Merge pull request #198 from aparibocci/feature/boyer_moore_horspool_search

Implementing `Boyer-Moore-Horspool` substring search algorithm

strings/boyer_moore_horspool_search.rb

@@ -0,0 +1,60 @@
+##
+# This class represents a table of {bad_match_character => slide_offset}
+# to be used in Boyer-Moore-Horspool substring finding algorithm.
+
+class BadMatchTable
+
+    attr_reader :pattern
+    attr_reader :table
+
+    def initialize(pattern)
+        @pattern = pattern
+        @table = {}
+        for i in 0...pattern.size
+            @table[pattern[i]] = pattern.size - 1 - i
+        end
+    end
+
+    ##
+    # Given a mismatch character belonging to the search string, returns
+    # the offset to be used when sliding the pattern towards the right.
+
+    def slide_offset(mismatch_char)
+        table.fetch(mismatch_char, pattern.size)
+    end
+end
+
+##
+# Returns the first starting index of the given pattern's occurrence (as a substring)
+# in the provided search string if a match is found, -1 otherwise.
+
+def first_match_index(search_string, pattern)
+    matches = matches_indices(search_string, pattern, true)
+    matches.empty? ? -1 : matches[0]
+end
+
+##
+# Returns the list of starting indices of the given pattern's occurrences (as a substring)
+# in the provided search string.
+# If no match is found, an empty list is returned.
+# If `stop_at_first_match` is provided as `true`, the returned list will contain at most one element,
+# being the leftmost encountered match in the search string.
+
+def matches_indices(search_string, pattern, stop_at_first_match=false)
+    table = BadMatchTable.new(pattern)
+    i = pattern.size - 1
+    indices = []
+    while i < search_string.size
+        for j in 0...pattern.size
+            if search_string[i-j] != pattern[pattern.size-1-j]
+                i += table.slide_offset(search_string[i-j])
+                break
+            elsif j == pattern.size-1
+                indices.append(i-j)
+                return indices if stop_at_first_match
+                i += 1
+            end
+        end
+    end
+    indices
+end

strings/boyer_moore_horspool_search_test.rb

@@ -0,0 +1,20 @@
+require 'minitest/autorun'
+require_relative 'boyer_moore_horspool_search'
+
+class TestBoyerMooreHorspoolSearch < Minitest::Test
+  def test_first_match_returns_negative_index_if_no_match
+    assert first_match_index('abcdefghijk', 'defz') < 0
+  end
+
+  def test_first_match_returns_first_match_index
+    assert first_match_index('abcdefghijkghilmno', 'ghi') == 6
+  end
+
+  def test_match_indices_returns_empty_list_if_no_match
+    assert matches_indices('abcdefghijk', 'defz').empty?
+  end
+
+  def test_match_indices_returns_list_of_match_indices
+    assert matches_indices('abcdefghijkghilmno', 'ghi') == [6, 11]
+  end
+end