Add QuiltData external service for zarr file browsing

[Copilot]

• Review and understand the existing externalServices.ts file structure
• Add logic to parse zarr_id from contentUrl (assetS3Url) for zarr files
• Add QuiltData as a new external service in EXTERNAL_SERVICES array
• Add zarr_id to ServiceUrlData interface
• Update serviceURL function to handle zarr_id placeholder replacement
• Add comment referencing dandi-schema issue Standardize 4xx error responses #356
• Test the changes with type-check and lint
• Build and verify the changes work correctly
• Improve UUID regex pattern to be more precise
• Enhance S3 bucket extraction to handle multiple URL formats
• Refactor code for better maintainability (extract constants and helper functions)
• Optimize regex compilation for performance
• Final code review completed
• Security check passed (CodeQL: 0 alerts)
• Update QuiltData name to "QuiltData: S3 Browser"
• Revert unrelated package-lock.json changes

Original prompt

---

This section details on the original issue you should resolve

<issue_title>External services for "Open with" for zarrs to browse them: add quiltdata</issue_title>
<issue_description>relates to good old

• Add URL to browse the zarr to the error message #1746

as ATM we do not provide file navigator within zarrs, and to an alternative solution suggested

• Add webdav.dandiarchive.org to "services" for the main instance and redirect /download/ for zarrs to webdav  #1993

but since "bundled" with handling for /download, the web UI aspect was forgotten/never implemented!

I think it is time to implement support for that and also quiltdata which provides a convenient navigator for S3 (demoed by @dyf at sfn 2025). So for a zarr, since it is a key prefix on S3, would work beautifully:

https://open.quiltdata.com/b/dandiarchive/tree/zarr/7b617177-ad57-4f7f-806b-060e18f42d15/ for the one from https://dandiarchive.org/dandiset/001675/draft/files?location=pipeline-aind%2Bephys/blobs/048d1ee9-83b7-491f-8f02-1ca615b1d455/run-1&page=1 .

Note that since it is embargoed, it would/could (needs DANDI API key entered, related fresh filed: flatironinstitute/neurosift#368) work only with neurosift and not webdav (public only ATM).

To add support for that type of URL we would need to parse out zarr_id from the contentUrl entry which points to amazon directly. Unfortunately we seems to not expose content ID here

• Add contentId to asset metadata dandi-schema#356</issue_description>

<agent_instructions>the post has all details - review the code in that web/src/utils/externalServices.ts, add zarr_id attribute which would get parsed from a contentUrl which would have /zarr/ in top of the bucket URL (not necessarily dandiarchive). Add a comment there to point to dandi/dandi-schema#356 as potentially in the future providing a better solution. And then compose target quiltdata url using that zarr_id</agent_instructions>

Comments on the Issue (you are @copilot in this section)

• Fixes External services for "Open with" for zarrs to browse them: add quiltdata #2655

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[yarikoptic]

Worked wonderfully on https://deploy-preview-2656--sandbox-dandiarchive-org.netlify.app/dandiset/216584/draft/files?location=one-more-icephys.zarr/another-icephys.zarr/different-icephys.zarr/new-icephys.zarr/new-icephys-zarr/new-icephys/icephys&page=1

(note: I asked to rename to rename to a bit longer name there since initial demo)

and lead to https://open.quiltdata.com/b/dandi-api-staging-dandisets/tree/zarr/48e9f28b-4a45-4ab4-b342-045dbad28072/

and lead to https://open.quiltdata.com/b/dandi-api-staging-dandisets/tree/zarr/3898620d-dfa3-4980-9cba-4c120ab8e520/

[yarikoptic]

Although I think we could also add bucket into the metadata record for reuse, WDYT folks?

fresh updated preview (tested that it works):