From 2ecf49a09976fa92de612d754de73a03f93cac83 Mon Sep 17 00:00:00 2001
From: dovisutu <40313014+dovisutu@users.noreply.github.com>
Date: Tue, 28 Apr 2020 10:13:23 +0800
Subject: [PATCH 1/2] Add code and image for huffman
---
.../huffman_encoding/code/scratch/huffman.svg | 98 +++++++++++
.../huffman_encoding/code/scratch/huffman.txt | 154 ++++++++++++++++++
contents/huffman_encoding/huffman_encoding.md | 5 +
3 files changed, 257 insertions(+)
create mode 100644 contents/huffman_encoding/code/scratch/huffman.svg
create mode 100644 contents/huffman_encoding/code/scratch/huffman.txt
diff --git a/contents/huffman_encoding/code/scratch/huffman.svg b/contents/huffman_encoding/code/scratch/huffman.svg
new file mode 100644
index 000000000..c23cce06c
--- /dev/null
+++ b/contents/huffman_encoding/code/scratch/huffman.svg
@@ -0,0 +1,98 @@
+
\ No newline at end of file
diff --git a/contents/huffman_encoding/code/scratch/huffman.txt b/contents/huffman_encoding/code/scratch/huffman.txt
new file mode 100644
index 000000000..05d23b9f7
--- /dev/null
+++ b/contents/huffman_encoding/code/scratch/huffman.txt
@@ -0,0 +1,154 @@
+define Huffman encodes (string)
+set [encoded v] to () // empty here
+if <(string) = ()> then
+else
+ delete all of [Letter-Node Pointer v]
+ delete all of [Nodes' Values List v]
+ delete all of [Operating Nodes' List v]
+ delete all of [Node-lhs Index v]
+ delete all of [bit list v]
+ delete all of [codebook v]
+ initialize leaves with (string)
+ make tree
+ if <(0) = (item (root) of [Node-lhs Index v])> then
+ binary count
+ else
+ repeat (length of [Letter-Node Pointer v])
+ add (plcaceholder) to [codebook v]
+ end
+ generate codebook from (root)
+ format (string) using codebook
+ end
+end
+
+define initialize leaves with (string)
+set [i v] to (1)
+repeat (length of (string))
+ if <[Letter-Node Pointer v] contains (letter (i) of (string))> then // increment
+ set [temp v] to ((1) + (item (item # of (letter (i) of (string)) in [Letter-Node Pointer v]) of [Nodes' Values List v]))
+ replace item (item # of (letter (i) of (string)) in [Letter-Node Pointer v]) of [Nodes' Values List v] with (tmp)
+ else
+ add (1) to [Nodes' Values List v]
+ add (letter (i) of (string)) to [Letter-Node Pointer v]
+ end
+ change [i v] by (1)
+end
+
+define Make tree
+set [i v] to (1)
+repeat (length of [Nodes' Values List v]) // initialize
+ add (item (i) of [Nodes' Values List v]) to [Operating Nodes' List v]
+ add (0) to [Node-lhs Index v] // leaves don't have children
+ add (0) to [Node-rhs Index v]
+ change [i v] by (1)
+end
+set [count v] to (length of [Nodes' Values List v])
+repeat until <(count) = (1)>
+ set [i v] to (2) // find smallest 2 nodes
+ set [minimum v] to (item (1) of [Operating Nodes' List v])
+ set [lhs-index v] to (1)
+ set [rhs-index v] to (1)
+ repeat ((length of [Operating Nodes' List]) - (1))
+ if <(item (i) of [Operating Nodes' List v]) < (minimum)> then
+ set [minimum v] to (item (i) of [Operating Nodes' List v])
+ set [lhs-index v] to (i)
+ end
+ change [i v] by (1)
+ end
+ set [i v] to (2)
+ replace item (lhs-index) of [Operating Nodes' List v] with (2147483647) // Int.MaxValue
+ set [2nd minimum v] to (item (1) of [Operating Nodes' List v])
+ repeat ((length of [Operating Nodes' List]) - (1))
+ if <(item (i) of [Operating Nodes' List v]) < (2nd minimum)> then
+ set [2nd minimum v] to (item (i) of [Operating Nodes' List v])
+ set [rhs-index v] to (i)
+ end
+ change [i v] by (1)
+ end
+ set [i v] to (2)
+ replace item (rhs-index) of [Operating Nodes' List v] with (2147483647) // Int.MaxValue
+ change [count v] by (-1)
+ add ((minimum) + (2nd minimum)) to [Nodes' Values List v]
+ add ((minimum) + (2nd minimum)) to [Operating Nodes' List v]
+ add (lhs-index) to [Node-lhs Index v]
+ add (rhs-index) to [Node-rhs Index v]
+end
+set [root v] to (length of [Nodes' Values List])
+
+define generate codebook from (root) // depth-first search
+if <(0) = (item (root) of [Node-lhs Index v])> then
+ set [tmp v] to () // empty hehre
+ set [i v] to (1)
+ repeat (length of [bit list v])
+ set [tmp v] to (join (tmp) (item (i) of [bit list v]))
+ change [i v] by (1)
+ end
+ replace item (root) of [codebook v] with (tmp)
+else
+ add (1) to [bit list v] // 1 for lhs
+ generate codebook from (item (root) of [Node-lhs Index v])
+ delete (length of [bit list v]) of [bit list v]
+ add (0) to [bit list v] // 0 for rhs
+ generate codebook from (item (root) of [Node-rhs Index v])
+ delete (length of [bit list v]) of [bit list v]
+end
+
+define format (string) using codebook
+set [i v] to (1)
+repeat (length of (string))
+ set [encoded v] to (join (encoded) (item (item # of (letter (i) of (string)) in [Letter-Node Pointer v]) of [codebook v]))
+ change [i v] by (1)
+end
+
+define binary count
+set [tmp v] to (item (1) of [Nodes' Values List v])
+repeat until <(0) = (tmp)>
+ add ((tmp) mod (2)) to [bit list v]
+ set [tmp v] to (((tmp) - ((tmp) mod (2))) / (2))
+end
+set [i v] to (length of [bit list v])
+repeat until <(0) = (i)>
+ set [encoded v] to (join (encoded) (item (i) of [bit list v]))
+ change [i v] by (-1)
+end
+
+define Huffman decodes (string)
+set [decoded v] to () // empty here
+if <(1) = (length of [Letter-Node Pointer v])> then
+ binary resume using (string) // special care: see issue #659
+else
+ set [i v] to (1)
+ set [index v] to (root)
+ repeat (length of (string))
+ if <(letter (i) of (string)) = (1)> then
+ set [index v] to (item (index) of [Node-lhs Index v])
+ else
+ set [index v] to (item (index) of [Node-rhs Index v])
+ end
+ if <(item (index) of [Node-lhs Index v]) = (0)> then
+ set [decoded v] to (join (decoded) (item (index) of [Letter-Node Pointer v]))
+ set [index v] to (root)
+ end
+ change [i v] by (1)
+ end
+end
+
+define binary resume using (string)
+set [tmp v] to (0)
+set [i v] to (1)
+repeat (length of (string))
+ set [tmp v] to (((tmp) * (2)) + (letter (i) of (string)))
+ change [i v] by (1)
+end
+repeat (tmp)
+ set [decoded v] to (join (decoded) (item (1) of [Letter-Node Pointer v]))
+end
+
+when @greenflag clicked
+set [to be encoded v] to [bibbity bobbity]
+Huffman encodes (to be encoded)
+say (join [Encoded, the string looks like: ] (encoded)) for (3) seconds
+Huffman decodes (encoded)
+say (join [And decoded, the string looks like: ] (decoded)) for (3) seconds
+say (join (join [As opposed to the original, which is ] ((length of (to be encoded)) * (8))) [bits,]) for (2) seconds
+say (join (join [The encoded has size ] (length of (encoded))) [.]) for (1) seconds
diff --git a/contents/huffman_encoding/huffman_encoding.md b/contents/huffman_encoding/huffman_encoding.md
index 72b70c9e5..f100a8f38 100644
--- a/contents/huffman_encoding/huffman_encoding.md
+++ b/contents/huffman_encoding/huffman_encoding.md
@@ -93,6 +93,11 @@ Whether you use a stack or straight-up recursion also depends on the language, b
[import, lang:"asm-x64"](code/asm-x64/huffman.s)
{% sample lang="scala" %}
[import, lang:"scala"](code/scala/huffman_encoding.scala)
+{% sample lang="scratch" %}
+The code snippet was taken from this [scratch project](https://scratch.mit.edu/projects/389604255/)
+
+