From e3edaf8144161b7ce2b7337179edd903375070ec Mon Sep 17 00:00:00 2001
From: Safwan Samsudeen <62411302+safwansamsudeen@users.noreply.github.com>
Date: Mon, 31 Jul 2023 20:45:56 +0530
Subject: [PATCH] Approaches for Atbash Cipher (#3457)

* first round

* correct typo

* correct yet another typo

* Apply suggestions from code review

Co-authored-by: BethanyG <BethanyG@users.noreply.github.com>

* replace most single variable names

* replace i to index

* Apply suggestions from code review

Committing suggestions so the PR can be merged.

---------

Co-authored-by: BethanyG <BethanyG@users.noreply.github.com>
---
 .../atbash-cipher/.approaches/config.json     | 21 ++++++++
 .../atbash-cipher/.approaches/introduction.md | 46 +++++++++++++++++
 .../.approaches/mono-function/content.md      | 46 +++++++++++++++++
 .../.approaches/mono-function/snippet.txt     |  8 +++
 .../.approaches/separate-functions/content.md | 51 +++++++++++++++++++
 .../separate-functions/snippet.txt            |  8 +++
 6 files changed, 180 insertions(+)
 create mode 100644 exercises/practice/atbash-cipher/.approaches/config.json
 create mode 100644 exercises/practice/atbash-cipher/.approaches/introduction.md
 create mode 100644 exercises/practice/atbash-cipher/.approaches/mono-function/content.md
 create mode 100644 exercises/practice/atbash-cipher/.approaches/mono-function/snippet.txt
 create mode 100644 exercises/practice/atbash-cipher/.approaches/separate-functions/content.md
 create mode 100644 exercises/practice/atbash-cipher/.approaches/separate-functions/snippet.txt

diff --git a/exercises/practice/atbash-cipher/.approaches/config.json b/exercises/practice/atbash-cipher/.approaches/config.json
new file mode 100644
index 0000000000..ed1edeb506
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/config.json
@@ -0,0 +1,21 @@
+{
+  "introduction": {
+    "authors": ["safwansamsudeen"]
+  },
+  "approaches": [
+    {
+      "uuid": "920e6d08-e8fa-4bef-b2f4-837006c476ae",
+      "slug": "mono-function",
+      "title": "Mono-function",
+      "blurb": "Use one function for both tasks",
+      "authors": ["safwansamsudeen"]
+    },
+    {
+      "uuid": "9a7a17e0-4ad6-4d97-a8b9-c74d47f3e000",
+      "slug": "separate-functions",
+      "title": "Separate Functions",
+      "blurb": "Use separate functions, and perhaps helper ones",
+      "authors": ["safwansamsudeen"]
+    }
+  ]
+}
diff --git a/exercises/practice/atbash-cipher/.approaches/introduction.md b/exercises/practice/atbash-cipher/.approaches/introduction.md
new file mode 100644
index 0000000000..dff79f96c6
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/introduction.md
@@ -0,0 +1,46 @@
+# Introduction
+Atbash cipher in Python can be solved in many ways.  
+
+## General guidance
+The first thing is to have a "key" mapping - possibly in a `dict` or `str.maketrans`, otherwise the value would have to be calculated on the fly.
+Then, you have to "clean" up the string to be encoded by removing numbers/whitespace.
+Finally, you break it up into chunks of five before returning it.
+
+For decoding, it's similar - clean up (which automatically joins the chunks) and translate using the _same_ key - the realization that the same key can be used is crucial in solving this in an idiomatic manner.
+
+## Approach: separate functions
+We use `str.maketrans` to create the encoding. 
+In `encode`, we use a [generator expression][generator expression] in `str.join`.
+```python
+from string import ascii_lowercase
+ENCODING = str.maketrans(ascii_lowercase, ascii_lowercase[::-1])
+
+def encode(text: str):
+    res = "".join(chr for chr in text.lower() if chr.isalnum()).translate(ENCODING)
+    return " ".join(res[index:index+5] for index in range(0, len(res), 5))
+
+def decode(text: str):
+    return "".join(chr.lower() for chr in text if chr.isalnum()).translate(ENCODING)
+```
+Read more on this [approach here][approach-seperate-functions].
+
+## Approach: mono-function
+Notice that there the majority of the code is repetitive? 
+A fun way to solve this would be to keep it all inside the `encode` function, and merely chunk it if `decode` is False:
+For variation, this approach shows a different way to translate the text.
+```python
+from string import ascii_lowercase as asc_low
+ENCODING = {chr: asc_low[id] for id, chr in enumerate(asc_low[::-1])}
+
+def encode(text: str, decode: bool = False):
+    res = "".join(ENCODING.get(chr, chr) for chr in text.lower() if chr.isalnum())
+    return res if decode else " ".join(res[index:index+5] for index in range(0, len(res), 5))
+
+def decode(text: str):
+    return encode(text, True)
+```
+For more detail, [read here][approach-mono-function].
+
+[approach-separate-functions]: https://exercism.org/tracks/python/exercises/atbash-cipher/approaches/separate-functions
+[approach-mono-function]: https://exercism.org/tracks/python/exercises/atbash-cipher/approaches/mono-function
+[generator expression]: https://www.programiz.com/python-programming/generator
diff --git a/exercises/practice/atbash-cipher/.approaches/mono-function/content.md b/exercises/practice/atbash-cipher/.approaches/mono-function/content.md
new file mode 100644
index 0000000000..879664ce20
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/mono-function/content.md
@@ -0,0 +1,46 @@
+## Approach: Mono-function
+Notice that there the majority of the code is repetitive? 
+A fun way to solve this would be to keep it all inside the `encode` function, and merely chunk it if `decode` is False:
+For variation, this approach shows a different way to translate the text.
+```python
+from string import ascii_lowercase as asc_low
+ENCODING = {chr: asc_low[id] for id, chr in enumerate(asc_low[::-1])}
+
+def encode(text: str, decode: bool = False):
+    res = "".join(ENCODING.get(chr, chr) for chr in text.lower() if chr.isalnum())
+    return res if decode else " ".join(res[index:index+5] for index in range(0, len(res), 5))
+
+def decode(text: str):
+    return encode(text, True)
+```
+To explain the translation: we use a `dict` comprehension in which we reverse the ASCII lowercase digits, and enumerate through them - that is, `z` is 0, `y` is 1, and so on. 
+We access the character at that index and set it to the value of `c` - so `z` translates to `a`.
+
+In the calculation of the result, we try to obtain the value of the character using `dict.get`, which accepts a default parameter. 
+In this case, the character itself is the default - that is, numbers won't be found in the translation key, and thus should remain as numbers.
+
+We use a [ternary operator][ternary-operator] to check if we actually mean to decode the function, in which case we return the result as is. 
+If not, we chunk the result by joining every five characters with a space.
+
+Another possible way to solve this would be to use a function that returns a function that encodes or decodes based on the parameters:
+```python
+from string import ascii_lowercase as alc
+
+lowercase = {chr: alc[id] for id, chr in enumerate(alc[::-1])}
+
+def code(decode=False):
+    def func(text):
+        line = "".join(lowercase.get(chr, chr) for chr in text.lower() if chr.isalnum())
+        return line if decode else " ".join(line[index:index+5] for index in range(0, len(line), 5))
+    return func
+
+    
+encode = code()
+decode = code(True)
+```
+The logic is the same - we've instead used one function that generates two _other_ functions based on the boolean value of its parameter.
+`encode` is set to the function that's returned, and performs encoding.
+`decode` is set a function that _decodes_.
+
+[ternary-operator]: https://www.tutorialspoint.com/ternary-operator-in-python
+[decorator]: https://realpython.com/primer-on-python-decorators/
\ No newline at end of file
diff --git a/exercises/practice/atbash-cipher/.approaches/mono-function/snippet.txt b/exercises/practice/atbash-cipher/.approaches/mono-function/snippet.txt
new file mode 100644
index 0000000000..84e8b79300
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/mono-function/snippet.txt
@@ -0,0 +1,8 @@
+from string import ascii_lowercase as asc_low
+ENCODING = {chr: asc_low[id] for id, chr in enumerate(asc_low[::-1])}
+
+def encode(text: str, decode: bool = False):
+    res = "".join(ENCODING.get(chr, chr) for chr in text.lower() if chr.isalnum())
+    return res if decode else " ".join(res[index:index+5] for index in range(0, len(res), 5))
+def decode(text: str):
+    return encode(text, True)
\ No newline at end of file
diff --git a/exercises/practice/atbash-cipher/.approaches/separate-functions/content.md b/exercises/practice/atbash-cipher/.approaches/separate-functions/content.md
new file mode 100644
index 0000000000..60e02a2205
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/separate-functions/content.md
@@ -0,0 +1,51 @@
+## Approach: Separate Functions
+We use `str.maketrans` to create the encoding. 
+`.maketrans`/`.translate` is extremely fast compared to other methods of translation.
+If you're interested, [read more][str-maketrans] about it.
+
+In `encode`, we use a [generator expression][generator-expression] in `str.join`, which is more efficient - and neater - than a list comprehension.
+```python
+from string import ascii_lowercase
+ENCODING = str.maketrans(ascii_lowercase, ascii_lowercase[::-1])
+
+def encode(text: str):
+    res = "".join(chr for chr in text.lower() if chr.isalnum()).translate(ENCODING)
+    return " ".join(res[index:index+5] for index in range(0, len(res), 5))
+
+def decode(text: str):
+    return "".join(chr.lower() for chr in text if chr.isalnum()).translate(ENCODING)
+```
+In `encode`, we first join together every character if the character is alphanumeric - as we use `text.lower()`, the characters are all lowercase as needed.
+Then, we translate it and return a version joining every five characters with a space in between.
+
+`decode` does the exact same thing, except it doesn't return a chunked output. 
+Instead of cleaning the input by checking that it's alphanumeric, we check that it's not a whitespace character.
+
+It might be cleaner to use helper functions:
+```python
+from string import ascii_lowercase
+ENCODING = str.maketrans(ascii_lowercase, ascii_lowercase[::-1])
+def clean(text):
+    return "".join([chr.lower() for chr in text if chr.isalnum()])
+def chunk(text):
+    return " ".join(text[index:index+5] for index in range(0, len(text), 5))
+
+def encode(text):
+    return chunk(clean(text).translate(ENCODING))
+
+def decode(text):
+    return clean(text).translate(ENCODING)
+```
+Note that checking that `chr` _is_ alphanumeric achieves the same result as checking that it's _not_ whitespace, although it's not as explicit.
+As this is a helper function, this is acceptable enough.
+
+You can also make `chunk` recursive:
+```python
+def chunk(text):
+    if len(text) <= 5:
+        return text
+    return text[:5] + " " + chunk(text[5:])
+```
+
+[generator-expression]: https://www.programiz.com/python-programming/generator
+[str-maketrans]: https://www.programiz.com/python-programming/methods/string/maketrans
\ No newline at end of file
diff --git a/exercises/practice/atbash-cipher/.approaches/separate-functions/snippet.txt b/exercises/practice/atbash-cipher/.approaches/separate-functions/snippet.txt
new file mode 100644
index 0000000000..fbfe0b75fa
--- /dev/null
+++ b/exercises/practice/atbash-cipher/.approaches/separate-functions/snippet.txt
@@ -0,0 +1,8 @@
+from string import ascii_lowercase
+ENCODING = str.maketrans(ascii_lowercase, ascii_lowercase[::-1])
+
+def encode(text: str):
+    res = "".join(chr for chr in text.lower() if chr.isalnum()).translate(ENCODING)
+    return " ".join(res[index:index+5] for index in range(0, len(res), 5))
+def decode(text: str):
+    return "".join(chr.lower() for chr in text if not chr.isspace()).translate(ENCODING)
\ No newline at end of file