gistfile1.rb

# The current version. Using self.inxex(field) makes this O(n^2)!

# Rebuild the HashWithChecksumAttribute with the current
# values of the fields Array
def reindex
  @tags = {}
  self.each do |field|
    @tags[field.tag] ||= []
    @tags[field.tag] << self.index(field) ##### AAAAAAAHHHHHHHH ####
  end
  @clean = true
end

### Instead, use this ###

def reindex
  @tags = {}
  self.each_with_index do |field, i|
    @tags[field.tag] ||= []
    @tags[field.tag] << i
  end
  @clean = true
end


### Benchmarking by  number of tags

tags = ['001','005', '100','110','111','240','243','245',  '700', '710', '711']
rec = MARC::Reader.new('batch.dat').first

Benchmark.bm do |x|
  (1..(tags.size)).each do |numtags|
    tagset = tags[0..(numtags -1 )]
    
    x.report("#{numtags} tag with #fields  ") do
      10000.times do
        rec.reindex
        tagset.each do |tag|
            t = rec.fields(tag)
        end
      end
    end
    x.report("#{numtags} tag with #find_all") do
      10000.times do
        tagset.each do |tag|
            t = rec.find_all {|f| f.tag == tag}
        end
      end
    end
    puts ""
  end
end

# 1 tag with #fields    0.630000   0.010000   0.640000 (  0.638195)
# 1 tag with #find_all  0.200000   0.000000   0.200000 (  0.206705)
# 
# 2 tag with #fields    0.730000   0.010000   0.740000 (  0.752267)
# 2 tag with #find_all  0.380000   0.000000   0.380000 (  0.389939)
# 
# 3 tag with #fields    0.840000   0.010000   0.850000 (  1.234076)
# 3 tag with #find_all  0.580000   0.000000   0.580000 (  0.679571)
# 
# 4 tag with #fields    0.910000   0.010000   0.920000 (  1.014209)
# 4 tag with #find_all  0.760000   0.010000   0.770000 (  0.796539)
# 
# 5 tag with #fields    1.010000   0.010000   1.020000 (  1.040470)
# 5 tag with #find_all  0.930000   0.000000   0.930000 (  0.972256)
# 
# 6 tag with #fields    1.090000   0.010000   1.100000 (  1.105366)
# 6 tag with #find_all  1.120000   0.010000   1.130000 (  1.180861)
# 
# 7 tag with #fields    1.190000   0.010000   1.200000 (  1.614809)
# 7 tag with #find_all  1.310000   0.010000   1.320000 (  1.359643)
# 
# 8 tag with #fields    1.270000   0.010000   1.280000 (  1.346432)
# 8 tag with #find_all  1.500000   0.020000   1.520000 (  1.579795)
# 
# 9 tag with #fields    1.360000   0.010000   1.370000 (  1.420000)
# 9 tag with #find_all  1.680000   0.010000   1.690000 (  1.863628)
# 
# 10 tag with #fields    1.460000   0.010000   1.470000 (  1.665508)
# 10 tag with #find_all  1.870000   0.020000   1.890000 (  1.976831)
# 
# 11 tag with #fields    1.520000   0.010000   1.530000 (  1.572052)
# 11 tag with #find_all  2.060000   0.020000   2.080000 (  2.151522)