active_median 0.2.7 → 0.3.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0060d29bd7e382bb2cd06e63df45278b9dfe1c2cb9a2210bd3e82dd973fc90ef
-  data.tar.gz: 6d12c8ecf51e432fdcd659df62f81f629b29ea7476cc9e3719ce072ae86df7cd
+  metadata.gz: 7589b66aa9d74af6afe94a766603a36feadf9b0f36d7be01212e97a4ae1de423
+  data.tar.gz: de6fa949edc2220610f97ff42a03ce7f50cc1cf9e66979846c570c3a13e3e6b8
 SHA512:
-  metadata.gz: 56c5060c952f8c729c6463946db701c2ff6a503a2258252ab8b3b9efacd43b6719ad542d22cb488cf8d4d83f41222ce0d4bf5e0985f948cc1adfea9bf5a5467b
-  data.tar.gz: 3c2ec783cbe6a85e7e35814359e767dbb714b36be9578e46a0e5583dec5b74c19feb920192dcd2e2ace9e6482e296ccdc24659135a519d9413d79f53bed8b3de
+  metadata.gz: d2739f3b083658090a61b283d5d416f61e4b998576fe8a704cbcb22a7c2fb621f8b1c89c422bae8cf67cf3784624d5f056837980c2fa80737df67edbc337490c
+  data.tar.gz: 9bac410d399bf81668e999d28da3f804ad4ab7e8891db726bbb07980a201902ca107d866b0f27866170df41c97c63d72cf4be4768666f67197f593d559da88b7

data/CHANGELOG.md

@@ -1,3 +1,22 @@
+## 0.3.2 (2021-08-16)
+
+- Added support for SQLite without an extension
+
+## 0.3.1 (2021-02-21)
+
+- Fixed error with relations with `select` clauses
+
+## 0.3.0 (2021-02-08)
+
+- Added column alias
+- Added support for percentiles with SQLite (switched extensions)
+- Raise `ActiveRecord::UnknownAttributeReference` for non-attribute arguments
+- Dropped support for Active Record < 5.2
+
+## 0.2.8 (2021-01-16)
+
+- Fixed bug with removing order
+
 ## 0.2.7 (2021-01-16)
 
 - Fixed error with order

data/README.md

@@ -4,11 +4,12 @@ Median and percentile for Active Record, Mongoid, arrays, and hashes
 
 Supports:
 
-- PostgreSQL 9.4+
-- MariaDB 10.3.3+
-- MySQL and SQL (with extensions)
-- SQL Server 2012+
-- MongoDB 2.1+
+- PostgreSQL
+- SQLite
+- MariaDB
+- MySQL (with an extension)
+- SQL Server
+- MongoDB
 
 :fire: Uses native functions for blazing performance
 
@@ -22,7 +23,7 @@ Add this line to your application’s Gemfile:
 gem 'active_median'
 ```
 
-For MySQL and SQLite, also follow [these instructions](#additional-instructions).
+For MySQL, also follow [these instructions](#additional-instructions).
 
 ## Models
 
@@ -44,19 +45,6 @@ Works with grouping, too
 Order.group(:store_id).median(:total)
 ```
 
-## User Input
-
-If passing user input as the column, be sure to sanitize it first [like you must](https://rails-sqli.org/) with other aggregate methods like `sum`.
-
-```ruby
-column = params[:column]
-
-# check against permitted columns
-raise "Unpermitted column" unless ["column_a", "column_b"].include?(column)
-
-User.median(column)
-```
-
 ## Arrays and Hashes
 
 Median
@@ -94,10 +82,18 @@ mysql <options> < load.sql
 
 ### SQLite
 
-SQLite requires a [community extension](https://www.sqlite.org/contrib). Download [extension-functions.c](https://www.sqlite.org/contrib/download/extension-functions.c) and follow the [instructions for compiling loadable extensions](https://www.sqlite.org/loadext.html#compiling_a_loadable_extension) for your platform. On Mac, use:
+Improve performance with SQLite with an extension. Download [percentile.c](https://www.sqlite.org/src/file?name=ext/misc/percentile.c&ci=d49c32e6e7cc341b) and follow the [instructions for compiling loadable extensions](https://www.sqlite.org/loadext.html#compiling_a_loadable_extension) for your platform.
+
+On Linux, use:
 
 ```sh
-gcc -g -fPIC -dynamiclib extension-functions.c -o extension-functions.dylib
+gcc -g -fPIC -shared -lsqlite3 percentile.c -o percentile.so
+```
+
+On Mac, use:
+
+```sh
+gcc -g -fPIC -dynamiclib -L/usr/local/opt/sqlite/lib -lsqlite3 percentile.c -o percentile.dylib
 ```
 
 To load it in Rails, create an initializer with:
@@ -105,15 +101,21 @@ To load it in Rails, create an initializer with:
 ```ruby
 db = ActiveRecord::Base.connection.raw_connection
 db.enable_load_extension(1)
-db.load_extension("extension-functions.dylib")
+db.load_extension("percentile.so") # or percentile.dylib
 db.enable_load_extension(0)
 ```
 
 ## Upgrading
 
-### 0.2.0
+### 0.3.0
+
+ActiveMedian 0.3.0 protects against unsafe input by default. For non-attribute arguments, use:
+
+```ruby
+Item.median(Arel.sql(known_safe_value))
+```
 
-A user-defined function is no longer needed for Postgres. Create a migration with `ActiveMedian.drop_function` to remove it.
+Also, percentiles are now supported with SQLite. Use the [percentile extension](#sqlite) instead of `extension-functions`.
 
 ## Contributing
 

data/lib/active_median.rb

@@ -4,6 +4,7 @@ require "active_median/enumerable"
 require "active_median/version"
 
 module ActiveMedian
+  # TODO remove in 0.4.0
   def self.drop_function
     ActiveRecord::Base.connection.execute <<-SQL
       DROP AGGREGATE IF EXISTS median(anyelement);

data/lib/active_median/model.rb

@@ -1,20 +1,34 @@
 module ActiveMedian
   module Model
     def median(column)
-      percentile(column, 0.5)
+      calculate_percentile(column, 0.5, "median")
     end
 
     def percentile(column, percentile)
+      calculate_percentile(column, percentile, "percentile")
+    end
+
+    private
+
+    def calculate_percentile(column, percentile, operation)
       percentile = percentile.to_f
       raise ArgumentError, "percentile is not between 0 and 1" if percentile < 0 || percentile > 1
 
       # basic version of Active Record disallow_raw_sql!
       # symbol = column (safe), Arel node = SQL (safe), other = untrusted
       # matches table.column and column
-      unless column.is_a?(Symbol) || column.is_a?(Arel::Nodes::SqlLiteral) || /\A\w+(\.\w+)?\z/i.match(column.to_s)
-        warn "[active_median] Non-attribute argument: #{column}. Use Arel.sql() for known-safe values. This will raise an error in ActiveMedian 0.3.0"
+      unless column.is_a?(Symbol) || column.is_a?(Arel::Nodes::SqlLiteral)
+        column = column.to_s
+        unless /\A\w+(\.\w+)?\z/i.match(column)
+          raise ActiveRecord::UnknownAttributeReference, "Query method called with non-attribute argument(s): #{column.inspect}. Use Arel.sql() for known-safe values."
+        end
       end
 
+      column_alias = relation.send(:column_alias_for, "#{operation} #{column.to_s.downcase}")
+      # safety check
+      # could quote, but want to keep consistent with Active Record
+      raise "Bad column alias: #{column_alias}. Please report a bug." unless column_alias =~ /\A[a-z0-9_]+\z/
+
       # column resolution
       node = relation.send(:arel_columns, [column]).first
       node = Arel::Nodes::SqlLiteral.new(node) if node.is_a?(String)
@@ -25,6 +39,9 @@ module ActiveMedian
 
       group_values = all.group_values
 
+      # replace select to match behavior of average
+      relation = unscope(:select)
+
       relation =
         case connection.adapter_name
         when /mysql/i
@@ -37,37 +54,36 @@ module ActiveMedian
               over = "PARTITION BY #{group_values.join(", ")}"
             end
 
-            select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column}) OVER (#{over})").unscope(:group)
+            relation.select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column}) OVER (#{over}) AS #{column_alias}").unscope(:group)
           else
             # if mysql gets native function, check (and memoize) version first
-            select(*group_values, "PERCENTILE_CONT(#{column}, #{percentile})")
+            relation.select(*group_values, "PERCENTILE_CONT(#{column}, #{percentile}) AS #{column_alias}")
           end
         when /sqlserver/i
           if group_values.any?
             over = "PARTITION BY #{group_values.join(", ")}"
           end
 
-          select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column}) OVER (#{over})").unscope(:group)
+          relation.select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column}) OVER (#{over}) AS #{column_alias}").unscope(:group)
         when /sqlite/i
-          case percentile.to_f
-          when 0
-            select(*group_values, "MIN(#{column})")
-          when 0.5
-            select(*group_values, "MEDIAN(#{column})")
-          when 1
-            select(*group_values, "MAX(#{column})")
-          else
-            # LOWER_QUARTILE and UPPER_QUARTILE use different calculation than 0.25 and 0.75
-            raise "SQLite only supports 0, 0.5, and 1 percentiles"
+          db = connection.raw_connection
+          unless db.instance_variable_get(:@active_median_handler)
+            if db.get_first_value("SELECT 1 FROM pragma_function_list WHERE name = 'percentile'").nil?
+              require "active_median/sqlite_handler"
+              db.create_aggregate_handler(ActiveMedian::SQLiteHandler)
+            end
+            db.instance_variable_set(:@active_median_handler, true)
           end
+
+          relation.select(*group_values, "PERCENTILE(#{column}, #{percentile} * 100) AS #{column_alias}")
         when /postg/i, /redshift/i # postgis too
-          select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column})")
+          relation.select(*group_values, "PERCENTILE_CONT(#{percentile}) WITHIN GROUP (ORDER BY #{column}) AS #{column_alias}")
         else
           raise "Connection adapter not supported: #{connection.adapter_name}"
         end
 
       # same as average
-      relation = relation.unscope(:order).distinct!(false) if relation.group_values.empty?
+      relation = relation.unscope(:order).distinct!(false) if group_values.empty?
 
       result = connection.select_all(relation.to_sql)
 

data/lib/active_median/sqlite_handler.rb

@@ -0,0 +1,32 @@
+module ActiveMedian
+  class SQLiteHandler
+    def self.arity
+      2
+    end
+
+    def self.name
+      "percentile"
+    end
+
+    def initialize
+      @values = []
+      @percentile = nil
+    end
+
+    # skip checks for
+    # 1. percentile between 0 and 100
+    # 2. percentile same for all rows
+    # since input is already checked
+    def step(ctx, value, percentile)
+      raise ActiveRecord::StatementInvalid, "1st argument to percentile() is not numeric" unless value.is_a?(Numeric)
+      @percentile ||= percentile
+      @values << value
+    end
+
+    def finalize(ctx)
+      if @values.any?
+        ctx.result = @values.percentile(@percentile / 100.0)
+      end
+    end
+  end
+end

data/lib/active_median/version.rb

@@ -1,3 +1,3 @@
 module ActiveMedian
-  VERSION = "0.2.7"
+  VERSION = "0.3.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: active_median
 version: !ruby/object:Gem::Version
-  version: 0.2.7
+  version: 0.3.2
 platform: ruby
 authors:
 - Andrew Kane
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-01-16 00:00:00.000000000 Z
+date: 2021-08-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -16,100 +16,16 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '5'
+        version: '5.2'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '5'
-- !ruby/object:Gem::Dependency
-  name: bundler
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: minitest
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: pg
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: rake
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: activerecord
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: groupdate
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
+        version: '5.2'
 description:
-email: andrew@chartkick.com
+email: andrew@ankane.org
 executables: []
 extensions: []
 extra_rdoc_files: []
@@ -121,6 +37,7 @@ files:
 - lib/active_median/enumerable.rb
 - lib/active_median/model.rb
 - lib/active_median/mongoid.rb
+- lib/active_median/sqlite_handler.rb
 - lib/active_median/version.rb
 homepage: https://github.com/ankane/active_median
 licenses:
@@ -134,14 +51,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '2.4'
+      version: '2.6'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.3
+rubygems_version: 3.2.22
 signing_key:
 specification_version: 4
 summary: Median and percentile for Active Record, Mongoid, arrays, and hashes