cmd: add --dedup-stats flag to ls; show per-file dedup stats

[Janmesh23]

This PR adds a --dedup-stats (-d) flag to sietch ls. When enabled, ls will:

• Show number of shared chunks per file (shared_chunks).
• Show estimated bytes avoided per file due to dedup (saved) using per-chunk EncryptedSize (fallbacks to Size, then 4MiB default).
• List other files that share those chunks (shared_with) — truncated to first 10 entries.

Notes:

• The saved bytes reflect how much duplicate data the file avoided storing (per-file view).
• For global reclaimed space metrics we could show (refs-1)*chunkSize aggregated per chunk; can be implemented in a follow-up PR.
• Includes unit test for computeDedupStatsForFile.

• The above image shows the implementation of the cmd sietch ls --dedup-stats

• The above image shows the implementation of the cmd sietch ls -l -d

Fixes: #96

[codecov]

Codecov Report

❌ Patch coverage is 20.17544% with 91 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
internal/deduplication/util.go	0.00%	41 Missing ⚠️
internal/ls/ui.go	0.00%	31 Missing ⚠️
cmd/ls.go	54.76%	15 Missing and 4 partials ⚠️

📢 Thoughts on this report? Let us know!

[S4tvara]

Hi @Janmesh23,
Thanks for your contribution. I’ve suggested some changes—please review them when you get a chance. Also, the target test code coverage is 11.64%, but your current PR is at 2.94%.

I was also able to reproduce the results:

[Janmesh23]

Hi @SubstantialCattle5 , thank you for reviewing , i would appreciate if you can tell me more about from where can i see the suggestions/changes (i'm a first time contributor hence need some help navigating )
again , thanks for your cooperation

[S4tvara]

there should be 3-4 comments tagged with review.

[S4tvara]

The Codecov report shows that the patch coverage for this PR is only 2.94%, with 99 lines in cmd/ls.go missing test coverage.
This means that most of the new or modified code in this PR isn’t covered by automated tests.

[Janmesh23]

The Codecov report shows that the patch coverage for this PR is only 2.94%, with 99 lines in cmd/ls.go missing test coverage. This means that most of the new or modified code in this PR isn’t covered by automated tests.

sure , i'll get on it

[Janmesh23]

there should be 3-4 comments tagged with review.

hi @SubstantialCattle5 , I could not see the comments you added in file changes
please let me know where I can see them
sorry for the inconvenience

[S4tvara]

there should be 3-4 comments tagged with review.

hi @SubstantialCattle5 , I could not see the comments you added in file changes please let me know where I can see them sorry for the inconvenience

@Janmesh23 ,
hi sure no worries I'll put them here itself.

[Janmesh23]

Refactored dedup logic for better modularity and testability.
Exported ComputeDedupStatsForFile, FormatSharedWith, and DisplayShortFormat so cmd and tests can reuse them.
Moved UI helpers under internal/ls and kept dedup logic in internal/deduplication.

[S4tvara]

@Janmesh23,

you’ve committed your test vaults and bak files as well. Please remove them.

Cheers,
Nilay

[Janmesh23]

@Janmesh23,

you’ve committed your test vaults and bak files as well. Please remove them.

Cheers, Nilay

yeah my bad , doing it right away :)

edit - done 👍

[S4tvara]

@Janmesh23 please follow the commit message guidelines explained in the `CONTRIBUTING.md`. Also, squash all your commits.

[S4tvara]

@Janmesh23,
I've squashed your commits from my side and also updated the commit message to follow the guidelines.
Please make sure to follow the standards defined for subsequent PRs.

Cheers,
Nilay

[Janmesh23]

@Janmesh23, I've squashed your commits from my side and also updated the commit message to follow the guidelines. Please make sure to follow the standards defined for subsequent PRs.

Cheers, Nilay

yess sure !
thankyou for the opportunity @SubstantialCattle5 , i would like to work on more such issues in sietch and learn more .
looking forward to do so :)

[S4tvara]

Sure just drop me a message whichever one you're interested in.