YAML Parsing Error When Importing OCI Image via URL Import

[souvikinator]

Current Behavior

When attempting to upload an exported OCI image from a URL in Meshery, the import process fails with a YAML parsing error. This prevents the successful import of the model.

file used: https://github.com/souvikinator/my_configs/blob/master/apisix.tar

The same file imports successfully when using the file import method instead of the URL import option.

Expected Behavior

The OCI image should import successfully without encountering YAML parsing issues. Meshery should validate and process the uploaded image seamlessly.

Screenshots/Logs

Environment

• Host OS: Mac Linux Windows
• Platform: Docker or Kubernetes
• Meshery Server Version: stable-v0.8.19
• Meshery Client Version: stable-v0.8.19

---

Contributor Guides and Handbook

• 🛠 Meshery Build & Release Strategy
• 📚 Instructions for contributing to documentation
  • Meshery documentation site and source
• 🎨 Wireframes and designs for Meshery UI in Figma (open invite)
• 🙋🏾🙋🏼 Questions: Discussion Forum and Community Slack

[souvikinator]

@leecalcote The issue occurred because the URL pointed to the GitHub page (HTML) instead of the raw file, leading to errors in parsing. Using the raw GitHub URL, such as https://github.com/souvikinator/my_configs/raw/master/apisix.tar, resolves the problem as it directly links to the file.

While the import now works, there's a need for:

1. Robust file type detection to prevent similar issues from propagating. The current error messaging can be misleading due to the content of type HTML being propagated down the pipeline.
2. Providing more context from a UI perspective would enhance clarity (what file types are supported over URL import, specifically mentioning about GitHub URLs, etc)

Switching to the raw file URL addresses the issue effectively, but improving detection and feedback mechanisms would make the process more user-friendly and reliable.

@amitamrutiya FYI

[souvikinator]

I'll create a pull request addressing the first point linking to this issue. Currently, with the URL import, if we cannot detect the file type, we set it to "unknown" and allow the content to proceed through the pipeline. Instead, we could block it at that point and provide a relevant message.

According to the code, the URL import should cover the following file types:

• OCI artifacts / Tar archives (.tar)
• Zip archives (.zip)
• Gzip archives (.tar.gz)
• YAML files (.yaml, .yml)

Am I missing any other file types?

[leecalcote]

@souvikinator .json

[leecalcote]

@souvikinator fixed?