Skip to content

Merging Configurations ​

One of Konfigo's most powerful features is its ability to intelligently merge multiple configuration sources. This guide explains how merging works and how to use it effectively.

Quick Start ​

Merge two configuration files:

bash
konfigo -s base.yaml,prod.yaml -of result.json

How Merging Works ​

Konfigo uses deep merging with clear precedence rules:

  1. Environment Variables (KONFIGO_KEY_*) - Highest precedence, merged last
  2. Sources in CLI order - Processed left-to-right as listed in -s; later sources override earlier ones
  3. Stdin (-) - Merged in whatever position it appears in the -s list (not inherently lowest)

Example: Three-Way Merge ​

bash
konfigo -s base.yaml,env.yaml,local.yaml

Processing order: base.yaml → env.yaml → local.yaml → Environment variables

Merging Rules by Data Type ​

Objects: Deep Merge ​

Objects are merged recursively, combining all keys:

yaml
# base.yaml
app:
  name: "myapp"
  port: 8080
  features:
    auth: true

# prod.yaml  
app:
  port: 9090
  features:
    monitoring: true

Result:

yaml
app:
  name: "myapp"        # from base
  port: 9090           # overridden by prod
  features:
    auth: true         # from base
    monitoring: true   # added by prod

Scalars: Complete Replacement ​

Strings, numbers, and booleans are completely replaced:

yaml
# base.yaml
database:
  host: "localhost"
  port: 5432
  ssl: false

# prod.yaml
database:
  host: "prod-db.com"
  ssl: true

# Result
database:
  host: "prod-db.com"  # replaced
  port: 5432           # preserved
  ssl: true            # replaced

Arrays: Complete Replacement (Default) ​

By default, arrays are replaced entirely, not merged:

yaml
# base.yaml
tags: ["app", "service"]

# override.yaml
tags: ["app", "production", "critical"]

# Result
tags: ["app", "production", "critical"]  # completely replaced

Arrays: Union with Deduplication (-m flag) ​

Use the -m flag to merge arrays by union instead of replacing:

bash
konfigo -m -s base.yaml,override.yaml
yaml
# base.yaml
tags: ["app", "service"]

# override.yaml
tags: ["app", "production", "critical"]

# Result
tags: ["app", "service", "production", "critical"]  # union with dedup

Environment Variable Overrides ​

Override any configuration value using environment variables:

bash
# Override nested keys using dot notation
export KONFIGO_KEY_database.host="override-host"
export KONFIGO_KEY_app.port=3000
export KONFIGO_KEY_features.auth=false

konfigo -s config.yaml

Key format: KONFIGO_KEY_<path.to.key>=<value>

Environment variables always have the highest precedence.

Protected Paths (Immutable) ​

Use schemas to protect critical configuration paths:

yaml
# schema.yaml
immutable:
  - "app.name"
  - "security.keys"
  - "database.credentials"

# These paths cannot be overridden by later sources
# (but environment variables can still override them)

Practical Examples ​

Multi-Environment Setup ​

bash
# Base configuration + environment-specific overrides
konfigo -s base.yaml,environments/prod.yaml -of prod-config.json

# With environment variables for runtime customization
KONFIGO_KEY_app.port=8080 konfigo -s base.yaml,prod.yaml

Configuration Layers ​

bash
# Layer configurations: defaults → environment → local customization
konfigo -s defaults.json,env/production.yaml,local.toml

Format Conversion During Merge ​

bash
# Merge YAML and JSON, output as TOML
konfigo -s config.yaml,overrides.json -ot -of final.toml

Best Practices ​

1. Organize by Precedence ​

Place more general configurations first, specific ones last:

bash
konfigo -s defaults.yaml,environment.yaml,local.yaml

2. Use Environment Variables for Runtime ​

Perfect for containerized deployments:

bash
# In Docker/Kubernetes
KONFIGO_KEY_database.host=$DB_HOST konfigo -s config.yaml

3. Protect Critical Settings ​

Use immutable paths for security-sensitive configuration:

yaml
immutable:
  - "security.apiKeys"
  - "database.credentials"

4. Test Your Merges ​

Always verify the merged result:

bash
# Output to screen to verify before saving
konfigo -s base.yaml,prod.yaml

Best Practices ​

✅ Do's ​

Structure source files by purpose

bash
# Good: Clear hierarchy and responsibility
konfigo -s defaults.yaml,env/prod.yaml,local-overrides.yaml

Use consistent naming patterns

yaml
# Good: Predictable structure
config/
  ├── base.yaml          # Common settings
  ├── env/
  │   ├── dev.yaml       # Development overrides
  │   ├── staging.yaml   # Staging overrides
  │   └── prod.yaml      # Production overrides
  └── local.yaml         # Developer-specific (git-ignored)

Keep environment variables for runtime-only values

bash
# Good: Runtime secrets and deployment-specific values
KONFIGO_KEY_database.password=$DB_PASSWORD \
KONFIGO_KEY_app.instanceId=$INSTANCE_ID \
konfigo -s base.yaml,prod.yaml

Test your merge logic

bash
# Good: Verify before deployment
konfigo -s base.yaml,prod.yaml --dry-run -v

❌ Don'ts ​

Don't rely on file order for critical logic

bash
# Bad: Unclear which file wins
konfigo -s prod.yaml,base.yaml,override.yaml

Don't mix configuration concerns

yaml
# Bad: Infrastructure mixed with application config
database:
  host: "prod-db.com"
kubernetes:
  replicas: 3
app:
  port: 8080

Don't put secrets in source files

yaml
# Bad: Secrets in version control
database:
  password: "supersecret123"  # Use environment variables instead!

Don't create deep nesting without purpose

yaml
# Bad: Unnecessarily complex structure
app:
  config:
    settings:
      database:
        connection:
          primary:
            host: "localhost"  # Too deep!

# Good: Reasonable nesting
database:
  host: "localhost"

⚠️ Common Pitfalls ​

Array Replacement

By default, arrays are completely replaced, not merged. Use the -m flag if you need union-based array merging:

bash
konfigo -m -s base.yaml,override.yaml

Alternatively, consider using objects with keys:

yaml
# Instead of this:
services: ["web", "api", "worker"]

# Consider this:
services:
  web: { enabled: true }
  api: { enabled: true }
  worker: { enabled: false }

Case Sensitivity

Configuration keys are case-sensitive by default. Be consistent:

yaml
# These are different keys:
Database: { host: "..." }
database: { host: "..." }

Environment Variable Precedence

Environment variables always win. Don't rely on file-based overrides for values that might be set in the environment:

bash
# This file setting will be ignored:
# config.yaml: { port: 8080 }
KONFIGO_KEY_port=9090 konfigo -s config.yaml  # port will be 9090

Common Patterns ​

Environment-Specific Deployment ​

bash
# Development
konfigo -s base.yaml,env/dev.yaml -of dev-config.json

# Production  
konfigo -s base.yaml,env/prod.yaml -of prod-config.json

Legacy System Integration ​

bash
# Merge legacy .env with modern YAML
konfigo -s legacy.env,modern.yaml -of integrated.json

Configuration Validation ​

bash
# Merge and validate against schema
konfigo -s base.yaml,overrides.yaml -S validation.schema.yaml

Troubleshooting ​

Unexpected Overrides ​

Use verbose mode to see merge order:

bash
konfigo -v -s file1.yaml,file2.yaml

Array Merging Issues ​

By default, arrays are replaced, not merged. Use the -m flag to enable union-based array merging with deduplication.

Case Sensitivity ​

By default, merging is case-sensitive. Use -c flag for case-insensitive merging:

bash
konfigo -c -s config1.yaml,config2.yaml

Next Steps ​