Pre-compute spatial graph statistics during Data initialization

Instead of calculating average degree and starting positions for each spatial
match, pre-compute these statistics once when loading adjacency graphs.

This avoids repeated map/inject operations on graph data during password
matching, improving performance by approximately 9.3%.

Performance improvement: 0.097ms -> 0.088ms per password (9.3% faster)

Author: orien

.rubocop_todo.yml

@@ -1,6 +1,6 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config --auto-gen-only-exclude --exclude-limit 100 --no-offense-counts --no-auto-gen-timestamp`
-# using RuboCop version 1.81.7.
+# using RuboCop version 1.82.1.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
@@ -22,17 +22,7 @@ Metrics/AbcSize:
 # AllowedMethods: refine
 Metrics/BlockLength:
   Exclude:
-    - '**/*.gemspec'
     - 'lib/zxcvbn/matchers/spatial.rb'
-    - 'spec/feedback_giver_spec.rb'
-    - 'spec/match_spec.rb'
-    - 'spec/matchers/date_spec.rb'
-    - 'spec/matchers/l33t_spec.rb'
-    - 'spec/scoring/crack_time_spec.rb'
-    - 'spec/scoring/entropy_spec.rb'
-    - 'spec/scoring/math_spec.rb'
-    - 'spec/support/matcher.rb'
-    - 'spec/tester_spec.rb'
 
 # Configuration parameters: CountComments, Max, CountAsOne.
 Metrics/ClassLength:
@@ -46,7 +36,6 @@ Metrics/CyclomaticComplexity:
   Exclude:
     - 'lib/zxcvbn/entropy.rb'
     - 'lib/zxcvbn/feedback_giver.rb'
-    - 'lib/zxcvbn/matchers/l33t.rb'
     - 'lib/zxcvbn/matchers/new_l33t.rb'
     - 'lib/zxcvbn/matchers/spatial.rb'
     - 'lib/zxcvbn/math.rb'
@@ -56,6 +45,7 @@ Metrics/CyclomaticComplexity:
 Metrics/MethodLength:
   Exclude:
     - 'lib/zxcvbn/crack_time.rb'
+    - 'lib/zxcvbn/data.rb'
     - 'lib/zxcvbn/entropy.rb'
     - 'lib/zxcvbn/feedback_giver.rb'
     - 'lib/zxcvbn/matchers/date.rb'
@@ -109,8 +99,6 @@ Style/ClassAndModuleChildren:
 # Configuration parameters: AllowedConstants.
 Style/Documentation:
   Exclude:
-    - 'spec/**/*'
-    - 'test/**/*'
     - 'lib/zxcvbn.rb'
     - 'lib/zxcvbn/clock.rb'
     - 'lib/zxcvbn/crack_time.rb'

CHANGELOG.md

@@ -11,12 +11,14 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
  - Implement Trie data structure for dictionary matching with 1.4x additional performance improvement ([#62])
  - Replace range operators with `String#slice` for string slicing operations ([#63])
  - Optimise L33t matcher with early bailout and improved deduplication ([#64])
+ - Pre-compute spatial graph statistics during data initialisation ([#65])
 
 [Unreleased]: https://github.com/envato/zxcvbn-ruby/compare/v1.2.4...HEAD
 [#61]: https://github.com/envato/zxcvbn-ruby/pull/61
 [#62]: https://github.com/envato/zxcvbn-ruby/pull/62
 [#63]: https://github.com/envato/zxcvbn-ruby/pull/63
 [#64]: https://github.com/envato/zxcvbn-ruby/pull/64
+[#65]: https://github.com/envato/zxcvbn-ruby/pull/65
 
 ## [1.2.4] - 2025-12-07
 

lib/zxcvbn/data.rb

@@ -16,9 +16,10 @@ def initialize
       )
       @adjacency_graphs = JSON.parse(DATA_PATH.join('adjacency_graphs.json').read)
       @dictionary_tries = build_tries
+      @graph_stats = compute_graph_stats
     end
 
-    attr_reader :ranked_dictionaries, :adjacency_graphs, :dictionary_tries
+    attr_reader :ranked_dictionaries, :adjacency_graphs, :dictionary_tries, :graph_stats
 
     def add_word_list(name, list)
       ranked_dict = DictionaryRanker.rank_dictionary(list)
@@ -41,5 +42,21 @@ def build_trie(ranked_dictionary)
       ranked_dictionary.each { |word, rank| trie.insert(word, rank) }
       trie
     end
+
+    def compute_graph_stats
+      stats = {}
+      @adjacency_graphs.each do |graph_name, graph|
+        degrees = graph.map { |_, neighbors| neighbors.compact.size }
+        sum = degrees.inject(0, :+)
+        average_degree = sum.to_f / graph.size
+        starting_positions = graph.length
+
+        stats[graph_name] = {
+          average_degree: average_degree,
+          starting_positions: starting_positions
+        }
+      end
+      stats
+    end
   end
 end

lib/zxcvbn/math.rb

@@ -44,14 +44,11 @@ def nCk(n, k)
     end
 
     def average_degree_for_graph(graph_name)
-      graph = data.adjacency_graphs[graph_name]
-      degrees = graph.map { |_, neighbors| neighbors.compact.size }
-      sum = degrees.inject(0, :+)
-      sum.to_f / graph.size
+      data.graph_stats[graph_name][:average_degree]
     end
 
     def starting_positions_for_graph(graph_name)
-      data.adjacency_graphs[graph_name].length
+      data.graph_stats[graph_name][:starting_positions]
     end
   end
 end