Created
May 9, 2014 23:32
-
-
Save jots/65215a6fce042b7d7857 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env ruby | |
require "pp" | |
def cdb_hash(s) | |
r = 5381 | |
# & 0xffffffff forces to 32 bit. (lots of hair for that) | |
s.split("").each { |c| r = (((r << 5) + r) ^ c.ord) & 0xffffffff } | |
return r | |
end | |
CDBFILENAME = "pure_ruby_cdb.cdb" | |
def writecdb(ofn=CDBFILENAME,infn="rubycdb.txt",sep="\t") | |
f = File.open(ofn,"wb") | |
# leave room for header pointers | |
p = 256*8 | |
f.seek(p) | |
buckets = []; 256.times {|x| buckets[x] = [] } | |
# open "rb" is very important for UTF8!! | |
File.open(infn,"rb").each do |line| | |
line.chomp! | |
k,v = line.split(sep) | |
f.write [klen=k.length,vlen=v.length].pack("LL") | |
f.write k + v | |
h = cdb_hash(k) | |
buckets[h % 256] << [h,p] | |
p += 8+klen+vlen | |
end | |
hashpos = p | |
# create the hash data | |
buckets.each do |buck| | |
next if (blen = buck.length) == 0 | |
ncells = blen*2 | |
cells = [] | |
ncells.times {|x| cells[x] = [0,0] } | |
buck.each do |h,p| | |
i = (h >> 8) % ncells | |
while cells[i][1] != 0 # in use... collision! | |
i = (i+1) % ncells | |
end | |
cells[i] = [h,p] | |
end | |
# now write them out. | |
cells.each do |c| | |
f.write c.pack("LL") | |
end | |
end | |
# write the header | |
f.seek(0) | |
buckets.each_with_index do |buck,x| | |
f.write [hashpos,buck.length*2].pack("LL") | |
hashpos += (buck.length*2)*8 | |
end | |
f.close | |
end | |
class CDBReader | |
BPOS = 0 | |
NCELLS = 1 | |
def initialize(fn=CDBFILENAME) | |
@f = File.open(fn,"rb") | |
@r = [] | |
# keep pointers handy | |
256.times do |x| | |
@r[x] = @f.read(8).unpack("LL") | |
end | |
end | |
def get(k) | |
h = cdb_hash(k) | |
buck_pos,ncells = @r[h % 256] | |
return "" if ncells == 0 # XXX no hit | |
start = (h >> 8) % ncells | |
ncells.times do |i| | |
@f.seek(buck_pos + ((start+i) % ncells)*8) | |
h1,p1 = @f.read(8).unpack("LL") | |
return "" if p1 == 0 # XXX no hit | |
if h1 == h | |
@f.seek(p1) | |
klen,vlen = @f.read(8).unpack("LL") | |
k1 = @f.read(klen) | |
return @f.read(vlen) if k1 == k | |
end | |
end | |
return "" # XXX no hit | |
end | |
end | |
if __FILE__ == $0 | |
# create test file | |
WORDS = "/usr/share/dict/words" | |
f = File.open("words.txt","wb") | |
File.open(WORDS,"rb").each do |line| | |
line.chomp! | |
f.puts [line,line.length].join("\t") | |
end | |
f.close | |
# create cdb file. | |
writecdb(CDBFILENAME,"words.txt") | |
# create cdb file using native cdb.. | |
system("cdb -m -c native_cdb.cdb words.txt") | |
# compare output | |
system("cmp -l #{CDBFILENAME} native_cdb.cdb") | |
end |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment