Jmix FetchPlan Analyzer

Early-stage static analyzer for Jmix/Spring services that extracts used entity paths from Java code and compares them against a declared fetch plan.

Status

v1 pilot

• suitable for internal evaluation
• intended for collecting feedback and unsupported cases
• not a production gatekeeper yet
• best-effort analysis, not a proof of correctness for arbitrary Java code

Problem

In Jmix/Spring applications, fetch plans tend to drift over time:

• some fields are used in code but missing from the fetch plan
• some fields remain in the fetch plan but are no longer used
• manual verification becomes slow and error-prone

This tool analyzes service methods, extracts used entity paths, normalizes them into canonical paths, and compares them with a declared fetch plan.

What the tool does

Given:

• a target service class
• a target method
• a root parameter
• a fetch plan

the tool produces:

• raw analysis tree
• normalized canonical paths
• comparison report:
  • Covered
  • Missing
  • Extra
  • Uncertain

Current capabilities

The current pilot supports:

• linear getter chains
• if / else
• foreach
• list.get(0) as collection element access
• stream().map(MethodRef) plus basic pass-through operators
• casts
• alias rebinding
• UNKNOWN_BREAK
• interprocedural analysis v1
• recursion guard
• Spring bean resolution for simple interface-based injections

See docs/SUPPORTED_SCOPE.md for details and limitations.

Example output

Canonical paths

• type.code
• type.name
• subcontracts.number
• subcontracts.status
• subcontracts.date
• employee.department.name
• employee.email

Comparison report

• Covered — analyzed paths already covered by fetch plan
• Missing — analyzed paths missing in fetch plan
• Extra — declared leaf fetch plan paths that are not used
• Uncertain — paths affected by unsupported or partially supported patterns

High-level pipeline

1. Parse Java source and locate target method
2. Build RawTree
3. Normalize RawTree into canonical PathTree
4. Flatten canonical paths
5. Extract declared paths from Jmix fetch plan
6. Compare analyzed vs declared paths
7. Produce report

Repository structure

• docs/ARCHITECTURE.md — architecture overview
• docs/SUPPORTED_SCOPE.md — supported/unsupported patterns
• docs/ROADMAP.md — v1 / v1.1 / v2
• docs/DECISIONS.md — important design decisions
• docs/EXAMPLES.md — sample inputs and outputs
• docs/FEEDBACK_TEMPLATE.md — how to report unsupported cases

Intended usage

Current intended usage is manual/internal analysis:

• run analysis for a real service method
• inspect normalized paths
• compare with fetch plan
• review Missing, Extra, and Uncertain
• collect unsupported patterns for future iterations

Non-goals for current version

This project does not try to:

• fully understand arbitrary Java code
• act as a mathematically complete verifier
• replace human review in all cases
• support reflection-heavy or runtime-generated logic

Contributing / feedback

If you want to help, please use the feedback format in docs/FEEDBACK_TEMPLATE.md.

License

• Apache-2.0