fix(orphanscan): use content_path to prevent false positives when Auto TMM changes save_path

[nitrobass24]

Summary

• When Auto TMM updates save_path to the category default without moving files, content_path still reflects the real location on disk
• Orphan scan now derives the actual save path by stripping the file name suffix from content_path, registering file entries for both paths
• Handles both single-file and folder torrents correctly (folder torrents include the folder prefix in the file name from qBittorrent's API)
• Aligns with the approach already used in managedDeleteCleanupDir for delete cleanup

Test plan

• Added TestBuildFileMapFromTorrents_ContentPathDivergesFromSavePath
• Verified with 60TB cross-seed setup — Auto TMM torrents with diverging save_path/content_path no longer flagged as orphans
• Both single-file (hash-suffixed folders) and folder torrents (tracker-name subdirs) confirmed working

Closes #1699

Summary by CodeRabbit

• Bug Fixes
  • Improved file scanning accuracy for torrents with alternate content paths. The system now correctly identifies files when torrent content is organized in tracker-specific directories separate from the category root.

[coderabbitai]

Note

Reviews paused

It looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the reviews.auto_review.auto_pause_after_reviewed_commits setting.

Use the following commands to manage reviews:

• @coderabbitai resume to resume automatic reviews.
• @coderabbitai review to trigger a single review.

Use the checkboxes below for quick actions:

• ▶️ Resume reviews
• ✅ Review completed - (🔄 Check again to review again)

Walkthrough

Modified buildFileMapFromTorrents to derive an actual file location from content_path when it diverges from the reported save_path. The function now adds entries for both paths when Auto TMM updates save_path without moving files. Added a test validating this scenario.

Changes

Cohort / File(s)	Summary
Production Logic internal/services/orphanscan/service.go	Enhanced buildFileMapFromTorrents to derive actualSavePath from content_path by trimming the first file's name suffix. When this path differs from both content_path and normalized save_path, adds it to scan roots and populates file entries, ensuring files are recognized despite Auto TMM path changes.
Test Coverage internal/services/orphanscan/settling_test.go	Added TestBuildFileMapFromTorrents_ContentPathDivergesFromSavePath to verify buildFileMapFromTorrents correctly handles torrents where SavePath diverges from ContentPath, validating file map entries and scan roots from both locations.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• PR #1193: Refactors buildFileMapFromTorrents and surrounding flow in the same file, directly overlapping with this function enhancement
• PR #1197: Introduces scan-root derivation logic that depends on accurate torrent path computation, complementing this fix's scan-root accuracy improvements

Suggested labels

bugfix, orphan, backend, tests

Poem

🐰 Ah, files that dance 'tween two paths did hide,
Content whispers true while SavePath lied,
Now both shall guide us to where they reside,
No orphans false — just the truth clarified! ✨

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: using content_path to fix false positives in orphan scanning when Auto TMM modifies save_path without moving files.
Linked Issues check	✅ Passed	The changes fully address issue #1699 by modifying buildFileMapFromTorrents to register file entries from both save_path and the actual on-disk location derived from content_path, preventing false orphan detections.
Out of Scope Changes check	✅ Passed	All changes are directly scoped to fixing the orphan scan false positive issue: one production change in service.go and one targeted test in settling_test.go covering the divergence scenario.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/orphanscan-content-path-false-positives

---

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.}

[coderabbitai]

🧹 Nitpick comments (1)

internal/services/orphanscan/service.go (1)

1170-1175: Remove title-specific debug branches from orphan scan logic.

Two hardcoded title checks ("True.Lies.1994") add non-functional, data-specific debug logging that should be removed per coding guidelines. These branches do not enforce behavior mechanically and serve only as documentation.

• service.go (lines 1170–1175): Remove the strings.Contains(entry, "True.Lies.1994") branch and its debug log.
• walker.go (lines 206–212): Remove the strings.Contains(path, "True.Lies.1994") branch and its debug log.

Removal examples

For service.go:

 				for _, f := range files {
 					entry := normalizePath(filepath.Join(actualSavePath, f.Name))
 					tfm.Add(entry)
-					if strings.Contains(entry, "True.Lies.1994") {
-						log.Debug().
-							Str("entry", entry).
-							Bool("verifyInMap", tfm.Has(normalizePath(entry))).
-							Msg("orphanscan: added content_path entry to file map")
-					}
 				}

For walker.go:

 	if strings.Contains(path, "True.Lies.1994") {
-		log.Debug().
-			Str("rawPath", path).
-			Str("normPath", normPath).
-			Bool("inMap", w.tfm.Has(normPath)).
-			Msg("orphanscan: walker debug for True.Lies")
-	}
 	if w.tfm.Has(normPath) {

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@internal/services/orphanscan/service.go` around lines 1170 - 1175, Remove the
title-specific debug branches that check for strings.Contains(...,
"True.Lies.1994") in the orphan scan logic: delete the conditional block in
service.go that wraps the log.Debug() call around strings.Contains(entry,
"True.Lies.1994") (the block that logs Str("entry", entry).Bool("verifyInMap",
tfm.Has(normalizePath(entry))).Msg(...)), and likewise remove the similar
conditional and debug log in walker.go that checks strings.Contains(path,
"True.Lies.1994"); leave the surrounding logic intact (e.g., keep any file-map
additions or normalization such as normalizePath(entry) and tfm.Has calls) but
eliminate these hardcoded, data-specific logging branches.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@internal/services/orphanscan/service.go`:
- Around line 1170-1175: Remove the title-specific debug branches that check for
strings.Contains(..., "True.Lies.1994") in the orphan scan logic: delete the
conditional block in service.go that wraps the log.Debug() call around
strings.Contains(entry, "True.Lies.1994") (the block that logs Str("entry",
entry).Bool("verifyInMap", tfm.Has(normalizePath(entry))).Msg(...)), and
likewise remove the similar conditional and debug log in walker.go that checks
strings.Contains(path, "True.Lies.1994"); leave the surrounding logic intact
(e.g., keep any file-map additions or normalization such as normalizePath(entry)
and tfm.Has calls) but eliminate these hardcoded, data-specific logging
branches.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: fc95d4be-6b46-4947-bb61-a1a3fe970915

📥 Commits

Reviewing files that changed from the base of the PR and between 2e590e1 and a38474a.

📒 Files selected for processing (1)

• internal/services/orphanscan/service.go

[coderabbitai]

🧹 Nitpick comments (1)

internal/services/orphanscan/service.go (1)

159-168: Add defensive check for empty or non-absolute actualSavePath.

The TrimSuffix operation could produce an empty string in edge cases (e.g., if contentPath is at the filesystem root like /file.mkv). While extremely unlikely in practice, this would result in relative paths being added to the file map, which wouldn't match during the absolute-path walk.

🛡️ Suggested defensive fix

 		contentPath := filepath.Clean(torrent.ContentPath)
 		if contentPath != "" && filepath.IsAbs(contentPath) && len(files) > 0 {
 			firstFileName := filepath.Clean(files[0].Name)
 			actualSavePath := strings.TrimSuffix(contentPath, string(filepath.Separator)+firstFileName)
-			if actualSavePath != contentPath && actualSavePath != savePath {
+			if actualSavePath != "" && filepath.IsAbs(actualSavePath) && actualSavePath != contentPath && actualSavePath != savePath {
 				scanRoots[actualSavePath] = struct{}{}
 				for _, f := range files {
 					tfm.Add(normalizePath(filepath.Join(actualSavePath, f.Name)))
 				}
 			}
 		}

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@internal/services/orphanscan/service.go` around lines 159 - 168, The loop
currently relies on normalizePath but doesn't guard against TrimSuffix producing
an empty or non-absolute actualSavePath; add a defensive check so empty or
non-absolute normalized paths are skipped. Update normalizePath (or immediately
after calling it in the loop that iterates over a and b) to return/skip when the
result is "" or when filepath.IsAbs(normalized) is false (i.e., treat those as
invalid by continuing the loop), and ensure this check covers values produced by
the TrimSuffix path transformation (actualSavePath) so no relative/empty paths
are added to the file map.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@internal/services/orphanscan/service.go`:
- Around line 159-168: The loop currently relies on normalizePath but doesn't
guard against TrimSuffix producing an empty or non-absolute actualSavePath; add
a defensive check so empty or non-absolute normalized paths are skipped. Update
normalizePath (or immediately after calling it in the loop that iterates over a
and b) to return/skip when the result is "" or when filepath.IsAbs(normalized)
is false (i.e., treat those as invalid by continuing the loop), and ensure this
check covers values produced by the TrimSuffix path transformation
(actualSavePath) so no relative/empty paths are added to the file map.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 6f05b550-d693-4f05-982d-ed7e3c141e27

📥 Commits

Reviewing files that changed from the base of the PR and between a38474a and c460aa3.

📒 Files selected for processing (2)

• internal/services/orphanscan/service.go
• internal/services/orphanscan/settling_test.go

✅ Files skipped from review due to trivial changes (1)

• internal/services/orphanscan/settling_test.go

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@internal/services/orphanscan/service.go`:
- Around line 1152-1168: The code derives an actualSavePath from
torrent.ContentPath but only applies it to the current instance’s file map; to
ensure cross-instance overlap checks see the same root, include that
content-derived root in the set returned/used by scanRootsFromTorrents() (and
therefore by buildInstanceScanRoots()); update scanRootsFromTorrents() (or the
scanRoots collection it returns) to add actualSavePath when computed (same
conditions used now) so buildFileMap()/MergeFrom() will consider the
content-derived root for overlap detection across instances.
- Around line 1159-1164: The current logic trims the first file name from
torrent.ContentPath using strings.TrimSuffix which yields empty string for
root-level paths (e.g., "/file.mkv") and drops the volume separator for Windows
(e.g., "D:\file.mkv" -> "D:"), causing root/volume roots to be skipped; replace
the TrimSuffix approach with filepath.Dir(torrent.ContentPath) to compute
actualSavePath, and if that returns "" or a volume-only path, normalize it to
the proper filesystem root (use string(filepath.Separator) for Unix roots or
filepath.VolumeName(contentPath)+string(filepath.Separator) for Windows) before
checking IsAbs and comparing to contentPath/savePath, then add the normalized
actualSavePath into scanRoots as before (referencing contentPath, files[0].Name,
actualSavePath, scanRoots, savePath).

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: f2be6044-f05d-4371-8bcf-33c9eab79be9

📥 Commits

Reviewing files that changed from the base of the PR and between a38474a and a0a368d.

📒 Files selected for processing (2)

• internal/services/orphanscan/service.go
• internal/services/orphanscan/settling_test.go

✅ Files skipped from review due to trivial changes (1)

• internal/services/orphanscan/settling_test.go

[nitrobass24]

@s0up4200 I have a few open PRs, but this one is the most urgent fix. Orphan Scan deleted about 10TB of torrents before I figured out what was happening.

[s0up4200]

@s0up4200 I have a few open PRs, but this one is the most urgent fix. Orphan Scan deleted about 10TB of torrents before I figured out what was happening.

Man, thats one nasty bug. Hope you can recover :/

Your fix seem correct, but there was one more place still relying on save_path only. I updated the cross-instance overlap check to use content_path as well so the same false positive cant happen in shared disk multi-instance setups. Let me know what you think.

[nitrobass24]

Code changes look good and I just did a test in my live environment and it works as intended. Ready to go!