead of getEnabledSetNames(), which reports empty for the noop controller even though it enables every set. Wait for the first-run reparse before opening Manage Sets so the dialog cannot bind to the pre-reload set instances. * [CardDB] Queue post-reload check and unblock first-run Manage Sets reloadCardDatabasesAndNotify() no longer emits cardDatabaseEnabledSetsChanged after the swap's Reset already rebuilt the models (second 36k rebuild froze startup), and its follow-up checkUnknownSets() is queued to the GUI thread so it sees swapped-in data. cardDatabaseAllNewSetsEnabled() chains Welcome/Manage Sets off a QFutureWatcher instead of waitForFinished() blocking the event loop during the reparse. * [CardDB] Batch the first-run enable-all into a single settings flush checkUnknownSets() enabled every set with ~3 QSettings rewrite cycles per set (sort key, known, enabled), and each setValue re-opens and re-syncs the whole cardDatabase.ini. With a full database (874 sets) that froze the GUI thread for ~9s after the database download (LagMonitor: event loop stalled for 9129 ms, measured offscreen with a real 59MB cards.xml), reading as a total application hang that recovers on its own. Persist the enable-all through the controller's batched saveSets() in a single flush instead, mirroring what SetsModel::save() already does: CardSetList::enableAllBatched() assigns fresh sort keys and marks every set known+enabled in memory, then writes all records at once. SetSaveData gains the isKnown flag so the batch covers all three options, and SetsModel::save() passes the sets' known state through (unchanged behavior there). Also drop the cardDatabaseEnabledSetsChanged emission from the all-sets-enabled branch: the reload chained off cardDatabaseAllNewSetsEnabled rebuilds the models from the re-parsed snapshot via cardDatabaseReset, so emitting both rebuilt the 36k-card models twice on the GUI thread. Adds a cache-invalidation test driving the real checkUnknownSets() first-run path. * [CardDB] Batch the new-sets-found enable-all the same way The first-run batching fixed the 874-set case, but the new-sets-found dialog path (enableAllUnknownSets/markAllSetsAsKnown) still wrote per set and per option, each setValue re-opening and re-syncing the whole cardDatabase.ini. Users with hundreds of custom sets hit the same multi-second GUI stall when answering the dialog. Persist both through the controller's batched saveSets() in a single flush, writing only changed sets (so the common no-op check skips the flush entirely). Removes the now-unused write-through CardSetList methods (enableAll, enableAllUnknown, markAllAsKnown, guessSortKeys) so the batched flush is the only way to persist bulk set changes. Extends the test controller mock with per-write/batched-flush counters and adds tests asserting both dialog paths persist with exactly one flush and zero individual writes. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-13 06:29:16 +02:00
BruebachL	96b819d7c8	Support tab-separated deck imports (#4222) (#7267) ... Third-party deck tools (e.g. Delver Lens) export decks as tab-separated rows such as "2 Card Name Set Name". Card names never contain tab characters, so a leading numeric column marks a tabular record: the first two columns are read as the quantity and card name, and any further columns are discarded as printer/set metadata. - deck_list: parse tab-delimited rows before whitespace collapsing - tests: cover tab-separated mainboard and sideboard imports - docs: document the accepted plain-text import format (fills the TODO) Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-11 21:19:10 +02:00
BruebachL	57a6a0d9cb	[Client] Warn when data paths point inside the program directory (#7473) ... Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-11 15:34:49 +02:00
BruebachL	4a010bd17f	[Report] Add report reopen action for closed reports (#7442) ... - New Command_ReportReopen moderator command (extension 1206) with server handler that reopens resolved/dismissed reports, resets resolution state and notified, and records the reopening note in the audit log - Reopen... button in the report queue action bar, gated to resolved and dismissed reports, with optional note prompt and tooltip - Success confirmations for Assign, Resolve/Dismiss and Reopen now survive the list refresh via a pending status message instead of being clobbered - Dispatch/auth coverage in server_moderator_role_test Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-11 15:28:45 +02:00
BruebachL	bdda3dedfa	[CI] Publish a gcovr line-coverage report from the test run (#7404) ... * [Tests] Register each test case with CTest instead of each test binary Registering a test cost four edits across a dozen CMakeLists files: add_test, add_executable, a conditional add_dependencies on gtest, and target_link_libraries. The conditional branch was the trap -- it is only needed when GTest was fetched and add_subdirectory'd, and omitting it yields undefined gtest symbols at link time -- so it was easy to leave out and impossible to notice until the link failed. Fold all four steps into cockatrice_add_test() and migrate every test directory onto it, with a cockatrice_add_manual_test() wrapper for the three benchmarks that are built but deliberately kept out of ctest. The calls now read as a declaration of what a test is rather than a transcript of how to register one. While rewriting the registrations, replace add_test(NAME x COMMAND x) with gtest_discover_tests(). Naming the binary as a single CTest entry meant a CI failure reported only which binary failed, and ctest -R Suite.Case could not select an individual case. Discovery raises the suite from 35 opaque entries to 469 real ones: ctest -R ThemeMigrationTest.IsIdempotent now runs exactly that case, and ctest -L server / -LE bench work off labels applied at registration. Give every test a 60s TIMEOUT. CTest's implicit default is 1500s, so one hung socket wait would stall CI for 25 minutes, and single_instance_manager_test already does real QLocalSocket waits. Dropping per-case granularity costs wall-clock time, since the suite becomes process-spawn bound rather than CPU bound: serial ctest went from 3.8s to 30.8s. Run it with -j in CI to recover it -- 7.6s at -j4, and 3.5s at -j8, both at or below the old serial time. The binaries are parallel-safe: the only test claiming a fixed resource scopes its local socket name to its own pid, so concurrent copies cannot collide. Also delete dummy_test, which asserted 1 == 1 and only proved the harness was wired up -- something the other 468 cases already establish. No test is added, removed, or changed by this commit; the 469 cases are the same ones the 35 binaries contained. * [Tests] Make QtTest available to test targets QtTest was not among the modules test targets could link, so the Qt model and signal contracts had no way to be verified from a test: QAbstractItemModelTester and QSignalSpy both live there. Add Test to _TEST_NEEDED in cmake/FindQtRuntime.cmake. That is the single place the union of test-only Qt modules is declared, so this is the only file that needs to change to make the module reachable from every test target. Nothing uses it yet. This lands the capability ahead of the model-layer tests that need it, so those can be added as tests alone rather than as test plus build-system plumbing. * [Tests] Correct the note on the oracle test compression defines add_definitions() populates the directory's COMPILE_DEFINITIONS property, which is consumed at generate time, so it reaches every target in the directory no matter where it is called relative to add_executable(). The previous note claimed the flags stayed confined to the benchmark because the find_package block came after it; oracle_importer_test, created above that block, carries both defines too. Verified harmless rather than a live bug: HAS_ZLIB and HAS_LZMA gate code only in oracle/src/pages.cpp and oracle/src/pagetemplates.cpp, and no test target compiles either. The ordering and the leaked defines are both pre-existing, so this only corrects the explanation. * [Tests] Carry one CTest label per test instead of several gtest_discover_tests() can hand CTest exactly one label. GoogleTestAddTests.cmake accumulates its generated script with `foreach(arg ${ARGN})`, so a label list reaches the CTestTestfile as bare `LABELS ci server`, and CTest reads only the first value of a property and drops the rest without a word. The `ci` label that is appended to every registered test was therefore absent from every test that also asked for `server` or `bench`: `ctest -L ci` reported 400 of 469 cases, and `ctest --print-labels` never listed a second label anywhere in the build. Reduce each test to exactly one of `ci`, `server`, or `bench`, which is all the selections need. The three now partition the suite exactly, 400 + 64 + 5 = 469, and `ctest -LE bench` still means everything but benchmarks. A second label is now a configure-time FATAL_ERROR rather than a line that quietly means something else. Anything orthogonal to a test's kind belongs in a helper flag beside the label, so it composes with the group instead of competing for the single slot. Also raise DISCOVERY_TIMEOUT off GoogleTestAddTests' 5s default. A plain debug binary answers --gtest_list_tests in 0.04s, but that budget is thin once several binaries discover concurrently behind a -j build, and being over it fails at the *link* step with a bare "Process terminated due to timeout" that points nowhere near the cause. * [Tests] Link gtest_main on the fetched-gtest path The downloaded-gtest branch set GTEST_BOTH_LIBRARIES to `gtest` alone, while FindGTest sets it to both halves. Every test target links that list rather than picking an entry point, so on this path nothing provides main() and any test source that does not define its own fails at link with `undefined reference to main` pointing at crt1.o, naming neither gtest nor the line responsible. That stayed latent while every test carried its own four-line main(). A test that omits it now fails, and the error points at the C runtime rather than at gtest, which is a long way from the cause. Restoring both halves makes the two paths agree and drops the boilerplate requirement on new test sources. The 18 sources that still define main() keep linking: gtest_main.cc defines nothing but main, so the linker leaves its object in the archive unused rather than reporting a duplicate. * [Build] Add an ENABLE_SANITIZERS option for ASan and UBSan Adds -DENABLE_SANITIZERS, off by default, building with AddressSanitizer and UndefinedBehaviorSanitizer. The bulk of the work is in the three ways the obvious implementation silently does the wrong thing. Probe the flags on the link step, not just the compile step. check_cxx_compiler_flag() only compiles, so every sanitizer flag comes back unsupported, because linking an ASan object without -fsanitize=address leaves __asan_init undefined. The result is not a build failure but a build that appears instrumented and is not, which is worse. Use check_cxx_source_compiles() with CMAKE_REQUIRED_LINK_OPTIONS instead, and treat an unsupported sanitizer pair as fatal since the cosmetic flags are allowed to degrade. Opt vendored sources out per source, through a cockatrice_disable_sanitizers_for() helper rather than a list of set_source_files_properties() calls at the root. The helper has to exist because CMake resolves a source-file property against the directory scope its target was declared in: setting the property from the root with absolute paths is a no-op, silently, which is why the call sites sit next to the add_library and add_executable that pull the vendored code in. SFMT, minizip, the LZMA SDK and Qxt SMTP are excluded. oracle/src/qt-json is in the history but not in the tree, removed in #7214. Also mark those sources SKIP_PRECOMPILE_HEADERS. The PCH is built with the sanitizer flags, so it defines __SANITIZE_ADDRESS__; a translation unit compiled with -fno-sanitize=all does not and GCC then refuses the PCH, which is a fatal -Werror=invalid-pch in the servatrice and oracle builds. -fno-sanitize=vptr is not optional. Qt constructs QObject subclasses into raw storage and writes the vptr afterwards, so the vptr check reports on essentially every QObject instantiation, and it has no value here because nothing outside moc-generated code inherits QObject. Verified on the resulting build: our own translation units carry ASan and UBSan calls, SFMT.o carries neither, and the full suite of 464 cases passes clean under ASAN_OPTIONS=detect_leaks=0:abort_on_error=1 with UBSAN_OPTIONS=halt_on_error=0. The test suite needs two adjustments to make this option usable at all, both of which land here because they are consequences of COCKATRICE_SANITIZERS existing rather than independent changes. DISCOVERY_TIMEOUT rises from 60s to 300s under sanitizers. An instrumented binary takes 2.8s to answer --gtest_list_tests against 0.04s uninstrumented, and discovery runs several binaries at once behind a -j build. One target is excluded via a NO_SANITIZE flag, server_developer_role_test. Reaching google::protobuf::TextFormat::Printer's destructor aborts on an absl assertion inside system protobuf 33.4 whenever the consumer translation unit is built without NDEBUG, which a Debug sanitizer build is. This reproduces with no Cockatrice code at all, just "TextFormat::Printer p; p.PrintToString(msg, &out);" built with -fsanitize=address, so it is upstream rather than ours. It is a flag and not a fourth label because "skip this in the sanitizer leg" has to compose with a test's group rather than compete for the single label slot, and because a flag excludes by not registering the cases at all, which cannot rot the way a label silently can. The target still builds and still runs on the legs without sanitizers. * [CI] Run the suite once under ASan and UBSan Adds --sanitize to .ci/compile.sh, mirroring --test, and turns it on for one entry of the existing Linux Docker matrix: Ubuntu 26.04, which already had a test-enabled leg. That reuses the image, the ccache layer and the existing step shape, so the CI diff is a matrix key plus the argument plumbing. Debian 13 and Fedora 44 keep covering the uninstrumented suite, so nothing is lost by instrumenting one leg. Sanitizers stay off the release and package legs, which build with --release for packaging: instrumenting those would change the shipped binaries' optimization profile and build time for no benefit. macOS and Windows stay uninstrumented too, since ASan is unreliable against the Qt build there and MSVC has no equivalent without /fsanitize=address. --sanitize does not imply --test, so an instrumented build can be produced locally without paying for ctest. When both are given, ctest runs with detect_leaks=0 because Qt and the sanitizer runtimes both leak at exit by design and LeakSanitizer's default buries real findings under hundreds of process-teardown reports, and with UBSan halt_on_error=0 so one run collects many findings instead of stopping at the first. ASan keeps abort_on_error=1, since a bad access leaves the process unusable and continuing only produces a second, more confusing report from the wreckage. * [CI] Publish a gcovr line-coverage report from the test run Sanitizers tell us what a run did wrong; they say nothing about what it never reached. This adds the other half: an instrumented build whose report says which of our own lines no test executes. Add an ENABLE_COVERAGE option that compiles with --coverage -O0 -g, mirroring the ENABLE_SANITIZERS opt-in including the per-source carve-out that keeps vendored trees out of the denominator. -O0 is deliberate: at higher optimisation levels GCC merges and relocates lines, so the report would describe code that no longer exists in that form. Only GCC is supported, since Clang's intermediate format needs llvm-cov; asking for coverage under Clang is a hard error rather than a build that silently produces unreadable .gcno files. Keep the filters in a committed gcovr.cfg rather than in the invocation. The tests themselves are excluded because they are compiled with the same instrumentation and are only ever executed by the run doing the measuring, so leaving them in would report a healthier project than we have. The process entry points and the client's window chrome go too: each needs a display, an interactive terminal or a live database before it does anything a headless test could observe, so they would otherwise sit at 0% forever and bury real movement. Give coverage its own matrix leg rather than a second flag on the sanitizer leg. The two answer different questions, and combining them slows both down without telling us more. The new variant key gives the leg its own Docker image and ccache entry so the two legs cannot race on the same tag. The leg publishes an HTML report with per-file pages, LCOV, Cobertura XML and a JSON summary. gcovr is a report-time dependency rather than a build dependency, so a missing one warns and skips instead of failing an otherwise green build. * [CI] Resolve the coverage source root before the build dir cd The previous commit computed SRC_ROOT after `cd "$BUILD_DIR"`, under a comment claiming it ran before. `$0` is relative to the directory the script was invoked from, so from inside the build dir `dirname "$0"` no longer resolves and the subshell's `cd` fails. Under `set -e` that is fatal rather than an empty variable, which took down every leg in the matrix, coverage or not, including the ones this change does not otherwise touch. Move it above the cd and guard it on MAKE_COVERAGE, since the coverage report is the only consumer. Resolving it unconditionally meant a coverage-only detail was paid for by every build in the fleet, and gave every one of them a way to fail. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-11 15:05:28 +02:00
BruebachL	8822ade49d	[Tests] Register each test case with CTest instead of each test binary (#7399) ... * [Tests] Register each test case with CTest instead of each test binary Registering a test cost four edits across a dozen CMakeLists files: add_test, add_executable, a conditional add_dependencies on gtest, and target_link_libraries. The conditional branch was the trap -- it is only needed when GTest was fetched and add_subdirectory'd, and omitting it yields undefined gtest symbols at link time -- so it was easy to leave out and impossible to notice until the link failed. Fold all four steps into cockatrice_add_test() and migrate every test directory onto it, with a cockatrice_add_manual_test() wrapper for the three benchmarks that are built but deliberately kept out of ctest. The calls now read as a declaration of what a test is rather than a transcript of how to register one. While rewriting the registrations, replace add_test(NAME x COMMAND x) with gtest_discover_tests(). Naming the binary as a single CTest entry meant a CI failure reported only which binary failed, and ctest -R Suite.Case could not select an individual case. Discovery raises the suite from 35 opaque entries to 469 real ones: ctest -R ThemeMigrationTest.IsIdempotent now runs exactly that case, and ctest -L server / -LE bench work off labels applied at registration. Give every test a 60s TIMEOUT. CTest's implicit default is 1500s, so one hung socket wait would stall CI for 25 minutes, and single_instance_manager_test already does real QLocalSocket waits. Dropping per-case granularity costs wall-clock time, since the suite becomes process-spawn bound rather than CPU bound: serial ctest went from 3.8s to 30.8s. Run it with -j in CI to recover it -- 7.6s at -j4, and 3.5s at -j8, both at or below the old serial time. The binaries are parallel-safe: the only test claiming a fixed resource scopes its local socket name to its own pid, so concurrent copies cannot collide. Also delete dummy_test, which asserted 1 == 1 and only proved the harness was wired up -- something the other 468 cases already establish. No test is added, removed, or changed by this commit; the 469 cases are the same ones the 35 binaries contained. * [Tests] Make QtTest available to test targets QtTest was not among the modules test targets could link, so the Qt model and signal contracts had no way to be verified from a test: QAbstractItemModelTester and QSignalSpy both live there. Add Test to _TEST_NEEDED in cmake/FindQtRuntime.cmake. That is the single place the union of test-only Qt modules is declared, so this is the only file that needs to change to make the module reachable from every test target. Nothing uses it yet. This lands the capability ahead of the model-layer tests that need it, so those can be added as tests alone rather than as test plus build-system plumbing. * [Tests] Correct the note on the oracle test compression defines add_definitions() populates the directory's COMPILE_DEFINITIONS property, which is consumed at generate time, so it reaches every target in the directory no matter where it is called relative to add_executable(). The previous note claimed the flags stayed confined to the benchmark because the find_package block came after it; oracle_importer_test, created above that block, carries both defines too. Verified harmless rather than a live bug: HAS_ZLIB and HAS_LZMA gate code only in oracle/src/pages.cpp and oracle/src/pagetemplates.cpp, and no test target compiles either. The ordering and the leaked defines are both pre-existing, so this only corrects the explanation. * [Tests] Carry one CTest label per test instead of several gtest_discover_tests() can hand CTest exactly one label. GoogleTestAddTests.cmake accumulates its generated script with `foreach(arg ${ARGN})`, so a label list reaches the CTestTestfile as bare `LABELS ci server`, and CTest reads only the first value of a property and drops the rest without a word. The `ci` label that is appended to every registered test was therefore absent from every test that also asked for `server` or `bench`: `ctest -L ci` reported 400 of 469 cases, and `ctest --print-labels` never listed a second label anywhere in the build. Reduce each test to exactly one of `ci`, `server`, or `bench`, which is all the selections need. The three now partition the suite exactly, 400 + 64 + 5 = 469, and `ctest -LE bench` still means everything but benchmarks. A second label is now a configure-time FATAL_ERROR rather than a line that quietly means something else. Anything orthogonal to a test's kind belongs in a helper flag beside the label, so it composes with the group instead of competing for the single slot. Also raise DISCOVERY_TIMEOUT off GoogleTestAddTests' 5s default. A plain debug binary answers --gtest_list_tests in 0.04s, but that budget is thin once several binaries discover concurrently behind a -j build, and being over it fails at the *link* step with a bare "Process terminated due to timeout" that points nowhere near the cause. * [Tests] Link gtest_main on the fetched-gtest path The downloaded-gtest branch set GTEST_BOTH_LIBRARIES to `gtest` alone, while FindGTest sets it to both halves. Every test target links that list rather than picking an entry point, so on this path nothing provides main() and any test source that does not define its own fails at link with `undefined reference to main` pointing at crt1.o, naming neither gtest nor the line responsible. That stayed latent while every test carried its own four-line main(). A test that omits it now fails, and the error points at the C runtime rather than at gtest, which is a long way from the cause. Restoring both halves makes the two paths agree and drops the boilerplate requirement on new test sources. The 18 sources that still define main() keep linking: gtest_main.cc defines nothing but main, so the linker leaves its object in the archive unused rather than reporting a duplicate. * [Build] Run the test suite under ASan and UBSan (#7400) * [Build] Add an ENABLE_SANITIZERS option for ASan and UBSan Adds -DENABLE_SANITIZERS, off by default, building with AddressSanitizer and UndefinedBehaviorSanitizer. The bulk of the work is in the three ways the obvious implementation silently does the wrong thing. Probe the flags on the link step, not just the compile step. check_cxx_compiler_flag() only compiles, so every sanitizer flag comes back unsupported, because linking an ASan object without -fsanitize=address leaves __asan_init undefined. The result is not a build failure but a build that appears instrumented and is not, which is worse. Use check_cxx_source_compiles() with CMAKE_REQUIRED_LINK_OPTIONS instead, and treat an unsupported sanitizer pair as fatal since the cosmetic flags are allowed to degrade. Opt vendored sources out per source, through a cockatrice_disable_sanitizers_for() helper rather than a list of set_source_files_properties() calls at the root. The helper has to exist because CMake resolves a source-file property against the directory scope its target was declared in: setting the property from the root with absolute paths is a no-op, silently, which is why the call sites sit next to the add_library and add_executable that pull the vendored code in. SFMT, minizip, the LZMA SDK and Qxt SMTP are excluded. oracle/src/qt-json is in the history but not in the tree, removed in #7214. Also mark those sources SKIP_PRECOMPILE_HEADERS. The PCH is built with the sanitizer flags, so it defines __SANITIZE_ADDRESS__; a translation unit compiled with -fno-sanitize=all does not and GCC then refuses the PCH, which is a fatal -Werror=invalid-pch in the servatrice and oracle builds. -fno-sanitize=vptr is not optional. Qt constructs QObject subclasses into raw storage and writes the vptr afterwards, so the vptr check reports on essentially every QObject instantiation, and it has no value here because nothing outside moc-generated code inherits QObject. Verified on the resulting build: our own translation units carry ASan and UBSan calls, SFMT.o carries neither, and the full suite of 464 cases passes clean under ASAN_OPTIONS=detect_leaks=0:abort_on_error=1 with UBSAN_OPTIONS=halt_on_error=0. The test suite needs two adjustments to make this option usable at all, both of which land here because they are consequences of COCKATRICE_SANITIZERS existing rather than independent changes. DISCOVERY_TIMEOUT rises from 60s to 300s under sanitizers. An instrumented binary takes 2.8s to answer --gtest_list_tests against 0.04s uninstrumented, and discovery runs several binaries at once behind a -j build. One target is excluded via a NO_SANITIZE flag, server_developer_role_test. Reaching google::protobuf::TextFormat::Printer's destructor aborts on an absl assertion inside system protobuf 33.4 whenever the consumer translation unit is built without NDEBUG, which a Debug sanitizer build is. This reproduces with no Cockatrice code at all, just "TextFormat::Printer p; p.PrintToString(msg, &out);" built with -fsanitize=address, so it is upstream rather than ours. It is a flag and not a fourth label because "skip this in the sanitizer leg" has to compose with a test's group rather than compete for the single label slot, and because a flag excludes by not registering the cases at all, which cannot rot the way a label silently can. The target still builds and still runs on the legs without sanitizers. * [CI] Run the suite once under ASan and UBSan Adds --sanitize to .ci/compile.sh, mirroring --test, and turns it on for one entry of the existing Linux Docker matrix: Ubuntu 26.04, which already had a test-enabled leg. That reuses the image, the ccache layer and the existing step shape, so the CI diff is a matrix key plus the argument plumbing. Debian 13 and Fedora 44 keep covering the uninstrumented suite, so nothing is lost by instrumenting one leg. Sanitizers stay off the release and package legs, which build with --release for packaging: instrumenting those would change the shipped binaries' optimization profile and build time for no benefit. macOS and Windows stay uninstrumented too, since ASan is unreliable against the Qt build there and MSVC has no equivalent without /fsanitize=address. --sanitize does not imply --test, so an instrumented build can be produced locally without paying for ctest. When both are given, ctest runs with detect_leaks=0 because Qt and the sanitizer runtimes both leak at exit by design and LeakSanitizer's default buries real findings under hundreds of process-teardown reports, and with UBSan halt_on_error=0 so one run collects many findings instead of stopping at the first. ASan keeps abort_on_error=1, since a bad access leaves the process unusable and continuing only produces a second, more confusing report from the wreckage. * [Build] Fix sanitizer flag leak into release package and make UBSan fail CI - compile.sh always passes -DENABLE_SANITIZERS explicitly so the release-package reconfigure in the same build dir cannot inherit ON from the sanitized debug step (which shipped libasan/libubsan DEB deps). - Sanitizers.cmake resets COCKATRICE_SANITIZERS to OFF when the option is off (CACHE INTERNAL otherwise sticks forever, keeping NO_SANITIZE skips and the 300s discovery timeout). - UBSan now fails the job: -fno-sanitize-recover=undefined at compile time plus UBSAN_OPTIONS halt_on_error=1 (recover mode exits 0, so the old halt_on_error=0 could never fail). --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-11 14:50:09 +02:00
BruebachL	a44757bd11	[Client] Improve card database update visibility (#7458) ... * [Client] Improve card database update visibility Addresses Cockatrice/Cockatrice#7431. Cockatrice previously offered no in-client feedback while the card database updater ran: the only progress UI lived in the first-run wizard, stderr/stdout were not parsed in interactive runs, and starting a second update with Oracle already open or with a hidden background run active just showed a confusing 'already running' guard - even though Oracle is launched as a separate process with no lifecycle tracking. - Add a status-bar widget (CardDatabaseUpdateStatusBar) that appears while an update runs and shows the Downloading/Parsing/Importing stages with percent progress; it is also used in the quit-confirm prompt so users see an unfinished update before cancelling it. - Parse the updater's PROGRESS lines (new CardUpdateProgressParser, unit-tested) in both foreground and background runs and log stage changes to the new card_database.update category, throttled to at most one line per second; forward Oracle stderr to the logs. - Show a tray balloon when a background update starts and on finish or failure instead of staying silent. - Request a running Oracle instance to raise itself (QLocalServer CockatriceOracleRaise-<user>) when the user starts another update, and demote the 'already running' guard to an informative message only when no running instance can be found. - Emit download/scan/import progress unconditionally from Oracle so interactive and background runs behave the same. - Document the new log category in the troubleshooting guide. * [Client] Address card database update review feedback * [Client] Add setting to control card database update tray notifications * [Client] Address remaining card database update review nits - Rename card_update_progress_parser.{h,cpp} to card_update_progress.{h,cpp} - Store the CardUpdateProgress struct in the status bar widget * Lint. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-09 11:07:48 +02:00
BruebachL	205e1ecebf	[CardDB] Keep parse-time set filtering and bust the cache on enablement changes (#7448) ... The parsers keep dropping printings for disabled sets so that disabling a set hides them everywhere, not just in deck editor search. The cached parse output therefore depends on enablement, which the source hash did not cover: the first parse froze whatever enablement state existed at that moment and later toggles hit the stale cache. - Fold a parser schema version and the sorted enabled set names into the binary cache source hash so toggling a set or enabling newly discovered sets (including custom ones) forces a re-parse on the next load. - Harden the legacy set migration to only write keys that exist so missing legacy values can no longer clobber cardDatabase.ini and silently disable sets on upgrade. - Make CardDatabase::reloadCardDatabasesAndNotify() bypass the binary cache so an explicit reload always re-reads the files on disk, even when their size/mtime hash still matches. - Cover cache invalidation end to end: sandboxed loads assert that disabling a set, adding a custom database and enabling a fresh set each rewrite the cache, that a no-op load keeps it, and that an explicit reload picks up edits the source hash cannot see. Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-09 06:09:18 +02:00
BruebachL	95f02ac0a1	[CardDB] Drop the rebalanced-cards setting entirely (#7457) ... * [CardDB] Drop rebalanced-card filtering from the v4 parser Rebalanced cards no longer exist, so the includeRebalancedCards preference has no data to filter. Always keep printings marked isRebalanced and remove the now-unused ICardPreferenceProvider plumbing from CockatriceXml4Parser and CardDatabaseLoader. The setting itself (CardsDisplaySettings, Manage Sets checkbox, Oracle attribute mapping) is removed in a follow-up. * [CardDB] Drop the rebalanced-cards setting entirely Rebalanced Alchemy cards no longer exist, so remove the includeRebalancedCards preference surface: the CardsDisplaySettings getter/setter/signal, the ICardPreferenceProvider and ICardsDisplaySettingsProvider methods, the Manage Sets checkbox, the legacy settings migration mapping, the isRebalanced attribute mapping in the Oracle importer, and the isRebalanced attribute in the v4 schema. Builds on #7448, which already stopped the v4 parser from filtering on the preference. * [CardDB] Drop the unused isOnlineOnly printing property The Arena-card detection switched from isOnlineOnly to isRebalanced in #5778, so nothing has read this printing property since; it only round-trips from the Oracle importer into cards.xml. Remove the importer mapping, the benchmark fixture field and the v4 schema attribute so regenerated databases stop carrying it. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-09 06:02:47 +02:00
BruebachL	d603a2bb6a	[Tooling] Apply include-what-you-use across the tree (#7396) ... * [DeckList] Apply include-what-you-use to libcockatrice_deck_list * [IWYU] Apply include-what-you-use across the tree Run one whole-tree IWYU pass over the libcockatrice_* libraries, the Cockatrice client, Oracle, Servatrice and the test suite, so each translation unit includes what it uses directly and forward-declares what it does not need. IWYU only analyses real translation units, never the moc code CMake generates from Q_OBJECT headers, so it strips includes that moc output silently relies on. Where a type is passed by value (or nested in a QList/QMap) in a signal or slot signature, the generated moc code needs the complete type; those includes are kept with an explanatory `// IWYU pragma: keep`. The same holds for Q_DECLARE_METATYPE targets such as QSqlDatabase. A handful of includes are added by hand where callers brace-initialize a const-reference parameter, which also requires the complete type. Applied as a single commit on purpose: stripping includes from the libcockatrice_* headers is precisely what breaks the client, test and server translation units, and the fixes for those live in the same translation units. Splitting by component would therefore leave intermediate commits that do not compile. All 426 modified Q_OBJECT/Q_GADGET headers were verified to be moc-self-contained by compiling each generated moc_*.cpp standalone; several built only because of moc include ordering in mocs_compilation.cpp before the keep-includes were added. * [Tooling] Emit public Qt headers and fix IWYU fallout CI found Three problems the local build could not see, all because a green local build only proves the configuration it was run under. Private Qt headers. IWYU emitted private, lowercase Qt headers, and qtypes.h does not exist at all before Qt 6.9, so the Debian 12 and Ubuntu 24.04 jobs on Qt 6.4.2 failed to compile. The generator only emitted a mapping when the capital header name mirrored the private one (QAbstractAnimation for qabstractanimation.h), so qtypes.h, which is wrapped by QtTypes instead, got no mapping and kept the private spelling. Map it explicitly to <QtTypes>, which is public since Qt 5 and provides the same typedefs either way, and rewrite the 100 affected includes. Do not generalize this to "prefer any Qt* wrapper". The table is generated from whichever Qt is installed, and the wrappers are not equally old: QtLogging only appeared in Qt 6.8 and QtMinMax in 6.5, so that heuristic trades one breakage for another. Every other lowercase header still in the tree has shipped publicly since Qt 5, so those are left alone. Platform-guarded includes. IWYU only sees the preprocessor configuration it is handed, so it dropped an include that only the macOS branch of main.cpp needs, and the macOS jobs failed on an undeclared UrlSchemeEventFilter. Restore the include. Force-compiling every file that carries a platform guard under Q_OS_MAC, Q_OS_MACOS and __APPLE__ reports nothing else, so the macOS branches are covered rather than assumed. shellcheck. Drop a leftover variable and stop assigning the include-what-you-use lookup, which was only ever an existence check. * [Tooling] Map Qt 6.5+ split headers to QtGlobal for Qt 6.4 compatibility IWYU emitted QtTypes, QtPreprocessorSupport, QtVersionChecks, QtAssert, QtMinMax, QtClassHelperMacros, QtEnvironmentVariables and QtTranslation. Qt 6.5 split those declarations out of qglobal.h into separate headers, so neither the Qt* wrapper nor the q*.h private name exists on the oldest Qt we support (6.4.2), and Debian 12 / Ubuntu 24.04 fail with "No such file or directory". qtypes.h is worse still: it only appeared in Qt 6.9. QtGlobal is the one spelling that works across the whole supported range. It has been public since Qt 5, defines these declarations itself up to Qt 6.4, and includes the split headers from Qt 6.5 on, so redirect all eight private headers to it and collapse the affected includes in the tree. Curated overrides now take precedence over the name-mirroring rule, which cannot know about Qt version floors because the mapping is generated from whatever Qt is installed. That rule was mapping qtversionchecks.h to <QtVersionChecks> even with an override present, because the override was only consulted as a fallback. A curated key that matches nothing is now a hard error rather than a silent fallback, and the redirected headers are no longer advertised as the home of a symbol. Verified every Qt include in the tree resolves against the real qt6-base-dev 6.4.2 package: 15 distinct includes, 0 missing. * [Tooling] Map QOverload to QtGlobal and drop the QMetaMethodArgument include Two more headers IWYU emitted do not exist on Qt 6.4.2: QOverload was split out of qglobal.h in Qt 6.5, so on 6.4.2 it is declared in qglobal.h directly and the standalone header is a hard error. QtGlobal covers both, so map qoverload.h to it like the other 6.5+ split headers. QMetaMethodArgument is never referenced anywhere in the tree. It is a bare forwarder for qobjectdefs.h that our generator advertised as a class symbol, which is how IWYU came to add the include at all. The symbol mappings are generated from header names, so they cannot tell a forwarder that happens to be capitalised from a real class; removing the unused include is the fix, and the generic case stays a known limitation of the generator. Extending the audit to every Q* header (an earlier pass only matched headers starting with "Qt" or lowercase q*.h, so it missed all of them) against the real qt6-base-dev, qt6-multimedia-dev, qt6-declarative-dev, qt6-svg-dev and qt6-websockets-dev 6.4.2 packages: 341 distinct Qt includes, 0 missing. * [Game] Give moc the complete CardItem type in player_actions.h player_actions.h forward declares CardItem but passes QList<CardItem *> by value in actPlay/actPlayFacedown/actHide. moc instantiates the metatype for those parameters, and on Qt 6.4 that expands to sizeof(CardItem), so Ubuntu 24.04 fails with "invalid application of 'sizeof' to incomplete type 'CardItem'". Newer Qt tolerates the incomplete type, which is why only the 6.4.2 jobs failed. Include the defining header with an IWYU pragma keep, matching how the other MOC-required includes in this area are handled. * [Game] Give moc the complete ServerInfo_Card type player_logic.h, game_scene.h, view_zone.h and view_zone_widget.h forward declare ServerInfo_Card but pass QList<const ServerInfo_Card *> through signal and slot signatures. moc instantiates the metatype for those parameters, which Qt 6.4 expands to sizeof(ServerInfo_Card), so the connect in game_scene.cpp fails on Debian 12 and Ubuntu 24.04 with "invalid application of 'sizeof' to incomplete type 'const ServerInfo_Card'". Newer Qt tolerates it, so only the 6.4.2 jobs failed. Include the generated protobuf header with an IWYU pragma keep, following the pattern already used for ServerInfo_Player and ServerInfo_User in game_event_handler.h. * [Client] Complete the types Qt 6.4 needs for connect and moc Qt 6.4 expands QMetaType::fromType on the normalised argument list of a PMF connect and on moc-generated metatypes, which requires sizeof on every signal and slot parameter. Newer Qt tolerates forward declarations there, so only the Debian 12 and Ubuntu 24.04 jobs failed, with "invalid application of 'sizeof' to incomplete type" for QMouseEvent, QPrinter and GameReplay. Replace the forward declarations with the real includes: - QMouseEvent in the nine widget headers that pass it through signals - QPrinter in abstract_tab_deck_editor.cpp, for the actPrintDeck lambda - game_replay.pb.h in tab_replays.h. GameReplay is a protobuf message, so the project's game/replay.h only forward declares it and cannot supply the complete type. Found by recompiling all 1918 project TUs against the Debian Qt 6.4.2 headers, which reproduces those two CI jobs locally in one pass. * [Client] Complete Release and QMenu for Qt 6.4 moc and connect Continuing the Qt 6.4 complete-type sweep: - window_main.cpp connects ClientUpdateChecker::finishedCheck to MainWindow::checkClientUpdatesFinished, whose Release * parameter made moc 6.4 instantiate the metatype. GameReplay turned out to be a protobuf message, so Release likewise needed release_channel.h in the three headers declaring those signals: window_main.h, dlg_update.h and client_update_checker.h. - moc 6.4 also needs sizeof(QMenu) for the QList<QMenu *> carried by TabSupervisor::setMenu and MainWindow::updateTabMenu. Verified by regenerating moc output with the Debian moc 6.4.2 for all 434 Q_OBJECT/Q_GADGET classes and compiling it, plus recompiling all 1612 source TUs, against the Qt 6.4.2 headers: no remaining failures. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-08 11:13:50 +02:00
BruebachL	0934380a5c	[Translations] Load Qt's own translations and build ours with the binary (#7384) ... * [Translations] Load Qt's own translations and build ours with the binary Three separate bugs kept translations from working. installNewTranslator() loaded the application translation into qtTranslator. QTranslator discards its previous contents on every load, so that wiped the qt_<lang>.qm loaded one line earlier and installed an empty translator instead. Qt's own strings were never translated on any platform. Give each file its own translator. QLibraryInfo::path(TranslationsPath) was the only directory searched for Qt's translations. A relocated Qt can report a path that does not exist: the KDE Flatpak runtime reports /usr/translations while its files live in /usr/share/qt6/translations. Probe the usual install layouts and only hand back the ones that exist. The .qm files were attached to the default build target only, so building the executable on its own left it without any translations and the language selector empty. Attach them to the target as well. Move the shared lookup into TranslationLoader so client and Oracle cannot drift again, and have it report a missing translation at debug level. An unset language is the default for English, which has no file to begin with, and warning about it only added noise to every fresh debug log. Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> * [Tests] Define main() in translation_loader_test Without a GTest installation the fallback in tests/CMakeLists.txt links only gtest, not gtest_main, so the test has to bring its own main(). Every other test already does; only this one relied on the distribution package and failed to link on CI. Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> * [Tests] Put the translation loader fixture next to the test executable A multi-config generator (MSBuild, Xcode) writes the executable to a per-config subdirectory while the .qm stays in the current binary directory, so applicationDirPath() never saw the fixture and ABuildTreeFindsTheTranslationsNextToTheExecutable failed on Windows. * [Translations] Depend on Qt's lrelease target and skip the loader test without lrelease Attaching the .qm custom commands to a second custom target races with Qt's own <Project>_lrelease target on MSBuild/Makefiles and is rejected by Xcode, so depend on Qt's target instead. Only build the translation loader test when Qt6::lrelease is available. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de> Co-authored-by: ZeldaZach <zahalpern+github@gmail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>	2026-10-03 15:04:30 +02:00
BruebachL	3aba1ef929	[Settings] Back up the settings folder before migrating (#7386) ... * [Settings] Back up the settings folder before migrating The per-file settings migration only preserved global.ini, as global.ini.old. Everything else the migrations write was overwritten in place with no way back: the legacy migration and the 17 per-domain migrateXxxSettings functions all write straight into the per-domain .ini files, so any of them that got a key mapping wrong silently replaced a working file. The re-migration path is the sharpest case, because migrateSettingsFromGlobalIni() runs again whenever an older build wrote non-sentinel keys back into global.ini, and those values then land on top of per-domain files the user had already edited. Snapshot the whole settings folder before either migration runs, into a timestamped sibling settings-backup-<yyyyMMdd-HHmmss>. A sibling rather than a child, so a backup is never copied into the next backup and the "Open Settings Folder" action keeps showing only live files. Keep every snapshot; settings files are kilobytes, so growth is not a concern. SettingsMigration::isMigrationPending() gates the copy so a normal start creates nothing. It reuses the same condition migrateSettingsFromGlobalIni() uses to decide whether it has work, factored into a shared helper so the two cannot drift, and it ignores the legacy migration on an empty profile where that migration only writes its sentinel. The backup itself skips a folder whose files are all empty, which covers a fresh profile whose global.ini was created but never written. A failed copy logs a warning and returns an empty path rather than aborting the migration, since a full disk or an unwritable data folder must not leave the client unable to start. global.ini.old still works as the single-file escape hatch. Document the folder layout, the automatic backups, and the fact that migrations only run forwards, so users know to copy the settings folder before downgrading to a version that cannot read the per-group files. * [Client] Match the desktop file name case for the portal app ID QGuiApplication::setDesktopFileName() was given "Cockatrice"/"Oracle" while the installed desktop entries are cockatrice.desktop and oracle.desktop. Linux paths are case-sensitive, so xdg-desktop-portal failed the app-info lookup for "Cockatrice" and Qt warned on every QDesktopServices call that it could not register with the host portal. The same string becomes the xdg-shell appId, so the window icon lookup failed on Wayland for the same reason. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-03 14:59:58 +02:00
BruebachL	014983b5bc	[Theme] Carry Default theme directory customizations over to System (#7385) ... * [Theme] Carry Default theme directory customizations over to System The Default -> System rename rewrote the stored theme name but left the user profile's themes/Default directory behind, so anyone who had customized zone graphics, stylesheets or palettes there silently lost them. Migrate the directory alongside the name. Rename it when System does not exist yet, otherwise merge file by file without overwriting what System already has, then drop the legacy directory. Users who already ran the rename get their customizations back on the next start, and a hand-made Default folder of a user on another theme is left untouched. * [Tests] Give theme_migration_test an explicit main and gate it on WITH_CLIENT The build machines that cannot find GTest fall back to the downloaded googletest, which links the bare gtest library without gtest_main. Every other test defines its own main() for that reason; this one did not, so Debian 13, Fedora 44 and Ubuntu 26.04 failed to link it. The test compiles client sources, so -DTEST=1 -DWITH_CLIENT=0 has no way to build it. Gate the subdirectory on WITH_CLIENT like the other client-source test. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-03 13:43:25 +02:00
BruebachL	add65caae5	[Client] Hand off single-instance payloads outside the readyRead signal (#7373) ... * [Client] Hand off single-instance payloads outside the readyRead signal Opening a cockatrice:// link from outside a running client showed the "Join game ...?" confirmation and then crashed on both answers. The join reached the server first, so the other players saw the user join and immediately disconnect. SingleInstanceManager::handleNewConnection() emitted filesReceived from inside the forwarding socket's readyRead lambda and only disconnected afterwards. A joingame link makes the handler show a modal QMessageBox, which runs a nested event loop. The sending instance exits as soon as it has the acknowledgment, so the socket was disconnected and deleteLater'd inside that loop: it died in the middle of its own readyRead signal, and both this lambda and Qt's signal dispatch kept using the freed memory. Disconnect right after the acknowledgment and emit filesReceived through a queued call, so no handler and no nested event loop ever runs inside the socket's signal. Guarding the disconnect call alone is not enough, since Qt still touches the socket after the lambda returns. Add a regression test that reproduces the nested-event-loop hand-off. * [Tests] Apply cmake-format to the single-instance test wiring * [Client] Defer the single-instance socket teardown past the readyRead frame The forwarded payload is emitted synchronously from inside the forwarding socket's readyRead, which is what lets the modal "Join game?" confirmation appear. A handler can spin a nested event loop, and the sending instance exits as soon as it has the ACK, so the socket is disconnected and deleteLater'd while that loop runs. Control then returns to this frame with the socket already destroyed, and the inline disconnectFromServer() touched freed memory. Queue the teardown instead and guard it with a QPointer taken before the handlers run -- building the guard afterwards is itself a use-after-free, since QPointer dereferences the pointer it is constructed from. The payload still reaches the handlers synchronously, so the confirmation is unchanged. The regression test covers both halves: the handlers must run while the socket is still connected, and the socket must not be torn down until the readyRead frame has returned. * Lint. * [Client] Only delete the single-instance socket outside the readyRead frame The auto-delete on `disconnected` was the remaining use-after-free. A handler may spin a nested event loop (the modal "Join game?" box), the sending instance exits as soon as it has the ACK, and so the peer closes while that loop runs. deleteLater() issued from inside such a loop is not deferred to the outer loop: Qt frees the object as soon as the nested loop unwinds, which is still inside this readyRead emission, so Qt's own signal dispatch and the handler lambda both keep using a freed socket. When the notification happens to arrive late the bug hides, which is why this only reproduced on some platforms. Drop the auto-delete and let the queued teardown own the socket's lifetime: it disconnects and then deletes, once the frame has unwound. The QPointer guard was necessary but not sufficient -- it stopped us dereferencing a dead pointer, not the socket being freed underneath the frame. The test waits on the socket's destroyed signal instead of polling, so the teardown assertion does not depend on platform event timing. * [Client] Do not queue the single-instance socket teardown The previous commit moved the deleteLater() off the disconnected signal and into a queued invocation, believing that "queued" meant "after the readyRead frame returns". It does not. A queued call is delivered by whichever event loop is running, and the modal "Join game?" box runs a nested one -- so the teardown was still executed inside the readyRead emission, which is exactly the use-after-free it claimed to prevent. Queueing was never necessary. Because nothing connects deleted/close to the socket's disconnected signal any more, no nested loop can free it while the handlers run, and emit filesReceived() is synchronous: once it returns, the frame has unwound and the socket can be closed and deleted inline. The QPointer stays, covering the one case that remains -- the primary shutting down while the confirmation box is open, since QLocalServer destroys its pending connections. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-02 21:19:02 +02:00
BruebachL	750d5b3c21	[Client] Add option to use game time in game log timestamps (#7265) ... * [Client] Add option to use game time in game log timestamps (#2202) Game log timestamps showed the local clock time, which is meaningless in a replay or in a log read after the fact — nobody can tell when a play happened relative to the game. Add a "Use game time instead of local time in game logs" setting. When enabled, the message log stamps entries with the elapsed game time (HH:MM:SS since the game started); regular chat views keep the local clock. * [Game] Move formatElapsedTime helper out of public slots (#2202) The message log needs the HH:MM:SS formatter used by the game timer, but it lives in GameState's public slots, tying a pure formatting helper to the signal-slot machinery. Move formatElapsedTime into a public section of GameState so the log can call it directly, and add the missing trailing newline to game_state.cpp. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-10-02 17:30:48 +02:00
BruebachL	eae9591a17	[Server/Client] Unify report categories (#7345) ... * [Server/Client] Unify report categories into one shared source of truth * Lont. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-24 18:37:54 +02:00
BruebachL	71ef36374d	[DeckEditor] Show card counts in Visual Deck Editor banners (#7339) ... Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-23 19:19:30 +02:00
BruebachL	7a41dfe175	[PictureLoader] Fix custom-folder pictures in subdirectories not loading (#7340) ... CardPictureLoaderLocal compared the full file name (including extension) of the custom-folder candidate against the extension-stripped base name of each directory entry, so a CUSTOM subfolder image like pics/CUSTOM/poker/1 of Hearts.png never matched and the card fell through to the network. Compare the complete base names instead. Add matcher tests covering CUSTOM subfolder resolution. Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-21 21:48:34 +02:00
BruebachL	e01912d3c8	[Tests] Add manual picture loader benchmark against the real card hosts (#7288) ... * tests: add manual picture loader benchmark against the real card hosts (cherry picked from commit 588d71d3b9a8278ab98b847802908f322c1035b7) * tests: address review on picture loader benchmark - Sandbox via unique app/org names plus Linux-only XDG redirection, derive the warm-cache probe and data path from SettingsCache, and bail out when the temporary sandbox cannot be created. - Run each pass with a fresh worker and shut workers down before reading the 429 counters, so the redirect cache is persisted and no worker thread can outlive the stack-local counters or the installed message handler. - Make s_activeCounters atomic and always forward log output when the previous handler is the built-in (nullptr) one. - Validate --timeout-min, scale the cached-pass budget with --count, honour the first --url as the stress template, and add the missing trailing newlines. - Zero-initialise SettingsCache members so the benchmark mock cannot dereference an indeterminate pointer. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-21 18:30:04 +02:00
BruebachL	14acf3bf64	[PictureLoader] Add user-configurable per-host request caps (#7287) ... * [PictureLoader] Add user-configurable per-host request caps Picture downloads were throttled to a uniform 10 requests/second per host with no way to tune a specific server. A rate-limited API host (Scryfall caps at 10 req/s) can trip 429s during bursts, and CDN hosts with no rate limit were throttled needlessly. Introduce developer-owned per-host caps that users can only ever lower, never raise, exposed in the download settings page: - DownloadSettings::DEVELOPER_HOST_CAPS sets the ceiling per host (api.scryfall.com 9, cards.scryfall.io unlimited, others 10). - A new hostRequestLimits setting stores user overrides in downloads.ini; clampHostRequestLimit() bounds them to [1, devCap] so a user can reduce api.scryfall.com to 5 but never raise it above 9. - The picture worker seeds, halves on 429, and recovers its sustained per-host allowance against the effective ceiling instead of the global maximum, and skips per-host accounting entirely for unlocked hosts (cards.scryfall.io) while global pacing and 429 backoff still apply. - The deck editor settings page gains one spinbox per known host, each clamped to its developer cap. * [PictureLoader] Let unlocked hosts skip dispatch pacing; adjust limits per URL Two refinements to the per-host request caps: - Unlocked hosts (UNLIMITED_HOST_QUOTA, e.g. cards.scryfall.io) no longer wait on the 100ms dispatch pacing or consume the global per-second quota. dispatchQueuedRequest fires their queued requests back-to-back, bounded only by their 429 backoff window and Qt's per-host connection pool, so an unthrottled CDN is not artificially slowed. - The deck editor download settings page replaces the static grid of one spinbox per known host with an "Adjust Rate Limit" toolbar action on the URL list. It picks the host out of the selected URL and clamps the entry against the developer cap table (including for user-added URLs). Also fixes a review finding: resetRequestQuota could write the UNLIMITED_HOST_QUOTA sentinel (-1) into the sustained per-host quota when a host became unlocked mid-run, permanently poisoning its allowance. Stale entries for unlocked hosts are now dropped, and the per-second seed is clamped against the effective ceiling so a lowered limit applies immediately. * [PictureLoader] Cap unlocked host bursts and adapt them to 429s * [PictureLoader] Store per-host limits readably and show them per URL * [PictureLoader] Make dispatch and rate-limit bookkeeping key on the real host Addresses ZeldaZach's round-4 review nits: - Dispatch now resolves the cached-redirect chain before the in-flight gate, so a redirect learned after a URL was queued can no longer bypass the MAX_IN_FLIGHT_PER_HOST cap and drain the whole queue onto the redirect target, which may carry its own developer cap. processSingleRequest does the same so the allowance math keys on the host that is actually hit. - The per-host in-flight slot is released when the reply is destroyed (with the worker as the connection context) rather than on a 'finished' connection bound to the work object, so an aborted reply or a work object deleted while a reply is pending can never permanently shrink the fast path's concurrency. - storeSettings only prunes limits for hosts with neither a URL nor a developer cap, so throttles on redirect targets (api.scryfall.com -> cards.scryfall.io) survive URL removal. - Unlocked hosts are offered 0..UNLOCKED_HOST_LIMIT_MAX (50) in the rate limit dialog, matching clampHostRequestLimit() and the documented hand-editable range, so values written into downloads.ini are no longer silently rewritten on the next edit. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-21 18:30:03 +02:00
BruebachL	e2a4556546	[Filter] Restore exact, case-insensitive set code search (#7336) ... * [Filter] Restore exact, case-insensitive set code search (#7332, #7333) * [Filter] Group set query suffix modes into a subrule Address review feedback: keep the 'e'/'set' prefix in one place and move the three suffix modes (exact, negated, release-date comparison) into a dedicated choice-like rule so sv.choice() still dispatches on them. Add coverage for the release-date comparison mode. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-21 14:22:05 +02:00
BruebachL	ef68a7bdcc	[Card] Add a setting for the language used in card search (#7314) ... * [Card] Add a setting for the language used in card search Localized card names and texts can now be searched too, controlled by a 'Language used in card search' toggle (English, selected card language, or both) on the general settings page. Untranslated cards always keep matching in English. Removed the redundant local copy of the URL templates list in the localized picture loader while here. * [Card] Bind search language per FilterString instance The peg parser rules are set up once per process, so the GenericQuery and OracleQuery rule actions could not capture per-instance state. Instead of storing the search language in a process-global that FilterString instance methods mutate, hand it to the rule actions through a thread-local parse context and copy it into the filter closures they produce. Card evaluation in FilterString::check no longer reads any process-global state, and each instance keeps the language it was built with; constructing one instance no longer changes what unrelated instances (deck filter, drop-to-hand, zone views) match against. The card database display model stores the raw query and rebuilds the FilterString when the search language changes, since the language is now bound at parse time. Add tests for the English/Selected/Both search modes, the English fallback for untranslated cards, and per-instance language independence. * [Card] Pass the card search language to deck and zone card searches Wire the two remaining FilterString consumers to the configured card search language so card-name matches respect it everywhere: - DeckFilterString now takes the search language and mode, exposes them to its [[card name]] rule action via a thread-local parse context (same pattern as FilterString), and the engine's card database uses them for content search. - ZoneViewZone reads the card language from CardsDisplaySettings when applying its search filter, and the reveal-zone widget re-applies the active search when the language setting changes. - The deck-storage search re-runs its filter against the current card language setting, including live re-application when the setting changes. Game-action targeting (DlgMoveTopCardsUntil) intentionally keeps evaluating against English card names. * [Card] Rename CardSearchLanguage to SearchLanguageMode * [Card] Restore displaced namespace doc in card_localization.h * [Filters] Pass CardSearchLanguage as a single struct * [CardSearchModel] Match English and localized names in Both mode Card names are stored in both English and localized forms, so search for matches in both during the 'Both' search mode instead of checking only the localized name. * [CreateTokenDialog] Fetch cardsDisplay settings inside the apply lambda Avoid capturing the raw settings pointer in the lambda: resolve the card language and card search language from the settings cache at call time so the values are always current when the search language is re-applied. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-21 08:50:48 +02:00
BruebachL	6823d54c1e	[DeckShare] Browse and open public decks with loading, error and accessibility states (#7245) ... * [DeckShare] Browse and open public decks with loading, error and accessibility states Add a public-decks tab that lists decks published by other users using the server's deck visibility feature, previewing each deck's banner card, color identity, tags and upload time without downloading the deck list until the user opens it. - Add a public-decks tab with a shared-settings widget and a remote model that fetches the target user's decks and refreshes both automatically and on user request, with a loading indicator and a server-error message instead of a blank tab when the fetch fails or the connection drops - Render each deck as a focusable preview tile whose banner, color identity, tags and upload time follow the existing Preview settings, with the deck name announced as the tile's accessible name and Space/Enter opening the deck, mirroring the shared-deck preview tile - Show a message box when opening a public deck fails or arrives corrupted - Publish and unpublish decks from the server storage toolbar and context menu, toggling the deck's own visibility bit (what the server persists) rather than the inherited effective state, and batch the visibility refresh until the last in-flight change is acknowledged - Add the Show Upload Time setting so the tile's upload stamp can be hidden like the other preview details - Update the retranslateUi wiring for the new public-decks tab and rename the share action tooltip from "Deck share" to "Share link" * [DeckShare] Adapt deck upload to the server-derived banner and tag protocol The server now derives the banner card and tags from the uploaded deck list itself, so Command_DeckUpload only carries the client-computed color identity. Drop the reserved banner/tag setters from the editor and storage uploads, send the color identity on remote saves, and read tags from the now-repeated ServerInfo_DeckStorage_TreeItem field. * [DeckStorage] Refresh the visibility column with a guarded timer instead of a latch counter A dropped visibility reply used to leave the pendingVisibilityChanges counter permanently positive, so the Public/Private column never refreshed again and nothing reset it on disconnect. A restartable single-shot timer with a boolean guard re-reads the tree whenever publishes quiet down and is stopped on disconnect, so a lost reply costs one stale refresh instead of killing the column for the session. * [DeckStorage] Summarize batch publish failures when the batch drains Each rejected node stacked its own modal dialog, so publishing a ten-deck selection against a rejecting server made the user dismiss ten dialogs one at a time. Failures are now collected while the batch is in flight and shown as a single summary when the visibility refresh timer fires; a reply that lands outside an active batch still reports right away. * [PublicDecks] Time out the loading state so a dropped reply cannot wedge the tab loading only cleared in decksReceived, but the ping sweep can drop a pending command without ever emitting finished, leaving the tab stuck on 'Loading public decks...' and the refresh button permanently inert. A single-shot timer started per refresh clears the latch and reports a timeout; the latch also clears when the client disconnects. * [PublicDecks] Escape remote-crafted text in tooltips and the tab title Deck names and usernames come from other users' records and Qt renders QLabel tooltips as AutoText, so a name like '<h1><table>...' parsed as markup. Escape and bound the deck-name tooltip and escape the username interpolated into the title label. * [DeckStorage] Distinguish an inherited public state in the visibility column The column reported the effective state while publishing toggles the node's own bit, so a private deck inside a public folder already read 'Public' and toggling appeared to do nothing (and toggling again silently unpublished it). The cell now shows 'Public (inherited)' for that case and the tooltip explains why. * [PublicDecks] Run retranslateUi at construction and name the refresh button retranslateUi was never called from the constructor, so the tooltips set there were absent until a language change. Call it before the first refresh, and give the icon-only refresh button an accessible name for screen readers. * [PublicDecks] Keep the empty and status variants correct across language changes retranslateUi unconditionally rewrote the empty label to the 'nothing published' variant, stomping the 'no decks match your filters' choice rebuildGrid had made, and a visible loading message stayed in the old language. Let retranslateUi pick the same variant rebuildGrid does and re-show the status so it retranslates. * [VDS] Share one color-identity match rule between the two deck grids The remote public decks model verbatim-copied updateColorMatches' switch, down to the ExactMatch normalization and the fact that Includes/Excludes do not normalize case. Extract colorIdentityMatches() next to the FilterMode enum and call it from both so the subtle rule cannot drift. * [DeckStorage] Drop the unused tree widget model accessor The accessor handed the model out past the wrapper methods that exist to keep it encapsulated, and nothing in the stack called it. * [DeckShare] End the public-decks files with a trailing newline keeps the final line's diff clean and stops clang-format CI from flagging the files. * [VDS] Reuse the shared quick settings widget for the public decks tab PublicDecksQuickSettingsWidget was VisualDeckStorageQuickSettingsWidget minus the folders, banner and tooltip controls, with identical wiring for the shared keys and a version of the near-identical file to keep in sync by hand. Fold the Show Upload Time checkbox into the shared widget, give it a setPublicDecksMode() that hides the controls that do not apply, and delete the duplicate. * [PublicDecks] Drop stale deck-list replies after the loading timeout A reply that lands after its own loading timeout (the reverse of the ping sweep dropping the command) could stop the newer request's timeout timer and repaint the grid with out-of-date data. Each refresh now captures a monotonically increasing request id, and only the newest request's reply updates the grid. * [PublicDecks] Re-show a displayed failure message on language changes The status label carries both the loading and the failure message, and retranslateUi hid it whenever the model was not loading, so a language change while a server-error or timeout message was on screen swapped it for the (empty) grid. The tab now keeps the last failure text and re-shows it when not loading, clearing it once a new refresh starts. * [DeckStorage] Keep the visibility refresh armed until replies land The single-shot drain was armed with the 500 ms delay at send time, so a round trip slower than that drained before the server applied the change, re-read the old state and never re-armed, leaving the column stale until a manual refresh. The timer is now armed with the full network timeout at send time (a lost reply still costs one stale refresh) and re-armed with the short delay every time a reply lands. * [DeckShare] Close public decks tabs when the client disconnects TabSupervisor::stop() built tabsToDelete from the room and game tabs only, so a public decks tab survived a disconnect, sitting with stale contents and a refresh button that kept hitting the dead client. Its values are now folded into the same cleanup. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-20 20:22:17 +02:00
BruebachL	db2e159dca	[Game] Allow judges to enter any game regardless of restrictions (#7315) ... Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-19 10:39:52 +02:00
BruebachL	7b39cf98d5	[Themes] Add identity default palettes for the image themes (#7280) ... * [Themes] Add identity default palettes for the image themes The Fabric, Leather, Plasma and VelvetMarble themes shipped only zone art and no palette, so their chrome fell back to the bare OS palette. Each theme now ships light and dark curated defaults written in the palette editor's own conventions, with [AppColors] so the home-tab buttons keep the theme's identity over static backgrounds. - Add palette-default-light.toml / palette-default-dark.toml for the four themes; scheme resolution follows the OS since the themes declare no color scheme. - Tint the window/base/button chrome and bevels towards each theme's identity: navy cloth (Fabric), black-brown with brass accents (Leather), electric violet with cyan sparks (Plasma), charcoal velvet with slate marble (VelvetMarble). - Fill [AppColors] AccentStrong / AccentSoft per theme so the home-tab button gradient matches instead of falling back to the default greens. - Derive Disabled and Inactive groups with the same conventions as PaletteGenerator::fromAccent. * [Themes] Add home backgrounds for the image themes * [Themes] Ship Fusion style + scheme backgrounds for the image themes Address PR review: - Add theme.cfg ([Style] Name = Fusion, ColorScheme = System) to Fabric, Leather, Plasma and VelvetMarble so their curated palettes actually apply. Without it the native style paints button chrome from the OS (windowsvista on Windows has no dark mode), making the dark palettes' light ButtonText unreadable on light buttons. - Leverage the scheme-variant asset resolution: register home-dark.png and home-light.png in resources and add them to the qrc so the built-in home background also flips with the palette instead of staying static. - Fix Leather [Palette.Inactive] Accent, which copied the Active Highlight color instead of the theme accent (dark #4a5f8f -> #c9995a, light #34508c -> #a5712f) in both palette files. * [Themes] Align light plasma home background with the dark variant Recolored the Plasma light home background to mirror the dark one: brighter violet/cyan blooms, vivid azure spark arcs, and higher accent saturation against the same pale-lavender key. Brightness is unchanged so it still reads as a light scheme. * [Home] Add option to disable the home tab background dim Adds a 'Dim the home tab background' checkbox to the Home tab settings page (Appearance). When unchecked, HomeWidget skips the translucent black overlay it paints over the whole background. Default is on, preserving current behavior; the home tab repaints live on change. * Update leather backgrounds * Update velvet marble backgrounds * Update light plasma background * Update light fabric background * WIP [UI] Theme-aware onboarding banner with frosted light mode Banner colours now derive from palette tokens at ~60fps (tick-driven, equality-guarded setters) so scheme switches and live accent-picker previews apply instantly: - dark stages: byte-for-byte the original treatment (near-black stage from window hue, Highlight accent, white centre halo, vignette 0.62) - light stages: pastel accent-hue wash instead of a neutral grey copy, brightness-lifted accent for additive glow legibility, deep-Highlight halo (uGlowColor) instead of white blowout, gentler vignette (uVignetteMin 0.88) so corners don't go muddy - black logo silhouette variant selected on light stages - theme picker preseeded with brand green (brand_colors.h single source) WIP notes for next session: - real-pixel wizard screenshot check still pending (headless capture exists: Xvfb :77 + isolated XDG_DATA_HOME; shader vs fallback pixel analysis not finished) - user plans separately: promote Fusion to default theme, Default -> system * [Themes] Align Fusion accent tokens with the SVG brand gradient Align AccentStrong (#139740) and AccentSoft (#c9fd62) and the linked Link/Accent roles with cockatrice.svg's linearGradient4265-7-8 stops so the identity gradient used by the logo emulation matches the static art the icon shipped. * [Onboarding] Draw the banner logo as a static gradient plate Replace the black/white logo tint switch with a ShaderEffect plate that repaints the SVG's brand gradient (light AccentSoft -> dark AccentStrong along the baked-in userSpaceOnUse axis) clipped to the full-color logo's alpha silhouette, with the white highlight path overlaid on top — matching the home widget's QPainter composite. The plate is static: no glow or breathing. Brand colors flow from BannerShaderConfig's new brandStrong/ brandSoft pair instead of the removed logoDark flag, and the background motifs get a touch more accent so the mark keeps its coloured surround. * [Home] Draw the featured logo as a theme gradient composite Repaint cockatrice.svg in Qt instead of showing the baked-in static art: fill the full-color logo's alpha silhouette with the same brand gradient the banner plate uses (light AccentSoft grading to dark AccentStrong along the SVG's userSpaceOnUse axis), then overlay the white highlight path. Renders an explicit QPixmap so the mark stays crisp at the 200px display size, and re-seeds it on theme/palette/appearance changes so it never goes stale. * [Resources] Drop the unused black logo asset No consumer remains after the banner's logoDark toggle was replaced by the static gradient plate (unit-tested in d7529e0c6a), so remove cockatrice-logo-black.svg and its qrc entry. * [Onboarding] Seed the theme picker from the theme's identity accent Replace the hardcoded brand-green preseed with the shipped theme's own AccentStrong (Plasma seeds violet, Fusion green), resolved from the default palette so auto/user-generated palettes can't mask it, and re-seed whenever the theme changes so the swatch never goes stale. Also consult the shipped palette in maybeAutoGeneratePalette so scheme flips don't regenerate a fresh palette over curated theme colors. * Regenerate zone assets for Fabric, Leather, VelvetMarble themes Align zone textures with each theme's palette identity: - Fabric: linen weave in navy/steel tones (was generic green/blue/red/gray) - Leather: grain texture in warm brown/amber tones - VelvetMarble: smooth charcoal marble veining (replaces .jpg with .png) All zones now have dark + light scheme variants for OS color scheme adaptation. Plasma zones deferred to separate iteration. * Regenerate Plasma zone assets: horizontal-hand layout, seamless tiling Bake each zone's phase from its true world position in the horizontal-hand layout (player 0, stack 172, table 280, hand 172+406) so the diagonal sheen continues continuously across zone boundaries. All zones share one mathematically tileable (1,1) diagonal at frequency 3; playerzone gets a distinct blue-violet identity and the light scheme gets extra contrast. * Regenerate VelvetMarble zones: classic Perlin marble, seamless tiling * Regenerate VelvetMarble zones: domain-warped fractal veins, stone grain * Promote Fusion to default theme, rename Default to System Fresh installs and new profiles now default to the Fusion (dark) theme instead of the platform-native theme. The old "Default" theme is renamed "System" to better describe its purpose — using the OS-native Qt style (windowsvista, macOS, etc.). Existing users who had "Default" selected are automatically migrated to "System" so they keep their platform-native styling. Users with an empty or invalid theme name now fall back to Fusion. * Move checkbox. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-18 15:16:07 +02:00
BruebachL	cca4af0ec7	[PictureLoader] Add local override storage and resolution with matcher tests (#7311) ... * [PictureLoader] Add local override storage and resolution with matcher tests * [PictureLoader] Address review comments - Make deleteAllLocalOverrides static; it does not touch instance state - Drop the now-unused hasCustomArt dead code - Rename the override install methods to installPrintingOverride / installPrintingOverrideOnLoad * [Tests] Give loader matcher tests a writable HOME in CI Under GitHub's docker runner the process uid has no passwd entry, so HOME resolves to '/' and the test-mode qttest data dir cannot be created. SettingsCache's QSettings then drops every write, getPicsPath() comes back empty, and the loader searches a blank path. Point HOME at a QTemporaryDir for the duration of the run. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-18 15:03:08 +02:00
BruebachL	1c93309952	[Client] Show localized card names, texts and pictures (#7294) ... * [Client] Show localized card names, texts and pictures Localization wiring now runs end to end: the oracle importer collects foreignData for the configured language and the client renders it. - [Oracle] Import localized names and rules texts for the selected cardLang - single-face cards store their foreignData name and full text - multi-face (split/adventure/aftermath/prepare) cards collect the joined name once and join each face's translated text with the same separator as the English merge; an incomplete translation falls back to English; the joined text follows the same highest-priority-set policy as the single-face path and is only collected when localization is enabled - the wizard switching languages re-imports the card database - [Client] Display localized card info throughout the client - card info text/picture widgets and the game board re-render on language change - pictures resolve cardLang art through Scryfall's named endpoint using the localized name, falling back to id-based art when no match exists - deck editor keeps canonical English names as card identity (EditRole) while showing localized names (DisplayRole), so decks and wire names stay stable - [Card] Add CardLocalization-backed name/text lookup and cards.xml v4 localization elements with a bounded-size translation cache - [Tests] Cover oracle foreignData import (incl. multi-face joins, priority and fallback paths), XML v4 localization parsing, deck model localized display and the language-aware settings default Existing installations need to re-run Oracle to see translations: localized data only lands in cards.xml when the Oracle app is started with the preferred language selected — launch the separate "Oracle" program that ships with Cockatrice, pick the language in the wizard and let it re-import the card database. The client's database cache (cards.xml.cache) is invalidated by the cache format bump and the source-hash checks, but a cache written before the re-import can still hold English-only entries (the hash uses file size and mtime, so a same-size/same-timestamp rewrite may be served as-is); delete cards.xml.cache and relaunch if no localized names/texts show up after re-importing. * [Card] Pass localized card names and texts into CardInfo construction Address review: instead of constructing the card and then calling setLocalizedName/setLocalizedText (which emit a cardInfoChanged signal per language), both constructors, both newInstance overloads and their callers (cards.xml v4 parser and the binary cache reader) now pass the localized maps as constructor arguments. * [Client] Rename LocalizedCard:: helpers namespace to CardLocalization The namespace now matches its header file name, as the review pointed out; LocalizedCard reads more like a class or struct. Callers (card info text widget, board card name rendering) are updated to match. * [Client] Drop unused info member from the card info text widget The CardInfoPtr member was only ever initialized to nullptr and never read; remove it together with its initializer. * [PictureLoader] Add the localized picture URL explicitly, not implicitly Address review: silently prepending the Scryfall named-picture URL to the download list whenever a non-English card language was active was surprising, consumed quota per card when it failed, and could grab the wrong (canon) art on name collisions, with no way to turn it off. The insert is now opt-in and user-controlled: changing the card language adds the template to the top of the download URLs once (persisted, documented in the re-import prompt, and editable/removable in the deck editor settings), while the picture loader no longer injects it at request time. * [Card] Show card languages in the same native (English) format as the UI Address review: the card text & images language dropdown listed bare native names, some in inconsistent lowercase (e.g. "čeština", "español de España"), which makes the languages easy to mix up for users that do not read the script (e.g. 日本語 vs 한국어). It now mirrors the UI language dropdown and always pairs the native name with its English name (e.g. "Deutsch (German)", "日本語 (Japanese)"), using the same fixed casing. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-18 12:03:07 +02:00
BruebachL	202a5ac958	[Client/Server/Protocol] Surface live metrics in the Developer tab (#7212) ... * [Server] Instrument command processing, game starts, and event loops Add a lock-free MetricsRegistry that accumulates per-command processing times in preallocated histogram slots (one per protobuf command type, bucketed at 1/5/10/25/50/100/250/500/1000/2500/5000 ms +Inf). The hot-path observeCommand() uses only relaxed atomic adds — no locks, no allocations, no cache-line ping-pong beyond the unavoidable counter updates. Wire the registry into AbstractServerSocketInterface::processCommandContainer() so every processed command is attributed with its container's wall-clock time. When a container exceeds metrics/slow_command_ms (default 500), a warning is logged including the connected username. Add an EventLoopWatchdog heartbeat that runs on every socket pool thread. If a heartbeat overshoots metrics/stall_warn_ms (default 2000 ms), the overshoot is recorded in atomic counters and a warning is logged. Both thresholds are configurable in servatrice.ini; setting stall_warn_ms to 0 disables the watchdogs entirely. Track game-start durations via a separate histogram in MetricsRegistry. Server_Game::startGameNow() measures the time from zone creation through player materialization and reports it via Server::observeGameStartDurationMs(). Add a live card-count gauge: Server_Game exposes getCardsInGame() and Servatrice::getCardsInGamesTotal() sums across all running games under the appropriate read locks. Include a standalone metrics_registry_test (Google Test) that validates empty registries, single/multi-sample histograms, kind encoding, overflow-slot collapse, negative-duration clamping, gauge rendering, and the game-start histogram separation. Took 10 minutes * [Client/Server/Protocol] Surface live metrics in the Developer tab Extend Response_GetServerStats with live counters from the in-process MetricsRegistry: cards in games, event loop stall totals/worst, total commands processed, average command time, active command types, and game-start count/duration. Add a repeated CommandStats message carrying per-command breakdowns (kind, extension number, resolved protobuf name, count, total ms) for every type that has seen at least one sample. Server-side cmdGetServerStats() populates all new fields after the existing DB uptime snapshot query, resolving protobuf extension names via the descriptor pool for human-readable labels like session/Command_Ping. Expand TabDeveloper with two tables: an overview section (existing DB stats plus the new live metrics) and a per-command breakdown table (Command / Count / Total ms / Avg ms) sorted by total_ms descending so the hottest commands surface first. Took 55 minutes Took 47 seconds * [Server] Drop dead Prometheus histogram, add developer command metrics, fix watchdog init order - metrics_registry: remove toPrometheusText/appendCumulativeBuckets and the time-bucket histogram that nothing in production ever emitted (the future /metrics exporter can bring it back); keep counts/totals read by the Developer tab - Fix +Inf bucket routing that never incremented, and its test that locked the bug in - Instrument developer_command container (kind 6) in processCommandContainer and stats label resolution - Read metrics/{slow_command_ms,stall_warn_ms} at the top of initServer() so stall_warn_ms=0 disables the watchdogs before pool threads start - Shrink KindStride to 1280 (largest extension in use is 1206) with a static_assert; document scrape cost of getCardsInGamesTotal; note slow_command logging has no rate limit in servatrice.ini.example * [Tests] Give metrics_registry_test an explicit main * [Server] Record only the dispatched command family; drop unused totals processCommandContainer recorded every family in a container even though the base if/else-if dispatch processes at most one. An unauthenticated client could batch a session command (login) with fabricated developer, moderator, and admin entries and forge genuine-looking samples that were never executed or authorized. Mirror the base's selection, skip when the handler was already deleted, and skip entries whose extension number is -1 (which would otherwise wrap into the previous kind's id range). [Server] Drop dead process-lifetime byte/uptime counters txBytesTotal/rxBytesTotal added an atomic RMW to every socket write and read for counters nothing consumes (cmdGetServerStats fills tx_bytes, rx_bytes, and uptime_secs from the DB snapshot). Remove the two atomics and the getTxBytesTotal/getRxBytesTotal/getUptimeSeconds getters; the incTxBytes/incRxBytes slots and mutexes remain for the ISL legacy counters. [Protocol] Document kind 5 as developer in CommandStats NumKinds is 6 and the server emits kind_index = 5 for developer commands; the comment stopped at 4. * [Client] Togglable auto-refresh for Developer stats tab * [Oracle] Fix clang-format alignment of card type priority list --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-11 17:18:56 +02:00
BruebachL	d5d99e4dfb	[Server/Client/Protocol] Add developer staff role (#7211) ... * [Server/Client/Protocol] Add developer staff role Introduce a Developer staff level (proto flag 32, DB admin bit 8) that sits between admin and moderator: no kick/ban/warn/report/admin powers, but gets server log access via a new developer command container family (GET_SERVER_STATS, VIEWLOG_HISTORY) and an idle-timeout exemption. - Protocol: IsDeveloper flag, developer_commands.proto envelope, Command_GetServerStats/Command_GetLogHistory, Response_GetServerStats, Command_AdjustMod.should_be_developer - Servatrice: fail-closed developer dispatcher, uptime snapshot handler, shared log history handler reuse, bit-8 DB mapping - Client: burgundy pawn/badge/labels/sort order, prepareDeveloperCommand, minimal Developer stats tab, log tab access, promote/demote actions Took 24 minutes Took 18 seconds * [Server/Client/Protocol] Address developer role review feedback Address ZeizaZach's review of the developer staff role: - Nudge the developer log query to exclude private chat and sender IPs (the ModeratorCommand path still sees everything). - Deduplicate Command_GetLogHistory into Command_ViewLogHistory, which now extends both ModeratorCommand (ext) and DeveloperCommand (dev_ext); the client picks the DeveloperCommand-scoped extension by extendee, and the server reads it via the extension number. - Pull the uptime snapshot SQL into Servatrice_DatabaseInterface as getLatestUptimeSnapshot() and widen the reported counters to 64-bit. - Document the admin bitfield (1 admin, 2 moderator, 4 judge, 8 developer) and add a server-side test for the developer command path. * Add missing trailing newline to user_context_menu.cpp * Remove stale includes of deleted command_get_log_history proto The Command_GetLogHistory message was folded into Command_ViewLogHistory, which deleted command_get_log_history.proto, but serversocketinterface still #included its generated header. Fresh CI builds fail on the missing file; local builds masked it by reusing a previously generated header. * [Server] Exclude chat rows when private-chat filter is bypassable A developer who omits log_location entirely — or sends only "chat" — leaves chatType, gameType, roomType all false, so getMessageLogHistory skips the target_type clause and returns every row, private messages included. When !allowPrivateChat the server now forces game+room when no surviving location was requested, guaranteeing the query always carries a target_type restriction. [Client] Demote mod+dev to moderator path in log-tab dispatch The developer command family is strictly weaker than the moderator one (no private chat, no sender_ip, ip filter ignored), so granting the developer bit to an existing moderator must not silently strip their capabilities. useDeveloperCommands is now true only when the user holds the developer bit and not the moderator bit. [Client] Hide the IP-address filter for developer log tab users The developer path ignores the ip_address query field server-side. Showing the field lets a developer type an IP and get results that are silently unfiltered by it rather than an empty result set — reads as a broken filter. Hide labelFindIPAddress/findIPAddress alongside the privateChat checkbox. * Developer pawn is silver. * [Client] Fix indentation of merged Card Art Rules / Developer tabs --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-11 17:18:56 +02:00
BruebachL	7d867b9745	[Oracle/Client] Report card database download progress (#7253) ... * [Oracle/Client] Report card database download progress Card database updates ran invisibly: MTGJSON parsing spun an indeterminate bar on the UI thread and the set import blocked the window, while the onboarding wizard spawned `oracle -b` with no progress to show at all. - Add byte-level scan progress to `RawJson::scanSetRanges` via an optional callback, throttled to ~100 reports per scan. - Emit `OracleImporter::dataReadProgress` during the scan and import sets on a worker thread, driving the wizard's progress bar per set. - With `-b`, write machine-readable `PROGRESS <stage> <done> <total>` lines to stdout for the download/scan/import stages; stderr keeps the log output. - Parse the oracle stdout in `MainWindow` and forward it to the onboarding wizard, giving the card database step a determinate bar with stage-specific status text. - Guard the async workers against the wizard being closed mid-run. - Add Google Test coverage for scan progress reporting. * [Oracle/Client] Harden oracle progress workers and quit prompt Address review feedback on the download-progress change: decompress and read sets files off the UI thread, cancel the load/import workers before the wizard can tear down the importer, and show an 'Extracting file...' status plus a clean 100% tail so the poll never looks stuck. Quitting Cockatrice while a card database update runs now asks for confirmation. * Show 100% for 500ms on complete. * Disable buttons on set import until done. * Clean up progress bar. * Drop wrapper around lambda --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-09 22:27:16 +02:00
BruebachL	ebeae48652	[Game] Preserve hand card order when restoring connection (#7266) ... When a player restores connection to a game, the client rebuilds each zone from the cards the server sends in the game state. Non-coordinate zones (hand, piles, stack) report x == 0 on every card, so inserting each rebuilt card at that coordinate reversed the received order one card at a time. Append rebuilt cards in the order they arrive for zones without coordinates; coordinate-based zones (table) keep using x/y. Fixes #2759 Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-07 08:21:44 +02:00
BruebachL	760fe88fa3	[Client] Add setting to ignore all private messages (#7260) ... * [Client] Add setting to ignore all private messages (#1250) * Early return --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-07 08:20:46 +02:00
BruebachL	9677fad342	[Client] Add zone management to the deck state manager (#7204) ... * [Client] Add zone management to the deck state manager State-layer operations for custom deck zones, plus the shared prompt dialog that later editor menus will call into. - moveCardToZone relocates every copy of a card row into any zone, refusing non-card rows and tokens so miswired selections can never shred a group or turn tokens into deck cards. The current zone is found by walking ancestors, which also handles legacy top-level zones. - createCustomZone, renameCustomZone, moveCustomZone and removeCustomZone wrap the tree API with memento history, model rebuilds and deck hash refreshes via modifyTree. - Same-board zone moves return success without minting a history entry, keeping the undo log honest. - promptForNewZone asks for a name and the parent zone, keeps Ok disabled until the trimmed name passes a caller-supplied validator (shown inline as an error), and reports its own translation context. Took 14 minutes # Commit time for manual adjustment: # Took 6 minutes # Commit time for manual adjustment: # Took 33 seconds * [DeckEditor] Address zone-management review feedback - Expose DecklistNodeTree::hasZoneName and use it in validateNewZoneName so the uniqueness scan covers custom zones on every board, not just the standard ones. - Hide the board selector in the rename dialog path where it is not used. - Emit deckHashChanged after refreshDeckHash so the deck hash label stays current after zone create/rename/move/remove. * [DeckEditor] Notify card set changes after zone edits and drop the board scan - modifyTree emits cardNodesChanged alongside deckHashChanged so the banner-card combo and printing in-deck counts refresh after removing a zone that still holds cards - DecklistNodeTree::findCustomZoneByName is public and moveCustomZone uses it, locating zones under non-standard boards (e.g. tokens) instead of scanning only main/side/maybeboard --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-05 22:00:21 +02:00
BruebachL	e8ec28572f	[Models] Mirror custom deck zones in the deck list model (#7203) ... * [Models] Mirror custom deck zones in the deck list model DeckListModel now surfaces the custom zones from the deck tree so views can render and edit them alongside criteria groups. The custom-zone bookkeeping that made the model unwieldy is extracted into DeckListModelCustomZones (deck_list_model_custom_zones.h/.cpp), a single self-contained unit owning every "what is / where is a custom zone" decision for the model's shadow tree: - rebuildTree mirrors each custom zone as a DecklistModelSubZoneNode under its board zone, cards flat inside (no further grouping). - The freshly built shadow tree is sorted while the model reset is still open, so views never observe unsorted intermediate order and proxies cannot desync. - Custom zones always sort after criteria groups within a board, regardless of their names. One shared sortWithCustomZonesLast backs both the live sortHelper (which remaps persistent indexes from the movement mapping) and the silent reset-time sortShadowTree. - addCard inserts flat into a custom zone by name and keeps grouping by active criteria for board zones. findCardNode resolves cards in both layouts, legacy top-level zones unchanged. - New IsCustomZoneRole lets views tell zones apart from groups. - Empty custom zones survive row removal. Zone rows themselves are only mutable through the deck tree API. A new deck_list_model_custom_zones_test suite locks the extracted shadow-tree logic (type testing, mirroring, name lookup, and the sort-with-custom-zones-last mapping). No behavior change. * [Models] Route group lookups around mirrored custom zones Group lookups (createNodeIfNeeded, findCardNode) must not resolve a mirrored custom zone that shares the group name. Introduce findGroupChild to search only non-custom children, and make addCard consult the deck tree before falling back to creating a top-level zone so cards added to an un-mirrored custom zone land inside it. mirrorCustomZones now flattens cards nested at any depth into the mirrored zone so no card is left without a model row. Add model behaviour tests (addCard routing, same-name group/zone collision, removeRows guard, empty-zone survival, findCard inside a custom zone) and fix the missing main() in the unit test binaries. * [Models] Fix addCard routing for card-named zones and nested custom zones - hasDeckZone no longer matches board cards that merely share the zone name, which previously caused infinite addCard/rebuildTree recursion - Adding to a custom zone whose deck side holds nested sub-zones appends to the deck tree instead of writing past its direct children --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-05 22:00:20 +02:00
BruebachL	c011ea7ceb	[Oracle] Parse sets lazily to slash importer peak memory (#7217) ... * [Oracle] Parse sets lazily to slash importer peak memory - Add a raw JSON scanner that splits the document into per-set byte ranges without materializing the JSON tree - Keep only the raw document bytes and parse one set at a time in startImport - Take readSetsFromByteArray by value so the wizard's buffer is moved, not copied - Clear the retained raw data in releaseSetData()/clear() - Cover the scanner and lazy parsing with tests Took 2 minutes * [Oracle] Fix nesting-depth cap, tolerate unescaped control chars, lazy-parse review fixes --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-05 20:35:28 +02:00
BruebachL	1dc54617ba	[Oracle] Add RAM usage benchmarks for the oracle importer (#7216) ... * [Oracle] Add RAM usage benchmarks for the oracle importer - Measure process peak/current RSS via procfs (Linux) or getrusage (macOS) - Add a synthetic-scale RAM benchmark and an opt-in real AllPrintings run gated by COCKATRICE_ORACLE_RAM_BENCHMARK=1 - Mirror the wizard's magic-byte handling to decompress .xz/.zip payloads - Wire optional ZLIB/LibLZMA into the benchmark target and raise its timeout Took 2 minutes * [Oracle/Tests] Measure parse against post-fixture baseline; assert release empties sets --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-05 20:35:27 +02:00
BruebachL	61e6a9913e	[Oracle] Add oracle importer tests and fix set parsing details (#7215) ... * [Oracle] Add oracle importer tests and fix set parsing details - Add oracle_importer_test and oracle_importer_benchmark_test targets - Preserve the first printing's legalities when an existing card is reused - Concatenate split-card coloridentity and sort/dedupe card colors - Use a raw string for the Basic Land format regex - Pre-allocate the card hash and micro-optimize string handling Took 2 minutes * [Oracle/Tests] Pin cmc coercion in CI run; scope the reserve pass The #7214 coercion assertion lived only in oracle_importer_benchmark_test, which gets no add_test and so never runs under ctest. Add NumericManaValueCoercedToCmc and LegacyConvertedManaCostCoercedToCmc to oracle_importer_test (a CI-ran binary): manaValue/convertedManaCost are JSON numbers in AllPrintings, and QJsonValue::toString() would drop them to an empty cmc without the #7214 coercion fix. Wrap the distinct-name reserve pass in a bare block so the ~35k name QStrings are handed back before the memory-heavy import loop starts. --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-04 22:14:16 +02:00
BruebachL	d6fbfb32a1	[Security] Use a CSPRNG for salts, tokens, and RNG seeding (#7192) ... * [Security] Use a CSPRNG for salts, tokens, and RNG seeding Password salts and activation tokens were generated with the global SFMT RNG, which was seeded from a 32-bit timestamp, making registration salts and activation tokens predictable. The game RNG used the same timestamp seed across restarts. Add CryptoUtil backed by OpenSSL RAND_bytes and use it for salt/token generation and to seed RNG_SFMT with a 64-bit CSPRNG value in both the client and server. Link libcockatrice_utility against OpenSSL::Crypto. Took 30 seconds Took 25 minutes * Lint. Took 4 minutes Took 36 seconds --------- Co-authored-by: Lukas Brübach <Bruebach.Lukas@bdosecurity.de>	2026-09-04 13:49:18 +02:00