KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#developer-tools #typescript #dsh #dsh-plugin #ai-agent #codex #ai-agents #deepseek-harness #claude-code #cli #react #mcp

Validator

★ 33 Open GitHub ↗

A simple ruby object validator

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Validator

[![Build Status](https://secure.travis-ci.org/zombor/Validator.png)](http://travis-ci.org/zombor/Validator)

Validator is a simple ruby validation class. You don't use it directly inside your classes like just about every other ruby validation class out there. I chose to implement it in this way so I didn't automatically pollute the namespace of the objects I wanted to validate.

This also solves the problem of validating forms very nicely. Frequently you will have a form that represents many different data objects in your system, and you can pre-validate everything before doing any saving.

## Usage

Validator is useful for validating the state of any existing ruby object.

```ruby
  object = OpenStruct.new(:email => 'foo@bar.com', :password => 'foobar')
  validator = Validation::Validator.new(object)
  validator.rule(:email, [:email, :not_empty])             # multiple rules in one line
  validator.rule(:password, :not_empty)                    # a single rule on a line
  validator.rule(:password, :length => { :minimum => 3 })  # a rule that takes parameters

if validator.valid?
    # save the data somewhere
  else
    @errors = validator.errors
  end
```

The first paramater can be any message that the object responds to.

### Writing your own rules

If you have a custom rule you need to write, you can create a custom rule class for it:

```ruby
  class MyCustomRule
    def error_key
      :my_custom_rule
    end

def valid_value?(value)
      # Logic for determining the validity of the value
    end

def params
      {}
    end
  end
```

A rule class should have the following methods on it:

- `error_key` a symbol to represent the error. This shows up in the errors hash.  Must be an underscored_version of the class name
  - `valid_value?(value)` the beef of the rule. This is where you determine if the value is valid or not
  - `params` the params hash that was passed into the constructor

If you add your custom rule class to the `Validation::Rule` namespace, you can reference it using a symbol:

```ruby
  validator.rule(:field, :my_custom_rule)  # resolves to Validation::Rule::MyCustomRule
  validator.rule(:field, :my_custom_rule => { :param => :value })
```

Otherwise, just pass in the rule class itself:

```ruby
  validator.rule(:field, MyProject::CustomRule)
  validator.rule(:field, MyProject::CustomRule => { :param => :value })
```

### Writing self-contained validators

You can also create self-contained validation classes if you don't like the dynamic creation approach:

```ruby
  require 'validation'
  require 'validation/rule/not_empty'

class MyFormValidator < Validation::Validator
    include Validation

rule :email, :not_empty
  end
```

Now you can use this anywhere in your code:

```ruby
  form_validator = MyFormValidator.new(OpenStruct.new(params))
  form_validator.valid?
```

# Semantic Versioning

This project conforms to [semver](http://semver.org/).

# Contributing

Have an improvement? Have an awesome rule you want included? Simple!

1. Fork the repository
 2. Create a branch off of the `master` branch
 3. Write specs for the change
 4. Add your change
 5. Submit a pull request to merge against the `master` branch

Don't change any version files or gemspec files in your change.

Validator

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README