V0.1.1

[arumes31]

Summary by CodeRabbit

• New Features

  • Added configurable upload_workers setting to control concurrent uploads per sync pair
  • Automatic cleanup of empty encrypted parent folders on deletion
  • Enhanced status display showing active upload/download activity

• Improvements

  • Improved sync diagnostics logging capture
  • Added network timeouts for better reliability
  • Optimized concurrent folder operations to reduce redundant requests
  • Enhanced error reporting during sync operations

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: e8bb691b-f5c8-4898-aa4e-af802c3b01ed

📥 Commits

Reviewing files that changed from the base of the PR and between 3bb03e8 and dc0390d.

📒 Files selected for processing (6)

• README.md
• cmd/gcrypt/main.go
• internal/config/config.go
• internal/drive/client.go
• internal/service/fyneui.go
• internal/sync/engine.go

---

📝 Walkthrough

Walkthrough

Adds a configurable upload_workers field to AppConfig, removes the sync.Mutex from the Drive Client (replacing it with a properly configured HTTP transport and documented concurrent-safety), and updates the sync engine to use singleflight for folder resolution, derive worker count and rate-limit from config, prune empty encrypted parent folders after deletion, and emit structured warnings on retry. The UI status label now distinguishes Uploading/Downloading during transfers, and slog output is redirected to the app logger.

Changes

Concurrent Uploads and Drive Sync Improvements

Layer / File(s)	Summary
Config contract and Drive client concurrent-safety internal/config/config.go, internal/drive/client.go	AppConfig gains UploadWorkers int; Client removes mu sync.Mutex, documents concurrent safety, and NewClient builds a custom http.Transport with bounded timeouts and a background-context OAuth refresh. All Drive operation methods drop mutex lock/unlock calls.
Engine: configurable workers, rate limiting, and singleflight field internal/sync/engine.go	Engine gains a folderGroup singleflight.Group field. NewEngine reads UploadWorkers from config with a default fallback, derives reqPerSec capped by maxRequestsPerSec, initialises the rate-limiter ticker from that value, and emits slog.Warn on operation failures before retry.
Engine: singleflight folder resolution and empty-folder pruning internal/sync/engine.go	resolveRemoteDir fast-paths via cache and wraps resolve/create in folderGroup.Do; deleteRemote prunes empty encrypted parent directories; lookupRemoteDirID uses a paginated findChildFolderID helper; cache access is narrowed to brief lock windows.
UI transfer-direction label and slog routing internal/service/fyneui.go, cmd/gcrypt/main.go	refreshSummary accepts appstate.State and sets the status label to "Downloading…" or "Uploading…" based on active/pending transfers and smoothed rates. main.go routes the slog default logger to the application logger's writer at Info level.
README: upload_workers and encrypted folder cleanup docs README.md	V2 config example adds upload_workers with rate-limit/retry context; Remote Layout section updated to state gcrypt automatically removes empty encrypted parent folders after deletions.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~60 minutes

Poem

🐇 Hop hop, the mutex is gone!
Concurrent uploads carry on,
Singleflight deduplicates the trail,
Empty folders? They no longer prevail.
The slog now writes where it should be—
Rate-limited hops, fast and free! 🌿

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch v0.1.1

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}