Maven Profile Management

Classify Maven build profiles that weren't auto-matched during discovery.

When to Use

This skill is invoked by analyze-project-architecture when:

1. Project contains Maven modules (build_systems includes maven)
2. Discovery found NO-MATCH-FOUND profiles

Do not use directly - invoked conditionally from architecture analysis workflow.

Profile Classification

Maven profiles enable optional build features. Extension API classifies profiles during discovery:

Key Insight: Most NO-MATCH-FOUND profiles are correctly unmatched:

• apache-release → Release process, not a build command
• skip-unit-tests → Test skipping, not a positive command
• use-apache-snapshots → Repository config, not a command

Workflow

Step 1: Collect Unmatched Profiles Across All Modules

Step 1a: Get Module List

python3 .plan/execute-script.py plan-marshall:analyze-project-architecture:architecture modules

Output (TOON):

modules[N]:
  - module-a
  - module-b
  - module-c

Step 1b: Query Each Module for Profiles

For each module in the list:

python3 .plan/execute-script.py plan-marshall:analyze-project-architecture:architecture \
  derived-module --name {module-name}

Parse the TOON output:

• Check build_systems contains maven
• If Maven, check metadata.profiles for entries with canonical: NO-MATCH-FOUND

Step 1c: Build Unmatched Profile Set

Collect all NO-MATCH-FOUND profiles into a deduplicated set:

unmatched_profiles = {profile-id-1, profile-id-2, ...}

Note: Same profile ID may appear in multiple modules. Only ask once per unique profile ID.

If set is empty → Exit, nothing to do.

Step 2: Ask User About Each Unmatched Profile

For each NO-MATCH-FOUND profile:

AskUserQuestion:
  question: "Maven profile '{profile-id}' is unmatched. What should it do?"
  header: "Profile"
  options:
    - label: "Ignore"
      description: "Leave as NO-MATCH-FOUND, no command generated"
    - label: "Skip"
      description: "Add to skip list, exclude from all processing"
    - label: "Map to canonical"
      description: "Map to integration-tests, coverage, benchmark, or quality-gate"
  multiSelect: false

Step 3: Apply User Decision

Skip - Add to skip list:

# Get current value first
python3 .plan/execute-script.py plan-marshall:manage-plan-marshall-config:plan-marshall-config ext-defaults get \
  --key build.maven.profiles.skip

# Append new profile (comma-separated)
python3 .plan/execute-script.py plan-marshall:manage-plan-marshall-config:plan-marshall-config ext-defaults set \
  --key build.maven.profiles.skip --value "{existing},{profile-id}"

Map - Add canonical mapping:

# Get current mappings first
python3 .plan/execute-script.py plan-marshall:manage-plan-marshall-config:plan-marshall-config ext-defaults get \
  --key build.maven.profiles.map.canonical

# Append new mapping (comma-separated profile:canonical pairs)
python3 .plan/execute-script.py plan-marshall:manage-plan-marshall-config:plan-marshall-config ext-defaults set \
  --key build.maven.profiles.map.canonical --value "{existing},{profile-id}:{canonical}"

Step 4: Re-run Discovery

After any configuration change:

python3 .plan/execute-script.py plan-marshall:analyze-project-architecture:architecture discover --force

Canonical Classifications

Multiple Profiles to One Canonical

When multiple profiles map to the same canonical:

• Only ONE command is generated
• First discovered profile becomes primary
• All profiles listed in all_profiles

User override: Add unwanted profiles to skip list.

Storage

Configuration stored in marshal.json under extension_defaults:

{
  "extension_defaults": {
    "build.maven.profiles.skip": "itest,native",
    "build.maven.profiles.map.canonical": "local-integration-tests:integration-tests,perf:benchmark"
  }
}

Key Formats:

• build.maven.profiles.skip - Comma-separated profile IDs to exclude
• build.maven.profiles.map.canonical - Comma-separated profile:canonical pairs

Skip list and profile mappings are read during architecture discovery to classify profiles.

Related Documents