2 files changed, 38 insertions, 5 deletions

diff --git a/src/huffmantree.nim b/src/huffmantree.nim
index 1711879..adcaec7 100644
--- a/src/huffmantree.nim
+++ b/src/huffmantree.nim
@@ -14,6 +14,7 @@
 # You should have received a copy of the GNU Affero General Public License
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
+import tables, heapqueue
 import integers, bitreader, bitwriter
 
 const valueLengthFieldBitLength* = 6  # 64
@@ -28,19 +29,32 @@ type HuffmanTreeNode*[T: SomeUnsignedInt] = ref object
       left, right: HuffmanTreeNode[T]
     of leaf:
       value: T
+  weight: int
 
 proc huffmanBranch*[T](left, right: HuffmanTreeNode[T]): HuffmanTreeNode[T] =
-  HuffmanTreeNode[T](kind: branch, left: left, right: right)
+  HuffmanTreeNode[T](kind: branch, left: left, right: right, weight: left.weight + right.weight)
 
-proc huffmanLeaf*[T](value: T): HuffmanTreeNode[T] =
-  HuffmanTreeNode[T](kind: leaf, value: value)
+proc huffmanLeaf*[T](value: T, weight = 0): HuffmanTreeNode[T] =
+  HuffmanTreeNode[T](kind: leaf, value: value, weight: weight)
 
 proc `==`*[T](a, b: HuffmanTreeNode[T]): bool =
-  if a.kind != b.kind: return false
+  if a.kind != b.kind or a.weight != b.weight: return false
   case a.kind:
     of branch: a.left == b.left and a.right == b.right
     of leaf: a.value == b.value
 
+proc `~=`*[T](a, b: HuffmanTreeNode[T]): bool =
+  if a.kind != b.kind: return false
+  case a.kind:
+    of branch: a.left ~= b.left and a.right ~= b.right
+    of leaf: a.value == b.value
+
+proc `!~`*[T](a, b: HuffmanTreeNode[T]): bool =
+  not (a ~= b)
+
+proc `<`*[T](left, right: HuffmanTreeNode[T]): bool =
+  left.weight < right.weight
+
 proc maxValue*[T](node: HuffmanTreeNode[T]): T =
   case node.kind:
     of branch: max(node.left.maxValue(), node.right.maxValue())
@@ -68,3 +82,12 @@ proc serialise*[T](tree: HuffmanTreeNode[T], bitWriter: BitWriter) =
         bitWriter.writeBits(valueBitLength, node.value)
   bitWriter.writeBits(valueLengthFieldBitLength, valueBitLength.uint8)
   writeNode(tree)
+
+proc symbolQueue*[T](stats: CountTableRef[T]): HeapQueue[HuffmanTreeNode[T]] =
+  result = newHeapQueue[HuffmanTreeNode[T]]()
+  for item, count in stats.pairs: result.push(huffmanLeaf(item, count))
+
+proc buildHuffmanTree*[T: SomeUnsignedInt](stats: CountTableRef[T]): HuffmanTreeNode[T] =
+  var symbolQueue = symbolQueue(stats)
+  while symbolQueue.len > 1: symbolQueue.push(huffmanBranch(symbolQueue.pop(), symbolQueue.pop()))
+  result = symbolQueue[0]
diff --git a/src/lzsschain.nim b/src/lzsschain.nim
index 8203cb8..073aa5e 100644
--- a/src/lzsschain.nim
+++ b/src/lzsschain.nim
@@ -15,7 +15,7 @@
 # along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
 import lists, tables, sugar
-import polyfill, integers, lzssnode
+import polyfill, integers, lzssnode, huffmantree
 
 const maxChainByteLength = 32_000 * wordBitLength
 
@@ -34,3 +34,13 @@ proc decode*(lzssChain: LzssChain): seq[uint8] =
       of reference:
         let absolutePos = result.len - node.relativePos
         result.add(result.toOpenArray(absolutePos, absolutePos + node.length - 1))
+
+proc stats*(lzssChain: LzssChain): tuple[characters: CountTableRef[uint8], lengths, positions: CountTableRef[int]] =
+  result = (newCountTable[uint8](), newCountTable[int](), newCountTable[int]())
+  for node in lzssChain.items:
+    case node.kind:
+      of character:
+        result.characters.inc(node.character)
+      of reference:
+        result.lengths.inc(node.length)
+        result.positions.inc(node.relativePos)