schema-inference 1.2.1 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: b5da2e518ffc5bd291ce69ba382f20a3afbc2c06
-  data.tar.gz: 44f330a78d095a596841020bcf8a5e983d12921b
+  metadata.gz: c1d7af131c500e6f39e2215da26ad17f9baf961c
+  data.tar.gz: 6d9055d8b1ba6a50342c819fab15b20a2ac2192d
 SHA512:
-  metadata.gz: d030451925707394967f79dcd0e7ff287d735c5176e02bd68803aeff05800bc91273fe49157a868be8b4c21f82998c1e4619d372e20d42bd19c8d895c6156c47
-  data.tar.gz: 0ac460013951cb20a47db9d0e4d0ae0f25f61b421255603f64c206280c5b22fd0d55a8f5df2d97f97a7c4a54a2b8a1ab3b25639fbe7f67a8579dea3272139163
+  metadata.gz: 7355e4693fde74f238ef0d9e26ff469b327c61e4bdb7fb9d4fd19155e7db1dffc6221c46009c59b0eae3cb3ca6c65121cbf753f7bdc3a764ae8ea21355be3fcb
+  data.tar.gz: 01d7e5023340336bbc2d27dcf6b754dc66ad4d09ad34d3d5f2024decbbd7178ebd05f59c2d03079b7c1e88ef09bce5dbfd8077c5967e07742b53c25ac683eed2

data/README.md

@@ -1,3 +1,6 @@
+[![Build
+Status](https://travis-ci.org/Phybbit/schema-inference.svg?branch=master)](https://travis-ci.org/Phybbit/schema-inference)
+
 # Schema::Inference
 
 Supports inferring tabular schemas from deep nested data structures.

data/lib/schema/inference/schema_inferrer.rb

@@ -50,7 +50,7 @@ module Schema
         raise ArgumentError, 'dataset must be an array or a hash'
       end
 
-      FIXNUM_MAX = (2**(0.size * 8 -2) -1)
+      INT_MAX = 2_147_483_648
 
       def data_schema(data)
         table_schema = {}
@@ -76,8 +76,8 @@ module Schema
             field_schema[:types][field[:type]][:count] += 1
 
             if type_has_min_max?(field[:type])
-              field_size = value_length(field[:inferred_value])
-              field_schema[:types][field[:type]][:min] = [field_schema[:types][field[:type]][:min] || FIXNUM_MAX, field_size].min
+              field_size = value_length(field[:inferred_value], field[:type])
+              field_schema[:types][field[:type]][:min] = [field_schema[:types][field[:type]][:min] || INT_MAX, field_size].min
               field_schema[:types][field[:type]][:max] = [field_schema[:types][field[:type]][:max] || 0, field_size].max
             end
 
@@ -92,8 +92,8 @@ module Schema
         type == String || NumericTypes.include?(type)
       end
 
-      def value_length(value)
-        return value.length if value.is_a?(String)
+      def value_length(value, type)
+        return value.to_s.length if type == String
         value # leave as-is otherwise
       end
 
@@ -109,7 +109,7 @@ module Schema
               table_schema[k][:types][type] ||= { count: 0 }
               table_schema[k][:types][type][:count] += info[:count]
               if type_has_min_max?(type)
-                table_schema[k][:types][type][:min] = [table_schema[k][:types][type][:min] || FIXNUM_MAX, info[:min]].min
+                table_schema[k][:types][type][:min] = [table_schema[k][:types][type][:min] || INT_MAX, info[:min]].min
                 table_schema[k][:types][type][:max] = [table_schema[k][:types][type][:max] || 0, info[:max]].max
               end
             }
@@ -207,14 +207,20 @@ module Schema
 
       def detect_type_of(value)
         return Boolean  if value.is_a?(TrueClass) || value.is_a?(FalseClass)
-        return Integer  if value.is_a? Integer
+
+        if value.is_a? Integer
+          return Integer if value.abs <= INT_MAX
+          return String
+        end
+
         return Numeric  if value.is_a? Numeric
         return Time     if value.is_a? Time
         return NilClass if value.is_a? NilClass
 
+
         if value.is_a? String
-          return Integer if value =~ /^[-+]?[0-9]+$/
-          return Numeric if value =~ /^[-+]?[0-9]*\.?[0-9]+$/
+          return Integer if value =~ /^[-+]?[0-9]+$/ && value.to_i.abs <= INT_MAX
+          return Numeric if value =~ /^[-+]?[0-9]*\.[0-9]+$/
           return Boolean if %w(false true).include?(value.downcase)
           return Time if Timeliness.parse(value) != nil
           return String

data/lib/schema/inference/version.rb

@@ -1,5 +1,5 @@
 module Schema
   module Inference
-    VERSION = '1.2.1'
+    VERSION = '1.3.0'
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: schema-inference
 version: !ruby/object:Gem::Version
-  version: 1.2.1
+  version: 1.3.0
 platform: ruby
 authors:
 - Eurico Doirado
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-10-31 00:00:00.000000000 Z
+date: 2017-08-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -150,9 +150,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.1
+rubygems_version: 2.6.11
 signing_key: 
 specification_version: 4
 summary: Supports inferring tabular schemas from deep nested structures.
 test_files: []
-has_rdoc: