pg_fulltext 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 49c8f3b3250945a7e3e600a100951abffe80a0537e14fcf392d96778d9648777
-  data.tar.gz: efef8f9be9540e961015472dc3f498203c62a37856ef4f7c66f00fad30e7e5a2
+  metadata.gz: 6e4fa6010e29f5c15b3381805189d6b01e917324260dd2184aa9b15d2c33366f
+  data.tar.gz: 578ccbf2ffe79bb48c37971a33e86566693f10130c555f218bc0b86fd2190c08
 SHA512:
-  metadata.gz: 2c324ce42e53dd725bec4ee7e11a8763927fe8080059a504ba31e5546b8d528ee76c040ae4a93c6b272a744d413f695583d413d991d1db6afb846f2e8750162e
-  data.tar.gz: 90e7cd23c65e9ee883c4c65c8fb5c21d08078eedf134f9d42cf989faf8928e2edc05a40ab008c07b367191326873773704de5bc458588aa5c0bcff7edce2ae11
+  metadata.gz: 8b394cce3e4e6691c346a28252582eec471fff714dc09a95c2627a0c2e1dc411c11969ad5de86c5b2d26d5e156c8d05cc0bbdf9c367b40cee71d3b2eef6e7299
+  data.tar.gz: 17e35ccf3e285df319a30382220e3dd5faf1495d46b80d9a5b0fd6d9e8cc29ad98d98c0fc7e09da530462d5ec42fb3d5b2c455c1ce804be5a4297fa206df8403

data/README.md

@@ -0,0 +1,65 @@
+# pg_fulltext
+
+A pretty reasonable PostgreSQL fulltext implementation with minimal configuration
+
+## Installation
+
+```shell
+gem install pg_fulltext
+```
+
+or add the following to your `Gemfile`:
+
+```shell
+gem 'pg_fulltext'
+```
+
+and run `bundle install`
+
+## Rails Configuration
+
+This implementation assumes you have a `tsv` column on your model, and that you're generating a string appropriate for 
+the language you're using (we don't specify a default, so it will probably default to `'english'` depending on your 
+Postgres implementation):
+
+```ruby
+class MyModel
+  include PgFulltext::ActiveRecord
+
+  add_search_scope
+end
+```
+
+You can then use the `search` method (configurable via the first parameter of the `add_search_scope` method):
+
+```ruby
+MyModel.search('foo bar "include this phrase" !butnotthis !"and and also not this"')
+```
+
+The defaults for this include support for negation, phrases, phrase negation, and prefix searches, but those can be 
+configured per the following options:
+
+| Option            | Default | Description |
+| ----------------  | :------ | :----------- |
+| `tsvector_column` | `tsv`   | If you have a different column containing your tsvector, specify it here |
+| `search_type`     | `nil`   | Your PostgreSQL probably defaults to `'english'`, but set this to match the tsvector you've generated |
+| `order`           | `true`  | Whether or not the `order` method should be applied against the generated `rank` for the fulltext query. If you just care about returning matches and not their respective rank, set this to `false` |
+| `reorder`         | `false` | If you already have `order` set on this relation, it will take precedence over the fulltext `rank`. `reorder` will call clear, effectively clearing the existing order and applying `rank` |
+| `any_word`        | `false` | Default search uses the `&` operator, ensuring that all terms are matched in the query.  If you want to match _any_ term in the query, set this to `true` |
+| `ignore_accents`  | `false` | By default, search queries with accents will be sent through as-is. Setting this to `true` will `unaccent()` the query, which helps match `tsv` columns that have also been unaccented.  Alternatively, you can have your `tsv` column be a combination of both, and this option will be unnecesary. Requires the `unaccent` Postgres extension. |
+
+## Standalone Configuration
+
+There's not much, here, but the `PgFulltext::Query.to_tsquery_string` method will generate a nice `tsvector`-compatible 
+string for you to use as you wish.  
+
+Something this should do the trick:
+
+```ruby
+search_string = 'foo bar "include this phrase" !butnotthis !"and and also not this"'
+sql_query = <<~SQL
+  SELECT *
+  FROM my_model
+  WHERE tsv @@ to_tsvector('portuguese', #{PgFulltext::Query.to_tsquery_string(query)})
+SQL
+```

data/Rakefile

@@ -0,0 +1,7 @@
+require 'bundler'
+Bundler::GemHelper.install_tasks
+
+require 'rspec/core/rake_task'
+RSpec::Core::RakeTask.new(:spec)
+
+task default: %w[spec]

data/lib/pg_fulltext/active_record.rb

@@ -19,15 +19,15 @@ module PgFulltext
         relation,
         query,
         tsvector_column: :tsv,
-        search_type: :simple,
+        search_type: nil,
         order: true,
+        prefix: true,
         reorder: false,
         any_word: false,
         ignore_accents: false
       )
         serial = SecureRandom.hex(4)
-        table_quoted = connection.quote_table_name(table_name)
-        pk_quoted = "#{table_quoted}.#{connection.quote_column_name(primary_key)}"
+        pk_quoted = "#{quoted_table_name}.#{connection.quote_column_name(primary_key)}"
         fulltext_join_name = "pg_fulltext_#{serial}"
 
         # Build the search relation to join on
@@ -37,6 +37,7 @@ module PgFulltext
           tsvector_column: tsvector_column,
           search_type: search_type,
           any_word: any_word,
+          prefix: prefix,
           ignore_accents: ignore_accents,
         )
 
@@ -58,13 +59,13 @@ module PgFulltext
         tsvector_column: :tsv,
         search_type: nil,
         any_word: false,
+        prefix: true,
         ignore_accents: false
       )
-        tsquery_string_quoted = connection.quote(PgFulltext::Query.to_tsquery_string(query, operator: any_word ? '|' : '&'))
+        tsquery_string_quoted = connection.quote(PgFulltext::Query.to_tsquery_string(query, operator: any_word ? '|' : '&', prefix: prefix))
         tsquery_string_quoted = "unaccent(#{tsquery_string_quoted})" if ignore_accents
-        table_quoted = connection.quote_table_name(table_name)
         column_quoted = connection.quote_column_name(tsvector_column)
-        fqc_quoted = "#{table_quoted}.#{column_quoted}"
+        fqc_quoted = "#{quoted_table_name}.#{column_quoted}"
         tsquery = "to_tsquery(#{"#{connection.quote search_type}, " if search_type.present?}#{tsquery_string_quoted})"
 
         relation

data/lib/pg_fulltext/query.rb

@@ -2,10 +2,8 @@ module PgFulltext
   module Query
     def self.to_tsquery_string(query, prefix: true, operator: '&')
 
-      # Parse out all [unicode] non-word and non-quote characters
-      query.gsub!(/[^\s\p{L}"!]/, '')
-      query.gsub!(/"+/, '"')
-      query.gsub!(/\s+/, ' ')
+      # Normalize search string to a more parseable equivalent
+      query = normalize_query(query)
 
       # Collect terms
       terms = []
@@ -21,15 +19,19 @@ module PgFulltext
           # Skip if completely comprised of non-unicode word characters
           next if term.gsub(/[^\s\p{L}]/, '') == ''
 
-          if term.start_with?('!"') && !term.end_with?('"')
-            phrase_terms << format_term(term[2..-1], prefix: true)
+          if term.start_with?('!"') && term.end_with?('"')
+            terms << format_term("!#{term[2..-2]}", prefix: prefix)
+          elsif term.start_with?('"') && term.end_with?('"')
+            terms << format_term(term[1..-2], prefix: prefix)
+          elsif term.start_with?('!"') && !term.end_with?('"')
+            phrase_terms << format_term(term[2..-1], prefix: prefix)
             negate_phrase = true
           elsif term.start_with?('"') && !term.end_with?('"')
-            phrase_terms << format_term(term[1..-1], prefix: true)
+            phrase_terms << format_term(term[1..-1], prefix: prefix)
           elsif phrase_terms.length > 0
             if term.end_with?('"')
               phrase_terms << format_term(term[0..-2], prefix: prefix)
-              terms << "#{'!' if negate_phrase}(#{reject_falsy(phrase_terms, prefix: prefix).join(' <-> ')})"
+              terms << "#{'!' if negate_phrase}(#{reject_falsy(phrase_terms).join(' <-> ')})"
               phrase_terms = []
               negate_phrase = false
             else
@@ -41,7 +43,7 @@ module PgFulltext
         end
       else
         query.gsub! /["]/, ''
-        terms = reject_falsy(query.split(' ').map { |v| format_term(v, prefix: prefix) }, prefix: prefix)
+        terms = reject_falsy(query.split(' ').map { |v| format_term(v, prefix: prefix) })
       end
 
       # Join terms with operator
@@ -50,6 +52,16 @@ module PgFulltext
 
     private
 
+    def self.normalize_query(query)
+      query
+        .gsub(/[.,]/, ' ')        # Replace all periods and commas with spaces (reasonable delimiters)
+        .gsub(/[^\s\p{L}"!]/, '') # Remove all non-unicode, quote ("), and bangs (!)
+        .gsub(/"+/, '"')          # Replace Repeat quotes with single double-quote
+        .gsub(/!+/, '!')          # Replace Repeat bangs with single bang
+        .gsub(/\s+/, ' ')         # Replace repeat whitespace occurrences with single spaces
+        .strip                    # Strip space from beginning and end of line
+    end
+
     def self.format_term(term, prefix: true)
       # Remove any ! that's not at the beginning of the term, as it will break the query
       term.gsub!(/(?<!^)!/, '')
@@ -58,7 +70,7 @@ module PgFulltext
       "#{term}#{':*' if prefix}"
     end
 
-    def self.reject_falsy(terms, prefix: true)
+    def self.reject_falsy(terms)
       false_values = [nil, '', '"', '!', ':*', '":*', '!:*']
       terms.reject { |v| false_values.include?(v) }
     end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pg_fulltext
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - Adam Robertson
@@ -9,20 +9,80 @@ autorequire:
 bindir: bin
 cert_chain: []
 date: 2021-10-11 00:00:00.000000000 Z
-dependencies: []
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: activerecord
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '5.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '5.0'
+- !ruby/object:Gem::Dependency
+  name: activesupport
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '5.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '5.0'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  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: with_model
+  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'
 description: Allows simple searching with a variety of options
 email: adam@arcreative.net
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- README.md
+- Rakefile
 - lib/pg_fulltext.rb
 - lib/pg_fulltext/active_record.rb
 - lib/pg_fulltext/query.rb
 homepage: https://github.com/arcreative/pg_fulltext
 licenses:
 - MIT
-metadata: {}
+metadata:
+  homepage_uri: https://github.com/arcreative/pg_fulltext
+  source_code_uri: https://github.com/arcreative/pg_fulltext
 post_install_message:
 rdoc_options: []
 require_paths:
@@ -38,7 +98,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.4
+rubygems_version: 3.1.6
 signing_key:
 specification_version: 4
 summary: PostgreSQL fulltext search