#Hecta number code#

then make the substitution in the key, decode the message, score this attempt at a decode, and decide if this attempt is an improvement or not.

If we start with the known letter ‘E’ and we are searching for ‘THE’ the regular expression will look for occurrences of ‘.E’ in our message and find the letters that are most commonly represented by the ‘.’. We add branches to the tree by searching for words using masks based on ‘known’ letters in parent branches. For this I am using a tree structure that starts with an initial ‘known’ letter based on the fx sorted plain alphabet. Each letter object stores the information about each letter (name or initial usage, fx, di usage, tri usage, reversals, and classification).Īt this point we can start looking for common words based on the initial letter classification and our initial key based on letter fx. The key itself is a list of letter objects (defined in class Letter). The di-, tri- and reversals are done using re.findall() and counting and sorting the results. To do this I have created a class called Key that handles the fx, di-, tri-, and reversal calculations and the initial classification of each letter. Then we can classify each letter in the key as either a consonant or a vowel, and either high, medium or low fx. Next, we need to extract and analyze the digraphs, trigraphs and occurrences of reversals. The first step would be to understand the letter fx, and build an initial key based on fx sorted from highest to lowest. Very rare for a vowel not to be present.

it is common for consonant vowel alterations cvc or vcv.

If a digraph and its reversal are present one letter is likely a vowel.

Consonants except t and n tend to combine with vowels more often.

Letters present in a great number of digraphs are probably vowels.

Vowels tend to combine with consonants, not other vowels.

Use as alternatives when a high fx letter is used in the message.

Group letters by High, Medium, and Low fx.

Use high fx letters and compare letters that pre- and suc- cede them to identify vowels.

Common starting and ending letters (if word length known).

Longer words can have distinct letter patters.

Patterns in words (if word length known).

Notes from chapter 2, Non-Linear or Mixed Transformations:

#Hecta number code#

Since I am just getting back into my routine and haven’t written much code in a while, this post is going to be more of an overview of the second chapter covering non-linear and random substitution, and an outline of the work I have done so far to implement the ideas presented.