From c484938b7831fdb5540b18820be107aa47304dea Mon Sep 17 00:00:00 2001
From: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Date: Sun, 29 Jul 2007 01:42:11 +0200
Subject: [PATCH] Maximum markov order is now an initialization parameter

---
 mark2.rb | 42 +++++++++++++++++-------------------------
 1 file changed, 17 insertions(+), 25 deletions(-)

diff --git a/mark2.rb b/mark2.rb
index 3738ec5..a986627 100644
--- a/mark2.rb
+++ b/mark2.rb
@@ -83,32 +83,24 @@ class ChanceHash
 end
 
 class MarkovChainer
-  # Maximum depth
-  MAX_ORDER = 5
-
   # Word or nonword regexp:
   # can be used to scan a string splitting it into
   # words and nonwords.
   WNW = /\w+|\W/u
 
-  def initialize
+  attr_reader :max_order
+  def initialize(ord=5)
     # mkv[i] holds the chains of order i
-    @mkv = Array.new
-
-    # Each chain is in the form
-    # [:array, :of, :symbols] => {
-    #   :prev => ChanceHash,
-    #   :next => ChanceHash
-    # }
-    # except for order 0, which is a simple ChanceHash
-    # itself
-    @mkv[0] = ChanceHash.new
-    MAX_ORDER.times { |i|
-      @mkv[i+1] = Hash.new { |hash, key|
-        hash[key] = {:prev => ChanceHash.new, :next => ChanceHash.new}
-      }
+    @max_order = ord+1
+    @mkv = Array.new(@max_order) { |i|
+      if i==0
+        ChanceHash.new
+      else
+        @mkv[i] = Hash.new { |hash, key|
+          hash[key] = {:prev => ChanceHash.new, :next => ChanceHash.new}
+        }
+      end
     }
-
   end
 
   def add_one(sym)
@@ -118,7 +110,7 @@ class MarkovChainer
 
   def add_before(array, prev)
     raise "Not enough words in new data" if array.empty?
-    raise "Too many words in new data" if array.size > MAX_ORDER
+    raise "Too many words in new data" if array.size > @max_order
     size = array.size
     h = @mkv[size][array.dup]
     h[:prev].increase(prev)
@@ -126,14 +118,14 @@ class MarkovChainer
 
   def add_after(array, nxt)
     raise "Not enough words in new data" if array.empty?
-    raise "Too many words in new data" if array.size > MAX_ORDER
+    raise "Too many words in new data" if array.size > @max_order
     size = array.size
     h = @mkv[size][array.dup]
     h[:next].increase(nxt)
   end
 
   def add_multi(array)
-    raise "Too many words in new data" if array.size > MAX_ORDER + 1
+    raise "Too many words in new data" if array.size > @max_order + 1
     add_before(array.butfirst, array.first)
     add_after(array.butlast, array.last)
   end
@@ -152,7 +144,7 @@ class MarkovChainer
     syms.push(nil)
 
     syms.size.times { |i|
-      ([MAX_ORDER, syms.size-i].min+1).times { |ord|
+      ([@max_order, syms.size-i].min+1).times { |ord|
         v = syms[i, ord+1]
         # puts "Learning #{v.inspect}"
         add(*v)
@@ -175,7 +167,7 @@ class MarkovChainer
   end
 
   def raw_next(syms)
-    ar = syms.last([MAX_ORDER, syms.size].min)
+    ar = syms.last([@max_order, syms.size].min)
     ord = ar.size
     if ord == 0
       @mkv[0].random
@@ -194,7 +186,7 @@ class MarkovChainer
   end
 
   def raw_prev(syms)
-    ar = syms.first([MAX_ORDER, syms.size].min)
+    ar = syms.first([@max_order, syms.size].min)
     ord = ar.size
     if ord == 0
       @mkv[0].random
-- 
2.32.0.93.g670b81a890