Reduce filesystem reads in validate structure

[djjuhasz]

Refs #169.

• Refactor the validate strcuture activity to reduce the number of filesystem reads to a single walk of the SIP directory hierarchy
• Add a top level SIP directory to the unit tests to reflect the expected SFA SIP structure
• Add a check for a completely empty SIP and report with as a single failure message

[codecov]

Codecov Report

❌ Patch coverage is 97.05882% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 61.56%. Comparing base (c7adfe7) to head (155fd59).

Files with missing lines	Patch %	Lines
internal/activities/validate_structure.go	97.05%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #183      +/-   ##
==========================================
+ Coverage   61.23%   61.56%   +0.33%     
==========================================
  Files          43       43              
  Lines        2515     2511       -4     
==========================================
+ Hits         1540     1546       +6     
+ Misses        830      825       -5     
+ Partials      145      140       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jraddaoui]

LGTM, thanks @djjuhasz! Probably not the best time, but a nice reorg.

But this doesn't fully address #169, right? Thinking about the changelog, does it introduce any functional changes? "Added: Check for a completely empty SIP (#169)", would that be okay?

[djjuhasz]

Yes, that's right it doesn't fully address #169. Your suggested log message looks good. 👍🏻

…

On Wed, Mar 25, 2026, 17:39 José Raddaoui Marín ***@***.***> wrote: ***@***.**** approved this pull request. LGTM, thanks @djjuhasz <https://github.com/djjuhasz>! Probably not the best time, but a nice reorg. But this doesn't fully address #169 <#169>, right? Thinking about the changelog, does it introduce any functional changes? "Added: Check for a completely empty SIP (#169 <#169>)", would that be okay? — Reply to this email directly, view it on GitHub <#183?email_source=notifications&email_token=AAEY3V2JLQDUYLYLYCJX2AD4SR34DA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTIMBRGA3TINBSHA42M4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJPLQOJPXEZLWNFSXOX3ON52GSZTJMNQXI2LPNZZV6Y3MNFRWW#pullrequestreview-4010744289>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAEY3V3NXUJDDEGNIYR6LXL4SR34DAVCNFSM6AAAAACW7MFT4OVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHM2DAMJQG42DIMRYHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jraddaoui]

I was planning to merge this PR for the upcoming release, but I gave it a try with https://github.com/artefactual-sdps/enduro/blob/main/internal/testdata/zipped_transfer/small.zip and it failed with the SIP is empty error, among other issues as that SIP doesn't follow the SFA structure. It worked okay with other SFA test transfer, so maybe this is more related to that than to this changes. However, that's not an empty SIP and that makes me wonder if this is working as expected. Since it doesn't fully address #169, I think it's better to merge this when David gets back.