Find a way to use hibernate to query and use data from a file store

[ia3andy]

It would store the data as a file synchronised with git. This would allow querying the data and use it in the app that gets generated as a static website.

The prerequisite is to have something easy to run in a CI (like github action) and that gives readable diffs for git.

[ia3andy]

Zulip discussion (https://quarkusio.zulipchat.com/#narrow/stream/187038-dev/topic/Json.20File.20JDBC.20driver) abstract:

Summary for GitHub Issue

Title: Implementing a JDBC Driver for JSON File Storage

Description:

Andy Damevin is exploring the possibility of using a JDBC driver to access and manipulate data stored in human-readable text files (preferably JSON) synchronized through Git. The goal is to allow easy data editing and storage in a format that supports version control and diffing, without requiring an additional runtime process.

Key Points:

• Use Case: Similar to roq-data but using Hibernate. Accessing and modifying data from Quarkus with synchronization via Git.
• Current Solutions:
  • H2 Database: Provides a file-based storage but results in binary-like data files.
  • CSV JDBC Drivers: Exist but are read-only and not suitable for the use case.
  • SQL Dumps: Suggested as an alternative, but it involves using traditional databases and exporting data.
• Explored Options:
  • Implementing a custom JDBC driver for JSON/YAML files.
  • Considering existing solutions like Apache Drill, which supports JDBC drivers for various data stores.
  • Existing systems like Gatsby, which queries data and stores it as JSON for static use.

Discussion Highlights:

• Yoann Rodière advised caution, noting that using JDBC for file-based storage is unconventional and might lead to challenges.
• Max Andersen highlighted multiple historical and modern solutions for similar needs, emphasizing the variety of ways to manage data storage and synchronization.

Next Steps:

• Further exploration of a custom JDBC driver for JSON/YAML.
• Investigating Apache Drill and other potential tools for file-based JDBC connectivity.
• Opening an issue to gather more insights and track progress on this exploration.

Links:

• JSON JDBC Driver Example
• CSV JDBC Driver (readonly)

---

Feel free to provide additional input or alternative solutions for efficiently managing and syncing human-readable file-based data storage through JDBC.

[maxandersen]

"Find a way to use hibernate to query and use data from a file store" - what is the queries you are after doing?

[ia3andy]

This is a good example: https://github.com/FroMage/RivieraDEV-Quarkus/tree/main/src/main/java/model

[maxandersen]

@ia3andy can you point a bit more specific - I browsed around and didn't find any queries - just a model easy to iterate on.

[FroMage]

What I would do is:

• Import YAML or JSON or whatever into ORM
• Queries / mutation / whatever
• Export back to YAML / JSON or whatever from ORM

Possibly using H2.

[FroMage]

Did I mention I have a JSON import/export for ORM in 🦊? 😇