Fix accounts search by full/partial display name and others (mastodon…

…#11580)

- Restrict followers counts to local users to minimize local advantage
- Fix emoji shortcodes causing error in search
- Fix search syntax parse errors not being caught

app/chewy/accounts_index.rb

@@ -26,10 +26,17 @@ class AccountsIndex < Chewy::Index
   define_type ::Account.searchable.includes(:account_stat), delete_if: ->(account) { account.destroyed? || !account.searchable? } do
     root date_detection: false do
       field :id, type: 'long'
-      field :display_name, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content'
-      field :acct, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content', value: ->(account) { [account.username, account.domain].compact.join('@') }
-      field :following_count, type: 'long', value: ->(account) { account.active_relationships.count }
-      field :followers_count, type: 'long', value: ->(account) { account.passive_relationships.count }
+
+      field :display_name, type: 'text', analyzer: 'content' do
+        field :edge_ngram, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content'
+      end
+
+      field :acct, type: 'text', analyzer: 'content', value: ->(account) { [account.username, account.domain].compact.join('@') } do
+        field :edge_ngram, type: 'text', analyzer: 'edge_ngram', search_analyzer: 'content'
+      end
+
+      field :following_count, type: 'long', value: ->(account) { account.following.local.count }
+      field :followers_count, type: 'long', value: ->(account) { account.followers.local.count }
       field :last_status_at, type: 'date', value: ->(account) { account.last_status_at || account.created_at }
     end
   end

app/lib/search_query_parser.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+class SearchQueryParser < Parslet::Parser
+  rule(:term)      { match('[^\s":]').repeat(1).as(:term) }
+  rule(:quote)     { str('"') }
+  rule(:colon)     { str(':') }
+  rule(:space)     { match('\s').repeat(1) }
+  rule(:operator)  { (str('+') | str('-')).as(:operator) }
+  rule(:prefix)    { (term >> colon).as(:prefix) }
+  rule(:shortcode) { (colon >> term >> colon.maybe).as(:shortcode) }
+  rule(:phrase)    { (quote >> (term >> space.maybe).repeat >> quote).as(:phrase) }
+  rule(:clause)    { (prefix.maybe >> operator.maybe >> (phrase | term | shortcode)).as(:clause) }
+  rule(:query)     { (clause >> space.maybe).repeat.as(:query) }
+  root(:query)
+end

app/lib/search_query_transformer.rb

@@ -0,0 +1,88 @@
+# frozen_string_literal: true
+
+class SearchQueryTransformer < Parslet::Transform
+  class Query
+    attr_reader :should_clauses, :must_not_clauses, :must_clauses
+
+    def initialize(clauses)
+      grouped = clauses.chunk(&:operator).to_h
+      @should_clauses = grouped.fetch(:should, [])
+      @must_not_clauses = grouped.fetch(:must_not, [])
+      @must_clauses = grouped.fetch(:must, [])
+    end
+
+    def apply(search)
+      should_clauses.each { |clause| search = search.query.should(clause_to_query(clause)) }
+      must_clauses.each { |clause| search = search.query.must(clause_to_query(clause)) }
+      must_not_clauses.each { |clause| search = search.query.must_not(clause_to_query(clause)) }
+      search.query.minimum_should_match(1)
+    end
+
+    private
+
+    def clause_to_query(clause)
+      case clause
+      when TermClause
+        { multi_match: { type: 'most_fields', query: clause.term, fields: ['text', 'text.stemmed'] } }
+      when PhraseClause
+        { match_phrase: { text: { query: clause.phrase } } }
+      else
+        raise "Unexpected clause type: #{clause}"
+      end
+    end
+  end
+
+  class Operator
+    class << self
+      def symbol(str)
+        case str
+        when '+'
+          :must
+        when '-'
+          :must_not
+        when nil
+          :should
+        else
+          raise "Unknown operator: #{str}"
+        end
+      end
+    end
+  end
+
+  class TermClause
+    attr_reader :prefix, :operator, :term
+
+    def initialize(prefix, operator, term)
+      @prefix = prefix
+      @operator = Operator.symbol(operator)
+      @term = term
+    end
+  end
+
+  class PhraseClause
+    attr_reader :prefix, :operator, :phrase
+
+    def initialize(prefix, operator, phrase)
+      @prefix = prefix
+      @operator = Operator.symbol(operator)
+      @phrase = phrase
+    end
+  end
+
+  rule(clause: subtree(:clause)) do
+    prefix   = clause[:prefix][:term].to_s if clause[:prefix]
+    operator = clause[:operator]&.to_s
+
+    if clause[:term]
+      TermClause.new(prefix, operator, clause[:term].to_s)
+    elsif clause[:shortcode]
+      TermClause.new(prefix, operator, ":#{clause[:term]}:")
+    elsif clause[:phrase]
+      PhraseClause.new(prefix, operator, clause[:phrase].map { |p| p[:term].to_s }.join(' '))
+    else
+      raise "Unexpected clause type: #{clause}"
+    end
+  end
+
+  rule(query: sequence(:clauses)) { Query.new(clauses) }
+end

app/services/account_search_service.rb

@@ -67,7 +67,7 @@ def simple_search_results
   end
 
   def from_elasticsearch
-    must_clauses   = [{ multi_match: { query: terms_for_query, fields: likely_acct? ? %w(acct) : %w(acct^2 display_name), type: 'best_fields' } }]
+    must_clauses   = [{ multi_match: { query: terms_for_query, fields: likely_acct? ? %w(acct.edge_ngram acct) : %w(acct.edge_ngram acct display_name.edge_ngram display_name), type: 'most_fields', operator: 'and' } }]
     should_clauses = []
 
     if account

app/services/search_service.rb

@@ -52,7 +52,7 @@ def perform_statuses_search!
     preloaded_relations = relations_map_for_account(@account, account_ids, account_domains)
 
     results.reject { |status| StatusFilter.new(status, @account, preloaded_relations).filtered? }
-  rescue Faraday::ConnectionFailed
+  rescue Faraday::ConnectionFailed, Parslet::ParseFailed
     []
   end