Strip every comment from the source (parsed with the TypeScript compiler so
strings, template literals, regex literals and JSX are never touched), and drop
internal/working artifacts that do not belong in the public repository
(design mockups, internal analysis docs, a stray backup file and an old log).
No functional change: build is green, the full test suite passes.
User-Report: ZIP scheitert mit wrong_password -> Passwort zur Liste hinzufügen
+ Settings speichern -> "Jetzt entpacken" scheitert WIEDER -> erst nach
App-Neustart klappt "Jetzt entpacken".
Analyse (eigenständig + zweiter Analyse-Agent, gesamte TS+Java-Kette):
Renderer-Save -> updateSettings (fingerprint enthält archivePasswordList) ->
setSettings (this.settings aktualisiert) -> extractNow -> runPackagePostProcessing
-> extractPackageArchives (passwordList: this.settings.archivePasswordList) ->
archivePasswordCandidates (enthält neues PW) -> Daemon bekommt Passwörter PRO
REQUEST. Java-Daemon probiert alle PW frisch, kein Per-Archiv-Cache; zip4j/
sevenzipjbinding ohne relevanten static State. Alle Pfade propagieren die neue
Liste korrekt — statisch ist KEIN Bug auffindbar.
Verifikation per Ausschluss: die EINZIGE zustandsbehaftete Komponente, die ein
App-Neustart zurücksetzt und ein Settings-Save NICHT, ist der langlebige
JVM-Daemon-Prozess (+ der In-Memory Learned-Password-Cache). Der User bestätigt
empirisch, dass ein Neustart (= frischer Daemon) es fixt.
Fix: neue resetExtractorCachesForPasswordChange() repliziert den Neustart-Effekt
am Extractor-Subsystem — bei Änderung der Passwortliste in setSettings wird der
Learned-Password-Cache geleert und der idle JVM-Daemon heruntergefahren, sodass
die nächste Extraktion frisch mit der neuen Liste startet. Beschäftigter Daemon
wird nicht abgebrochen (laufende Extraktion bleibt unangetastet).
Diagnostik: setSettings loggt jetzt PW-Anzahl + Reset-Resultat. Zusammen mit den
bestehenden "Archiv-Passwortliste: passwordCount=N"-Logs lässt sich bei erneutem
Auftreten in 30s unterscheiden, ob das PW beim Extractor ankommt (H2, Fix greift)
oder nicht (H1, dann TS-upstream). Best-bet per Ausschluss + Logs zur finalen
Bestätigung beim nächsten Repro.
621 Tests grün, tsc-Fehlerzahl unverändert (9 pre-existing).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
collectArchiveCleanupTargets() now includes companion files (.sfv, .nfo,
.md5, .sha1, .sha256, .crc, .srr) that share the same base stem as the
archive parts. Previously these were left as orphans after archive cleanup.
Applies to all archive types: multipart RAR, single RAR, ZIP, split ZIP,
7z, and split 7z.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Dynamically compute JVM -Xmx based on system RAM instead of hardcoded 32g
- Reduce peak memory during session loading (inline JSON string, skip redundant clone)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Retry extraction with 2.5s delay on CRC/password errors (Windows file handle race)
- Improve auto-recovery: force re-download when known password fails (content corruption)
- Expand auto-recovery to wrong_password category for encrypted RAR5
- Add fsync after download for pre-allocated files
- Fix permanent extraction failure loop for archives with valid headers but corrupt content
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Switch JVM GC from SerialGC to G1GC with MaxGCPauseMillis=50
for shorter pause times in long-lived daemon
- Increase JVM heap from 512m/8g to 1g/32g to reduce GC pressure
on systems with plenty of RAM
- Raise hybrid thread cap from floor(cpuCount/2) capped 8
to ceil(cpuCount*0.75) capped 12 — downloads are I/O-bound
and don't need much CPU headroom
- Refresh daemon process priority before each request so hybrid
extraction gets correct CPU priority even if daemon was started
with a different priority level
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
If the primary extractor (7-Zip) fails with wrong_password/checksum
error on a .rar file, automatically try the alternative extractor
(UnRAR/WinRAR) which handles RAR format natively and more reliably.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Archives with absolute internal paths (e.g. scene groups storing full
Windows paths) fail all password attempts in normal mode at ~98%, then
succeed only after the flat-mode fallback kicks in. Previously every
archive in such a package wasted all password cycles before discovering
flat mode was needed again.
Now the first successful flat-mode extraction sets a package-level flag
so subsequent archives skip the normal loop entirely and go straight to
flat-mode extraction, saving ~4x password attempts per archive.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Retry failed wrong_password archives serially after parallel extraction
to recover from CRC mismatches caused by concurrent UnRAR I/O contention
- Stop resetting sessionDownloadedBytes on start/resume so the session
total accurately reflects all bytes downloaded since app launch
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- extractor: detect UnRAR "Cannot create...\..." error (archive with
leading-backslash internal paths) and retry in flat mode (-e) which
strips all paths and avoids the invalid double-separator on Windows
- types/download-manager: add providerLabel field to DownloadItem,
store full label (e.g. "Debrid-Link #1") set at unrestrict time
- App: display providerLabel in Service column (falls back to generic
provider name if label not yet set)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- JVM extractor now supports --daemon mode: starts once, processes
multiple archives via stdin JSON protocol, eliminating ~5s JVM boot
per archive
- TypeScript side: daemon manager starts JVM once, sends requests via
stdin, falls back to spawning new process if daemon is busy
- Fix extraction progress caching: replaced Object.create(null) + in
operator with Set<string> + linear Array scan — both Map.has() and
the in operator mysteriously failed to find keys that were just set
- Daemon auto-shutdown on app quit via shutdownDaemon() in before-quit
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Replace Map-based archive item cache with plain Object.create(null)
to work around mysterious Map.has() returning false despite set()
being called with the same key — this caused resolveArchiveItems
to run on every 1.1s pulse instead of being cached, preventing
extraction progress (Entpacken X%) from ever showing in the UI
- Apply same fix to both hybrid and full extraction paths
- Increase JVM heap from 512MB to 1GB for better extraction throughput
- Use SerialGC for faster JVM startup on short-lived extract processes
- Add download lifecycle logging (package add + item download start)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Change OS priority from IDLE/BELOW_NORMAL to NORMAL/BELOW_NORMAL so
extraction runs at full speed (matching manual 7-Zip/WinRAR performance)
- Use "high" priority in both hybrid and full extraction paths
- Increase hybrid extraction threads from hardcoded 2 to dynamic
calculation (half CPU count, min 2, max 8)
- Fix emitState forced emit being silently dropped when a non-forced
timer was already pending — forced emits now always replace pending
timers to ensure immediate UI feedback during extraction transitions
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Release post-process slot immediately after main extraction completes.
All slow post-extraction work (nested extraction, auto-rename, archive
cleanup, link/sample removal, empty directory cleanup, MKV collection)
now runs in background via runDeferredPostExtraction so the next package
can start unpacking without delay.
- Export hasAnyFilesRecursive, removeEmptyDirectoryTree, cleanupArchives
from extractor.ts for use in deferred handler
- Import removeDownloadLinkArtifacts, removeSampleArtifacts from cleanup
- Expand runDeferredPostExtraction with full post-cleanup pipeline:
nested extraction, rename, archive cleanup, link/sample removal,
empty dir tree removal, resume state clearing, MKV collection
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Fix ~70 issues across the entire codebase including security fixes,
error handling improvements, test stabilization, and code quality.
- Fix TLS race condition with reference-counted acquire/release
- Bind debug server to 127.0.0.1 instead of 0.0.0.0
- Add overall timeout to MegaWebFallback
- Stream update installer to disk instead of RAM buffering
- Add path traversal protection in JVM extractor
- Cache DdownloadClient with credential-based invalidation
- Add .catch() to all fire-and-forget IPC calls
- Wrap app startup, clipboard, session-log in try/catch
- Add timeouts to container.ts fetch calls
- Fix variable shadowing, tsconfig path, line endings
- Stabilize tests with proper cleanup and timing tolerance
- Fix installer privileges, scripts, and afterPack null checks
- Delete obsolete _upload_release.mjs
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Show "Entpacken vorbereiten..." while scanning archives and checking disk space
- Show "Archive scannen..." and "Speicherplatz prüfen..." phases from extractor
- Use dash separator in UI: "[10/10 - Done] - Entpacken 45% (3/6)"
- Handle new "preparing" phase in both hybrid and full extraction progress handlers
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
