tableschema 0.4.1 → 0.5.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: e38a1a391b93666fb188277be7cad9a1d553a315
-  data.tar.gz: 86cba39876c3d6111942964cef2b812c456d4ba5
+  metadata.gz: af9d2ac90e521ace72b172703da49b220a957978
+  data.tar.gz: 95a2e84830de62dbe00e0b4a107222b60d48bdd4
 SHA512:
-  metadata.gz: 4b593795d0ca14c17d876bbc28941454ad956ba72d4fc3d645eea82eb0631716b61382efd7a466cd4320ab133cea27534b66aa373a9e408b17c5bbf8323c73c4
-  data.tar.gz: 28b6e768e4190537b316ceb5ae8f45731b90114a664aa7ae99a7acd3e30a70df099e155af7c0078f6a72ff9b6c8c67eee4803a4e2cac3a1930b91f8449a44f2d
+  metadata.gz: e9de112e6d3f5bc137fde4fd9e972ca2ff97a6b818e3a66e184bc3d0868c9e77e71221ad76bbad56a2014753a24bef71705607063b08a96afa730b6dbd4114d0
+  data.tar.gz: 5caf4a5bde09a437ccf4d7a6cb91ae44fc07395366179ac682b6ba9b0a90dc966a5d5c51f2f80b70ae0911250aa5244e59b0b77e9df8821901929663807f631b

data/.travis.yml

@@ -2,8 +2,8 @@
 language: ruby
 
 rvm:
-- 2.3.1
-- 2.4.1
+- 2.3.3
+- 2.4.0
 
 before_install:
   gem install bundler -v 1.11.2

data/README.md

@@ -46,11 +46,11 @@ The gem `jsontableschema` is no longer maintained. Here are the steps to transit
 2. Replace module name `JsonTableSchema` with module name `TableSchema`. For example:
 
     ```ruby
-    JsonTableSchema::Table.infer_schema(csv)
+    JsonTableSchema::Table.new(source, schema)
     ```
     with
     ```ruby
-    TableSchema::Table.infer_schema(csv)
+    TableSchema::Table.new(source, schema)
     ```
 
 ## Usage
@@ -91,9 +91,9 @@ table.read
 ```
 
 Both `iter` and `read` take the optional parameters:
-- `row_limit`: integer, default `nil` - stop at this many rows
-- `cast`: boolean, default `true` - cast values for each row
 - `keyed`: boolean, default: `false` - return the rows as Hashes with headers as keys
+- `cast`: boolean, default `true` - cast values for each row
+- `limit`: integer, default `nil` - stop at this many rows
 
 ### Infer a schema
 
@@ -102,7 +102,8 @@ If you don't have a schema for a CSV, and want to generate one, you can infer a
 ```ruby
 csv = 'https://github.com/frictionlessdata/tableschema-rb/raw/master/spec/fixtures/simple_data.csv' # Can also be a url or array of arrays
 
-table = TableSchema::Table.infer_schema(csv)
+table = TableSchema::Table.new(csv, nil)
+table.infer()
 table.schema
 #=> {:fields=>[{:name=>"id", :title=>"", :description=>"", :type=>"integer", :format=>"default", :constraints=>{}}, {:name=>"title", :title=>"", :description=>"", :type=>"string", :format=>"default", :constraints=>{}}]}
 ```
@@ -169,26 +170,16 @@ schema_hash = {
 }
 schema = TableSchema::Schema.new(schema_hash)
 
-schema.headers
+schema.field_names
 #=> ["id", "height"]
-schema.required_headers
-#=> ["id"]
 schema.fields
 #=> [{:name=>"id", :type=>"string", :constraints=>{:required=>true}, :format=>"default"}, {:name=>"height", :type=>"number", :format=>"default", :constraints=>{}}]
-schema.primary_keys
+schema.primary_key
 #=> ["id"]
 schema.foreign_keys
 # => [{:fields=>"state", :reference=>{:resource=>"the-resource", :fields=>"state_id"}}]
 schema.get_field('id')
 # => {:name=>"id", :type=>"string", :constraints=>{:required=>true}, :format=>"default"}
-schema.has_field?('foo')
-#=> false
-schema.get_type('id')
-#=> 'string'
-schema.get_fields_by_type('string')
-# => [{:name=>"id", :type=>"string", :constraints=>{:required=>true}, :format=>"default"}, {:name=>"state", :type=>"string", :format=>"default", :constraints=>{}}]
-schema.get_constraints('id')
-# => {:required=>true}
 ```
 
 #### Cast row

data/lib/profiles/table-schema.json

@@ -24,13 +24,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -125,13 +120,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -162,6 +152,12 @@
                 ],
                 "default": "default"
               },
+              "bareNumber": {
+                "type": "boolean",
+                "title": "bareNumber",
+                "description": "a boolean field with a default of `true`. If `true` the physical contents of this field must follow the formatting constraints already set out. If `false` the contents of this field may contain leading and/or trailing non-numeric characters (which implementors MUST therefore strip). The purpose of `bareNumber` is to allow publishers to publish numeric data that contains trailing characters such as percentages e.g. `95%` or leading characters such as currencies e.g. `€95` or `EUR 95`. Note that it is entirely up to implementors what, if anything, they do with stripped text.",
+                "default": true
+              },
               "decimalChar": {
                 "type": "string",
                 "description": "A string whose value is used to represent a decimal point within the number. The default value is `.`."
@@ -170,10 +166,6 @@
                 "type": "string",
                 "description": "A string whose value is used to group digits within the number. The default value is `null`. A common value is `,` e.g. '100,000'."
               },
-              "currency": {
-                "type": "string",
-                "description": "A number that may include additional currency symbols."
-              },
               "constraints": {
                 "title": "Constraints",
                 "description": "The following constraints are supported for `number` fields.",
@@ -257,13 +249,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -294,6 +281,12 @@
                 ],
                 "default": "default"
               },
+              "bareNumber": {
+                "type": "boolean",
+                "title": "bareNumber",
+                "description": "a boolean field with a default of `true`. If `true` the physical contents of this field must follow the formatting constraints already set out. If `false` the contents of this field may contain leading and/or trailing non-numeric characters (which implementors MUST therefore strip). The purpose of `bareNumber` is to allow publishers to publish numeric data that contains trailing characters such as percentages e.g. `95%` or leading characters such as currencies e.g. `€95` or `EUR 95`. Note that it is entirely up to implementors what, if anything, they do with stripped text.",
+                "default": true
+              },
               "constraints": {
                 "title": "Constraints",
                 "description": "The following constraints are supported for `integer` fields.",
@@ -375,13 +368,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -462,13 +450,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -548,13 +531,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -634,13 +612,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -748,13 +721,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -841,13 +809,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -937,13 +900,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1038,13 +996,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1146,13 +1099,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1250,13 +1198,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1352,13 +1295,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1439,13 +1377,8 @@
             "properties": {
               "name": {
                 "title": "Name",
-                "description": "An identifier string. Lower case characters with `.`, `_`, `-` and `/` are allowed.",
-                "type": "string",
-                "pattern": "^([-a-z0-9._/])+$",
-                "context": "This is ideally a url-usable and human-readable name. Name `SHOULD` be invariant, meaning it `SHOULD NOT` change when its parent descriptor is updated.",
-                "examples": [
-                  "{\n  \"name\": \"my-nice-name\"\n}\n"
-                ]
+                "description": "A name for this field.",
+                "type": "string"
               },
               "title": {
                 "title": "Title",
@@ -1539,9 +1472,9 @@
           "fields",
           "reference"
         ],
-        "properties": {
-          "oneOf": [
-            {
+        "oneOf": [
+          {
+            "properties": {
               "fields": {
                 "type": "array",
                 "items": {
@@ -1572,8 +1505,10 @@
                   }
                 }
               }
-            },
-            {
+            }
+          },
+          {
+            "properties": {
               "fields": {
                 "type": "string",
                 "description": "Fields that make up the primary key."
@@ -1595,8 +1530,8 @@
                 }
               }
             }
-          ]
-        }
+          }
+        ]
       },
       "examples": [
         "{\n  \"foreignKeys\": [\n    {\n      \"fields\": \"state\",\n      \"reference\": {\n        \"resource\": \"the-resource\",\n        \"fields\": \"state_id\"\n      }\n    }\n  ]\n}\n",

data/lib/tableschema.rb

@@ -32,8 +32,6 @@ require "tableschema/types/duration"
 require "tableschema/defaults"
 
 require "tableschema/field"
-require "tableschema/validate"
-require "tableschema/model"
 require "tableschema/schema"
 require "tableschema/table"
 require "tableschema/infer"

data/lib/tableschema/defaults.rb

@@ -4,6 +4,9 @@ module TableSchema
     type: 'string',
     missing_values: [''],
     group_char: ',',
-    decimal_char: '.'
+    decimal_char: '.',
+    true_values: ['true', 'True', 'TRUE', '1'],
+    false_values: ['false', 'False', 'FALSE', '0'],
+    bare_number: true,
   }.freeze
 end

data/lib/tableschema/field.rb

@@ -4,7 +4,9 @@ module TableSchema
   class Field < Hash
     include TableSchema::Helpers
 
-    attr_reader :name, :type, :format, :missing_values, :constraints
+    # Public
+
+    attr_reader :name, :type, :format, :required, :constraints
 
     def initialize(descriptor, missing_values=nil)
       self.merge! deep_symbolize_keys(descriptor)
@@ -12,6 +14,7 @@ module TableSchema
       @type = self[:type] = self.fetch(:type, TableSchema::DEFAULTS[:type])
       @format = self[:format] = self.fetch(:format, TableSchema::DEFAULTS[:format])
       @constraints = self[:constraints] = self.fetch(:constraints, {})
+      @required = @constraints.fetch(:required, false)
       @missing_values = missing_values || default_missing_values
     end
 
@@ -19,15 +22,15 @@ module TableSchema
       self.to_h
     end
 
-    def cast_value(value, check_constraints: true)
+    def cast_value(value, constraints: true)
       cast_value = cast_type(value)
-      return cast_value if check_constraints == false
+      return cast_value if constraints == false
       TableSchema::Constraints.new(self, cast_value).validate!
       cast_value
     end
 
-    def test_value(value, check_constraints: true)
-      cast_value(value, check_constraints: check_constraints)
+    def test_value(value, constraints: true)
+      cast_value(value, constraints: constraints)
       true
     rescue TableSchema::Exception
       false
@@ -41,6 +44,8 @@ module TableSchema
       end
     end
 
+    # Private
+
     private
 
       def default_missing_values

data/lib/tableschema/helpers.rb

@@ -16,26 +16,6 @@ module TableSchema
       end
     end
 
-    def convert_to_boolean(value)
-      if value.is_a?(Boolean)
-        return value
-      elsif true_values.include?(value.to_s.downcase)
-        true
-      elsif false_values.include?(value.to_s.downcase)
-        false
-      else
-        nil
-      end
-    end
-
-    def true_values
-      ['yes', 'y', 'true', 't', '1']
-    end
-
-    def false_values
-      ['no', 'n', 'false', 'f', '0']
-    end
-
     def get_class_for_type(type)
       "TableSchema::Types::#{type_class_lookup[type.to_sym] || 'String'}"
     end

data/lib/tableschema/schema.rb

@@ -1,12 +1,14 @@
+require 'tableschema/defaults'
+
 module TableSchema
   class Schema < Hash
-    include TableSchema::Validate
-    include TableSchema::Model
     include TableSchema::Helpers
 
+    # Public
+
     attr_reader :errors
 
-    def initialize(descriptor, case_insensitive_headers: false, strict: false)
+    def initialize(descriptor, strict: false, case_insensitive_headers: false)
       self.merge! deep_symbolize_keys(parse_schema(descriptor))
       @case_insensitive_headers = case_insensitive_headers
       @strict = strict
@@ -17,26 +19,60 @@ module TableSchema
       self
     end
 
+    def validate
+      @errors = Set.new(JSON::Validator.fully_validate(@profile, self))
+      check_primary_key
+      check_foreign_keys
+      @errors.empty?
+    end
+
+    def validate!
+      validate
+      raise SchemaException.new(@errors.first) unless @errors.empty?
+      true
+    end
+
     def descriptor
       self.to_h
     end
 
-    def parse_schema(descriptor)
-      if descriptor.class == Hash
-        descriptor
-      elsif descriptor.class == String
-        begin
-          JSON.parse(open(descriptor).read, symbolize_names: true)
-        rescue Errno::ENOENT
-          raise SchemaException.new("File not found at `#{descriptor}`")
-        rescue OpenURI::HTTPError => e
-          raise SchemaException.new("URL `#{descriptor}` returned #{e.message}")
-        rescue JSON::ParserError
-          raise SchemaException.new("File at `#{descriptor}` is not valid JSON")
-        end
-      else
-        raise SchemaException.new("A schema must be a hash, path or URL")
-      end
+    def primary_key
+      [self[:primaryKey]].flatten.reject { |k| k.nil? }
+    end
+
+    def foreign_keys
+      self[:foreignKeys] || []
+    end
+
+    def fields
+      self[:fields]
+    end
+
+    def field_names
+      fields.map { |f| transform(f[:name]) }
+    rescue NoMethodError
+      []
+    end
+
+    def get_field(field_name)
+      fields.find { |f| f[:name] == field_name }
+    end
+
+    def add_field(descriptor)
+      self[:fields].push(descriptor)
+      validate!
+      descriptor
+    rescue TableSchema::SchemaException => e
+      self[:fields].pop
+      raise e if @strict
+      nil
+    end
+
+    def remove_field(field_name)
+      field = get_field(field_name)
+      self[:fields].reject!{ |f| f.name == field_name }
+      validate
+      field
     end
 
     def cast_row(row, fail_fast: true)
@@ -66,5 +102,125 @@ module TableSchema
       true
     end
 
+    # Deprecated
+
+    alias :headers :field_names
+
+    def missing_values
+      self.fetch(:missingValues, TableSchema::DEFAULTS[:missing_values])
+    end
+
+    def get_type(field_name)
+      get_field(field_name)[:type]
+    end
+
+    def get_constraints(field_name)
+      get_field(field_name)[:constraints] || {}
+    end
+
+    def required_headers
+      fields.select { |f| f.fetch(:constraints, {}).fetch(:required, nil).to_s == 'true' }
+            .map { |f| transform(f[:name]) }
+    end
+
+    def unique_headers
+      fields.select { |f| f.fetch(:constraints, {}).fetch(:unique, nil).to_s == 'true' }
+            .map { |f| transform(f[:name]) }
+    end
+
+    def has_field?(field_name)
+      get_field(field_name) != nil
+    end
+
+    def get_fields_by_type(type)
+      fields.select { |f| f[:type] == type }
+    end
+
+    # Private
+
+    private
+
+    def parse_schema(descriptor)
+      if descriptor.class == Hash
+        descriptor
+      elsif descriptor.class == String
+        begin
+          JSON.parse(open(descriptor).read, symbolize_names: true)
+        rescue Errno::ENOENT
+          raise SchemaException.new("File not found at `#{descriptor}`")
+        rescue OpenURI::HTTPError => e
+          raise SchemaException.new("URL `#{descriptor}` returned #{e.message}")
+        rescue JSON::ParserError
+          raise SchemaException.new("File at `#{descriptor}` is not valid JSON")
+        end
+      else
+        raise SchemaException.new("A schema must be a hash, path or URL")
+      end
+    end
+
+    def transform(name)
+      name.downcase! if @case_insensitive_headers == true
+      name
+    end
+
+    def expand!
+      (self[:fields] || []).each do |f|
+        f[:type] = TableSchema::DEFAULTS[:type] if f[:type] == nil
+        f[:format] = TableSchema::DEFAULTS[:format] if f[:format] == nil
+      end
+    end
+
+    def load_fields!
+      self[:fields] = (self[:fields] || []).map { |f| TableSchema::Field.new(f, missing_values) }
+    end
+
+    def load_validator!
+      filepath = File.join(File.dirname(__FILE__), '..', 'profiles', 'table-schema.json')
+      @profile ||= JSON.parse(File.read(filepath), symbolize_names: true)
+    end
+
+    def check_primary_key
+      return if self[:primaryKey].nil?
+      primary_key.each { |pk| check_field_value(pk, 'primaryKey') }
+    end
+
+    def check_foreign_keys
+      return if self[:foreignKeys].nil?
+      self[:foreignKeys].each do |key|
+        if field_type_mismatch?(key)
+          add_error("A TableSchema `foreignKey.fields` value must be the same type as `foreignKey.reference.fields`")
+        end
+        if field_count_mismatch?(key)
+          add_error("A TableSchema `foreignKey.fields` must contain the same number of entries as `foreignKey.reference.fields`")
+        end
+        foreign_key_fields(key).each { |fk| check_field_value(fk, 'foreignKey.fields') }
+        if key.fetch(:reference).fetch(:resource).empty?
+          foreign_key_fields(key.fetch(:reference)).each { |fk| check_field_value(fk, 'foreignKey.reference.fields')}
+        end
+      end
+    end
+
+    def check_field_value(key, type)
+      if headers.select { |f| key == f }.count == 0
+        add_error("The TableSchema #{type} value `#{key}` is not found in any of the schema's field names")
+      end
+    end
+
+    def foreign_key_fields(key)
+      [key.fetch(:fields)].flatten
+    end
+
+    def field_count_mismatch?(key)
+      foreign_key_fields(key).count != foreign_key_fields(key.fetch(:reference)).count
+    end
+
+    def field_type_mismatch?(key)
+      key.fetch(:fields).class.name != key.fetch(:reference).fetch(:fields).class.name
+    end
+
+    def add_error(error)
+      @errors << error
+    end
+
   end
 end

data/lib/tableschema/table.rb

@@ -1,27 +1,28 @@
 module TableSchema
   class Table
 
-    attr_reader :schema, :headers
+    # Public
 
-    def self.infer_schema(csv, csv_options: {})
-      TableSchema::Table.new(csv, nil, csv_options)
-    end
+    attr_reader :headers, :schema
 
     def initialize(csv, descriptor, csv_options: {})
       @csv_options = csv_options.merge(headers: true)
+      @descriptor = descriptor
       @csv = parse_csv(csv)
       @headers = initialize_headers
-      @schema = descriptor.nil? ? infer_schema : TableSchema::Schema.new(descriptor)
-      initialize_unique_colums
+      if !descriptor.nil?
+        @schema = TableSchema::Schema.new(@descriptor)
+        initialize_unique_colums
+      end
     end
 
-    def iter(row_limit: nil, cast: true, keyed: false)
+    def iter(keyed: false, cast: true, limit: nil)
       unless block_given?
-        return enum_for(:iter, row_limit: row_limit, cast: cast, keyed: keyed)
+        return enum_for(:iter, limit: limit, cast: cast, keyed: keyed)
       end
 
       @csv.each_with_index do |row, i|
-        break if row_limit && (row_limit <= i)
+        break if limit && (limit <= i)
         if cast == true
           cast_values = @schema.cast_row(row)
           row = CSV::Row.new(@headers, cast_values)
@@ -38,11 +39,21 @@ module TableSchema
       @csv.rewind
     end
 
-    def read(row_limit: nil, cast: true, keyed: false)
-      iterator = self.iter(row_limit: row_limit, cast: cast, keyed: keyed)
+    def read(keyed: false, cast: true, limit: nil)
+      iterator = self.iter(keyed: keyed, cast: cast, limit: limit)
       iterator.to_a
     end
 
+    def infer()
+      if !@schema
+        inferer = TableSchema::Infer.new(@headers, @csv)
+        @schema = inferer.schema
+        initialize_unique_colums
+        @csv.rewind
+      end
+      @schema.descriptor
+    end
+
     def save(target)
       CSV.open(target, "wb", @csv_options) do |csv|
         csv << @headers
@@ -51,6 +62,8 @@ module TableSchema
       true
     end
 
+    # Private
+
     private
 
     def parse_csv(csv)
@@ -62,12 +75,6 @@ module TableSchema
       array.map { |row| row.to_csv(row_sep: nil) }.join("\r\n")
     end
 
-    def infer_schema
-      inferer = TableSchema::Infer.new(@headers, @csv)
-      @csv.rewind
-      inferer.schema
-    end
-
     def initialize_headers
       headers = @csv.first.to_h.keys
       @csv.rewind

data/lib/tableschema/types/boolean.rb

@@ -25,9 +25,17 @@ module TableSchema
       end
 
       def cast_default(value)
-        value = convert_to_boolean(value)
-        raise TableSchema::InvalidCast.new("#{value} is not a #{name}") if value.nil?
-        value
+        true_values = @field.fetch(:trueValues, TableSchema::DEFAULTS[:true_values])
+        false_values = @field.fetch(:falseValues, TableSchema::DEFAULTS[:false_values])
+        if [true, false].include?(value)
+          return value
+        elsif true_values.include?(value)
+          return true
+        elsif false_values.include?(value)
+          return false
+        else
+          raise TableSchema::InvalidCast.new("#{value} is not a #{name}")
+        end
       end
 
     end

data/lib/tableschema/types/integer.rb

@@ -25,6 +25,10 @@ module TableSchema
         if value.is_a?(type)
           value
         else
+          bare_number = @field.fetch(:bareNumber, TableSchema::DEFAULTS[:bare_number])
+          if !bare_number
+            value = value.gsub(/((^\D*)|(\D*$))/, '')
+          end
           Integer(value)
         end
       rescue ArgumentError

data/lib/tableschema/types/number.rb

@@ -48,28 +48,16 @@ module TableSchema
           else
             group_char = @field.fetch(:groupChar, TableSchema::DEFAULTS[:group_char])
             decimal_char = @field.fetch(:decimalChar, TableSchema::DEFAULTS[:decimal_char])
-            formatted_value = value.gsub(group_char, '').gsub(decimal_char, '.')
-            if formatted_value.match(percent_chars)
-              process_percent(formatted_value)
-            elsif @field.fetch(:currency, nil)
-              process_currency(formatted_value)
-            else
-              Float(formatted_value)
+            bare_number = @field.fetch(:bareNumber, TableSchema::DEFAULTS[:bare_number])
+            formatted_value = value
+            formatted_value = formatted_value.gsub(group_char, '')
+            formatted_value = formatted_value.gsub(decimal_char, '.')
+            if !bare_number
+              formatted_value = formatted_value.gsub(/((^\D*)|(\D*$))/, '')
             end
+            Float(formatted_value)
           end
         end
-
-        def process_percent(value)
-          Float(value.gsub(percent_chars, '')) / 100
-        end
-
-        def process_currency(value)
-          Float(value.gsub(@field[:currency], ''))
-        end
-
-        def percent_chars
-          /%|‰|‱|％|﹪|٪/
-        end
     end
   end
 end

data/lib/tableschema/version.rb

@@ -1,3 +1,3 @@
 module TableSchema
-  VERSION = "0.4.1".freeze
+  VERSION = "0.5.0".freeze
 end

data/tableschema.gemspec

@@ -26,7 +26,7 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "coveralls", "~> 0.8.13"
   spec.add_development_dependency "rubocop", "~> 0.49.1"
 
-  spec.add_dependency "json-schema", "~> 2.6.0"
+  spec.add_dependency "json-schema", "~> 2.8.0"
   spec.add_dependency "uuid", "~> 2.3.8"
   spec.add_dependency "tod", "~> 2.1.0"
   spec.add_dependency "activesupport", "~> 5.1.0"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: tableschema
 version: !ruby/object:Gem::Version
-  version: 0.4.1
+  version: 0.5.0
 platform: ruby
 authors:
 - Open Knowledge Foundation
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2017-08-27 00:00:00.000000000 Z
+date: 2017-08-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -114,14 +114,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.6.0
+        version: 2.8.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.6.0
+        version: 2.8.0
 - !ruby/object:Gem::Dependency
   name: uuid
   requirement: !ruby/object:Gem::Requirement
@@ -201,7 +201,6 @@ files:
 - lib/tableschema/field.rb
 - lib/tableschema/helpers.rb
 - lib/tableschema/infer.rb
-- lib/tableschema/model.rb
 - lib/tableschema/schema.rb
 - lib/tableschema/table.rb
 - lib/tableschema/types/any.rb
@@ -220,7 +219,6 @@ files:
 - lib/tableschema/types/time.rb
 - lib/tableschema/types/year.rb
 - lib/tableschema/types/yearmonth.rb
-- lib/tableschema/validate.rb
 - lib/tableschema/version.rb
 - tableschema.gemspec
 homepage: https://github.com/frictionlessdata/tableschema-rb

data/lib/tableschema/model.rb

@@ -1,96 +0,0 @@
-require 'tableschema/defaults'
-
-module TableSchema
-  module Model
-
-    def headers
-      fields.map { |f| transform(f[:name]) }
-    rescue NoMethodError
-      []
-    end
-
-    alias :field_names :headers
-
-    def fields
-      self[:fields]
-    end
-
-    def primary_keys
-      [self[:primaryKey]].flatten.reject { |k| k.nil? }
-    end
-
-    def foreign_keys
-      self[:foreignKeys] || []
-    end
-
-    def missing_values
-      self.fetch(:missingValues, TableSchema::DEFAULTS[:missing_values])
-    end
-
-    def get_type(field_name)
-      get_field(field_name)[:type]
-    end
-
-    def get_constraints(field_name)
-      get_field(field_name)[:constraints] || {}
-    end
-
-    def required_headers
-      fields.select { |f| f.fetch(:constraints, {}).fetch(:required, nil).to_s == 'true' }
-            .map { |f| transform(f[:name]) }
-    end
-
-    def unique_headers
-      fields.select { |f| f.fetch(:constraints, {}).fetch(:unique, nil).to_s == 'true' }
-            .map { |f| transform(f[:name]) }
-    end
-
-    def has_field?(field_name)
-      get_field(field_name) != nil
-    end
-
-    def get_field(field_name)
-      fields.find { |f| f[:name] == field_name }
-    end
-
-    def get_fields_by_type(type)
-      fields.select { |f| f[:type] == type }
-    end
-
-    def add_field(descriptor)
-      self[:fields].push(descriptor)
-      validate!
-      descriptor
-    rescue TableSchema::SchemaException => e
-      self[:fields].pop
-      raise e if @strict
-      nil
-    end
-
-    def remove_field(field_name)
-      field = get_field(field_name)
-      self[:fields].reject!{ |f| f.name == field_name }
-      validate
-      field
-    end
-
-    private
-
-    def transform(name)
-      name.downcase! if @case_insensitive_headers == true
-      name
-    end
-
-    def expand!
-      (self[:fields] || []).each do |f|
-        f[:type] = TableSchema::DEFAULTS[:type] if f[:type] == nil
-        f[:format] = TableSchema::DEFAULTS[:format] if f[:format] == nil
-      end
-    end
-
-    def load_fields!
-      self[:fields] = (self[:fields] || []).map { |f| TableSchema::Field.new(f, missing_values) }
-    end
-
-  end
-end

data/lib/tableschema/validate.rb

@@ -1,70 +0,0 @@
-module TableSchema
-  module Validate
-
-    attr_reader :errors
-
-    def load_validator!
-      filepath = File.join(File.dirname(__FILE__), '..', 'profiles', 'table-schema.json')
-      @profile ||= JSON.parse(File.read(filepath), symbolize_names: true)
-    end
-
-    def validate
-      @errors = Set.new(JSON::Validator.fully_validate(@profile, self))
-      check_primary_keys
-      check_foreign_keys
-      @errors.empty?
-    end
-
-    def validate!
-      validate
-      raise SchemaException.new(@errors.first) unless @errors.empty?
-      true
-    end
-
-    private
-
-    def check_primary_keys
-      return if self[:primaryKey].nil?
-      primary_keys.each { |pk| check_field_value(pk, 'primaryKey') }
-    end
-
-    def check_foreign_keys
-      return if self[:foreignKeys].nil?
-      self[:foreignKeys].each do |key|
-        if field_type_mismatch?(key)
-          add_error("A TableSchema `foreignKey.fields` value must be the same type as `foreignKey.reference.fields`")
-        end
-        if field_count_mismatch?(key)
-          add_error("A TableSchema `foreignKey.fields` must contain the same number of entries as `foreignKey.reference.fields`")
-        end
-        foreign_key_fields(key).each { |fk| check_field_value(fk, 'foreignKey.fields') }
-        if key.fetch(:reference).fetch(:resource).empty?
-          foreign_key_fields(key.fetch(:reference)).each { |fk| check_field_value(fk, 'foreignKey.reference.fields')}
-        end
-      end
-    end
-
-    def check_field_value(key, type)
-      if headers.select { |f| key == f }.count == 0
-        add_error("The TableSchema #{type} value `#{key}` is not found in any of the schema's field names")
-      end
-    end
-
-    def foreign_key_fields(key)
-      [key.fetch(:fields)].flatten
-    end
-
-    def field_count_mismatch?(key)
-      foreign_key_fields(key).count != foreign_key_fields(key.fetch(:reference)).count
-    end
-
-    def field_type_mismatch?(key)
-      key.fetch(:fields).class.name != key.fetch(:reference).fetch(:fields).class.name
-    end
-
-    def add_error(error)
-      @errors << error
-    end
-
-  end
-end