comic-dl v0.0.4

[v0.0.4] - 2026-09-27

Added

• Ten new built-in sources — DivaScans, HD Porn Comics, IMHentai,
  KodokuEasyAccess, ManhuaTo, Nyx Scans, Tapas, ToonVerse, ValirScans, and
  Vortex Scans — taking built-in support from 21 to 29 sites.
• First-run legal notice.
• Stable site_error_code failure classification across scrapers, surfaced
  in errors and batch output.
• Per-page download state manifest with a resume-safe retry pass; partial
  chapters tracked separately in batch footers.
• Connectivity probe before blaming the network for failures.
• Library schema v4: one global database with per-row output roots, global
  library commands, and library update --dry-run.
• Leading m. mobile subdomains normalized away before matching.

Changed

• KodokuStudio replaced by KodokuEasyAccess (new domain).
• System-webview solver off by default.
• Archive writes stream to disk instead of buffering whole image sets.
• Website design refreshed; dev guides rewritten as end-to-end processes.

Removed

• ManhwaZ support dropped.

Fixed

• Kagane reports the Cloudflare challenge instead of a network error.
• Batch output honesty: Saved: gated on success, per-page failure reasons,
  chapter/partial accounting; single-URL polish (short chapter titles, one
  completion line).
• Tapas cover/creator/description; Asura Scans full series summary; GenzToons
  listing guard, locked chapters, metadata cleanup; Vortex/Nyx cover,
  synopsis, genres.
• Cookie collision resolution, charset-aware cache decode, cookie-replay
  probing with a JSON cache-poisoning guard; .partial markers restored to
  on-disk page names so reruns skip them.

Security

• Cookie jar encrypted at rest.
• Outbound fetches fail closed on DNS/validation errors; rebinding window
  narrowed.
• 2xx media payloads never treated as antibot challenges; DataDome markers
  tightened.

Full Changelog: v0.0.3...v0.0.4