โฏ๏ธ Json::Merge
if ci_badges.map(&:color).detect { it != "green"} โ๏ธ let me know on Discord or RubyForum, as I may have missed the notification.
if ci_badges.map(&:color).all? { it == "green"} ๐๏ธ send money so I can do more of this. FLOSS maintenance is now my full-time job.
๐ฃ How will this project approach the September 2025 hostile takeover of RubyGems? ๐๏ธ
I've summarized my thoughts in this blog post.
๐ป Synopsis

Json::Merge intelligently merges two versions of a JSON, JSONC, or JSON5 file using the StructuredMerge Ruby stack. It is built on ast-merge and tree_haver, with parser capability reported through the shared backend registry.
JSONC and JSON5 use the same Json::Merge API as JSON; use require "json/merge" for all three dialects.
JSONC Support
The JSON merge provider supports JSON, JSONC, and JSON5 dialects through Json::Merge. Pass JSONC content to the same merger API used for JSON.
JSONC-specific behavior:
- **Single API**: JSON and JSONC both use `Json::Merge`
- **Comment-Aware**: Preserves `//` and `/* */` comments when the parser exposes them
- **Trailing Commas**: Accepts trailing commas in JSONC objects and arrays
- **JSONC Boundaries**: Rejects JSON5-only syntax such as unquoted keys,
single-quoted strings, and JSON5 numeric literals
- **Freeze Blocks**: Uses the `json-merge` freeze token by default, with a custom token available when needed
JSON5 Support
JSON5 uses the same structural merge engine with dialect: :json5. It accepts
JSON5 syntax, including comments, trailing commas, unquoted object keys,
single-quoted strings, hexadecimal and signed numeric literals, Infinity, and
NaN. JSONC remains intentionally narrower: it accepts comments and trailing
commas, but rejects the JSON5-only forms.
Key Features
- **Tree-Sitter Powered**: Uses tree-sitter-json for strict JSON and the
normalized JSON5 tree for JSONC and JSON5 syntax
- **JSONC-Aware**: Preserves `//` and `/* */` comments when the parser exposes them
- **Intelligent**: Matches objects and arrays by structural signatures
- **Fuzzy Property Matching**: `ObjectMatchRefiner` matches similar property names
(e.g., `databaseUrl` โ `database_url`) using Levenshtein distance for naming convention differences
- **Full Provenance**: Tracks origin of every node
- **StructuredMerge Native**: Depends on `ast-merge` and `tree_haver`; parser availability comes from registered backend providers
- **Customizable**:
- `signature_generator` - callable custom signature generators
- `preference` - setting of `:template`, `:destination`, or a Hash for per-node-type preferences
- `node_splitter` - Hash mapping node types to callables for per-node-type merge customization (see [ast-merge][ast-merge] docs)
- `add_template_only_nodes` - setting to retain nodes that do not exist in destination
- `match_refiners` - array of refiners for fuzzy matching (e.g., `ObjectMatchRefiner`)
Supported Node Types
| Node Type | Signature Format | Matching Behavior |
|---|---|---|
| Object | [:object, key_signatures...] |
Objects match by their key structure |
| Array | [:array, element_count] |
Arrays match by position and type |
| Pair | [:pair, key_name] |
Key-value pairs match by key name |
| String | [:string, value] |
Strings match by value |
| Number | [:number, value] |
Numbers match by value |
| Boolean | [:boolean, value] |
Booleans match by value |
| Null | [:null] |
Null values always match |
Example
require "json/merge"
template = File.read("template.json")
destination = File.read("destination.json")
merger = Json::Merge::SmartMerger.new(template, destination)
result = merger.merge
File.write("merged.json", result.to_json)
๐ก Info you can shake a stick at
| Tokens to Remember | |
|---|---|
| Works with MRI Ruby 4 | |
| Support & Community | |
| Source | |
| Documentation | |
| Compliance | |
| Style | |
| Maintainer ๐๏ธ | |
... ๐ |
Compatibility
Compatible with MRI Ruby 4.0.0+, and concordant releases of JRuby, and TruffleRuby.
CI workflows and Appraisals are generated for MRI Ruby 4.0.0+.
This test floor is configured by ruby.test_minimum in .kettle-jem.yml and
may be higher than the gem's runtime compatibility floor when legacy Rubies are
not practical for the current toolchain.
The amazing test matrix is powered by the kettle-dev stack.
How kettle-dev manages complexity in tests
| Gem | Source | Role | Total downloads |
|---|---|---|---|
| appraisal2 | GitHub | multi-dependency Appraisal matrix generation | |
| appraisal2-rubocop | GitHub | RuboCop Appraisal generator integration | |
| kettle-dev | GitHub | development, release, and CI workflow tooling | |
| kettle-jem | GitHub | Appraisals & CI workflow templates | |
| kettle-soup-cover | GitHub | SimpleCov coverage policy and reporting | |
| kettle-test | GitHub | standard test runner and coverage harness | |
| rubocop-lts | GitHub | Ruby-version-aware linting | |
| turbo_tests2 | GitHub | parallel test execution |
โจ Installation
Install the gem and add to the application's Gemfile by executing:
bundle add json-merge
If bundler is not being used to manage dependencies, install the gem by executing:
gem install json-merge
โ๏ธ Configuration
Signature Match Preference
Control which version to use when nodes have matching signatures but different content:
# Use template version (for config updates)
merger = Json::Merge::SmartMerger.new(
template,
destination,
preference: :template,
)
# Use destination version (default - preserve customizations)
merger = Json::Merge::SmartMerger.new(
template,
destination,
preference: :destination,
)
Template-Only Nodes
Control whether to add nodes that only exist in the template:
# Add template-only nodes
merger = Json::Merge::SmartMerger.new(
template,
destination,
add_template_only_nodes: true,
)
Object Match Refiner
When JSON object properties (key-value pairs) don't match by exact key name, the
ObjectMatchRefiner uses fuzzy matching to pair entries with:
- Similar key names (e.g., `databaseUrl` vs `database_url`)
- Keys with typos or different naming conventions (camelCase vs snake\_case)
- Array elements with similar structure or content
# Enable object fuzzy matching
merger = Json::Merge::SmartMerger.new(
template,
destination,
match_refiners: [
Json::Merge::ObjectMatchRefiner.new(threshold: 0.5),
],
)
ObjectMatchRefiner Options
| Option | Default | Description |
|---|---|---|
threshold |
0.5 | Minimum similarity score (0.0-1.0) to accept a match |
key_weight |
0.7 | Weight for key name similarity |
value_weight |
0.3 | Weight for value similarity |
# Custom weights for key-centric matching
refiner = Json::Merge::ObjectMatchRefiner.new(
threshold: 0.6,
key_weight: 0.8, # Focus more on key names
value_weight: 0.2, # Less focus on values
)
Debug Logging
Enable debug logging to see merge decisions:
export JSON_MERGE_DEBUG=1
JSONC and JSON5 Options
JSONC and JSON5 files use the same options as JSON files. Set dialect when
constructing a direct merger:
merger = Json::Merge::SmartMerger.new(
template_content,
dest_content,
# Which version to prefer when nodes match
# :destination (default) - keep destination values
# :template - use template values
preference: :destination,
# Whether to add template-only nodes to the result
# false (default) - only include properties that exist in destination
# true - include all template properties
add_template_only_nodes: false,
# :json (strict), :jsonc (comments and trailing commas), or :json5
dialect: :jsonc,
# Token for freeze block markers
# Default: "json-merge"
# Looks for: // json-merge:freeze / // json-merge:unfreeze
freeze_token: "json-merge",
# Custom signature generator (optional)
# Receives a node, returns a signature array or nil
signature_generator: ->(node) { [:pair, node.key] if node.type == :pair },
)
๐ง Basic Usage
Merging Two JSON Files
require "json/merge"
template_content = File.read("template.json")
dest_content = File.read("destination.json")
merger = Json::Merge::SmartMerger.new(template_content, dest_content)
result = merger.merge
File.write("merged.json", result.to_json)
Analyzing a JSON File
require "json/merge"
source = File.read("config.json")
analysis = Json::Merge::FileAnalysis.new(source)
# Iterate over all top-level nodes
analysis.statements.each do |node|
sig = analysis.generate_signature(node)
puts "#{node.class}: #{sig.inspect}"
end
Fuzzy Property Matching
When property names differ between template and destination (e.g., naming convention changes),
use the ObjectMatchRefiner:
require "json/merge"
template = <<~JSON
{
"databaseUrl": "postgres://localhost/app",
"cacheTimeout": 3600,
"apiEndpoint": "https://api.example.com"
}
JSON
destination = <<~JSON
{
"database_url": "postgres://localhost/custom",
"cache_ttl": 7200,
"api_endpoint": "https://custom.example.com"
}
JSON
# Default merge won't match keys (names differ - camelCase vs snake_case)
# Use ObjectMatchRefiner for fuzzy matching
merger = Json::Merge::SmartMerger.new(
template,
destination,
match_refiners: [
Json::Merge::ObjectMatchRefiner.new(threshold: 0.5),
],
)
result = merger.merge
# Properties are matched despite naming convention differences:
# - databaseUrl โ database_url (similar when normalized)
# - cacheTimeout โ cache_ttl (similar: "cache")
# - apiEndpoint โ api_endpoint (similar when normalized)
Array Element Matching
The ObjectMatchRefiner also handles array elements with similar structure:
template = <<~JSON
{
"users": [
{ "id": 1, "userName": "alice" },
{ "id": 2, "userName": "bob" }
]
}
JSON
destination = <<~JSON
{
"users": [
{ "id": 1, "user_name": "alice_custom" },
{ "id": 3, "user_name": "charlie" }
]
}
JSON
merger = Json::Merge::SmartMerger.new(
template,
destination,
match_refiners: [
Json::Merge::ObjectMatchRefiner.new(threshold: 0.5),
],
)
# Array elements with matching IDs or similar structure are paired
Merging JSONC Files
require "json/merge"
template = File.read("template.jsonc")
destination = File.read("destination.jsonc")
merger = Json::Merge::SmartMerger.new(template, destination)
result = merger.merge
File.write("merged.jsonc", result)
Merging JSON5 Files
require "json/merge"
template = File.read("template.json5")
destination = File.read("destination.json5")
merger = Json::Merge::SmartMerger.new(template, destination, dialect: :json5)
result = merger.merge
File.write("merged.json5", result)
JSONC Freeze Blocks
Freeze blocks protect sections from being overwritten during merge:
{
"name": "my-app",
// json-merge:freeze Secret configuration
"api_key": "my_production_api_key",
"api_secret": "super_secret_value",
// json-merge:unfreeze
"debug": false
}
Content between // json-merge:freeze and // json-merge:unfreeze markers is preserved from the destination file, regardless of what the template contains.
Adding Template-Only JSONC Properties
merger = Json::Merge::SmartMerger.new(
template,
destination,
add_template_only_nodes: true,
)
result = merger.merge
# Result includes properties from template that do not exist in destination
๐ Security
See SECURITY.md.
๐ค Contributing
If you need some ideas of where to help, you could work on adding more code coverage, or if it is already ๐ฏ (see below) check issues or PRs, or use the gem and think about how it could be better.
We so if you make changes, remember to update it.
See CONTRIBUTING.md for more detailed instructions.
Code Coverage
Coverage service badges
๐ Versioning
This library follows for its public API where practical.
For most applications, prefer the Pessimistic Version Constraint with two digits of precision.
For example:
spec.add_dependency("json-merge", "~> 7.0")
๐ Is "Platform Support" part of the public API? More details inside.
Dropping support for a platform can be a breaking change for affected users. If a release changes supported platforms, it should be called out clearly in the changelog and versioned with that impact in mind.
To get a better understanding of how SemVer is intended to work over a project's lifetime, read this article from the creator of SemVer:
See CHANGELOG.md for a list of releases.
๐ License
The gem is available under the following licenses: AGPL-3.0-only, PolyForm-Small-Business-1.0.0. See LICENSE.md for details.
If none of the available licenses suit your use case, please contact us to discuss a custom commercial license.
