python-scripts

This is a repository of random Python scripts I've made.

Character Markov Chain

CharacterMarkovChain.py is an implementation of a Markov chain to generate realistic-looking words based on a list of source words, such as an English dictionary or a list of names. The markov_chain method goes through every word in the source list and creates a dictionary of the probabilities that each letter is followed by each other letter, and then the generate_word method uses that dictionary as a Markov chain to generate new words. The generate_word_within method is just like generate_word, but it restricts the length of the generated word to within a specified range.

Here are some example generate_word outputs given the source list ['lasagna', 'spaghetti', 'rigatoni', 'macaroni', 'ravioli', 'linguine', 'fettuccine']:

saspaguigui
rolasaghe
spararonetti
ragninaspa
lionghe

Homomorse

Homomorses are pairs or sets of words with the same configuration of dots and dashes, or mits (Morse digits), ignoring spaces. Homomorse.py finds every set of homomorses in the English dictionary text file, enable1.txt.

The homomorse set with the most mits turns out to be the words with the mit configuration -.....---.-..-----...-.-..-.-...-..-.--, which includes:

biologically, theologically

The largest homomorse set turns out to be the words with the mit configuration -....--...., which includes:

babe, bans, bates, bath, begs, deans, death, digs, dimes, duns, neeps, nips, tsadi

Run the script to see the entire collection of English homomorses.

LetterBoxed

LetterBoxed.py is a solver for Sam Ezersky's online puzzle Letter Boxed. The rules of the puzzle can be seen here.

This script takes one command line argument, namely a twelve-letter string consisting of each side of the Letter Boxed square. The order of sides or letters within a side doesn't matter, as long as letters within a side are consecutive in the input. The script outputs every possible two-word solution using the enable1.txt word list. This word list may vary from the one Sam uses, but it should still output one or more valid solutions whenever possible.

WordList OneLook

WordListOneLook.py takes in two command line arguments, a wordlist text file and a pattern of letters, and outputs every word in the wordlist that matches the given pattern. Patterns are case-insensitive and take on the style of OneLook searches, where the ? character indicates one wildcard character and the * character indicates a string of any number (including zero) of wildcard characters.

For example, the input P??T*T* indicates words with the first letter P, fourth letter T, and a T anywhere after the fourth letter. With my wordlist, the output looks like this:

I found 17 results:
PARTEDTHEREDSEA
PARTYROCKANTHEM
POETRYANTHOLOGY
PARTSTHEREDSEA
PARTTHEREDSEA
POTTYTRAINING
POETLAUREATE
PORTAPOTTIES
POTTYTRAINED
POTTYTRAINS
PENTATONIX
PLOTTWISTS
PORTAPOTTY
POTTYMOUTH
POTTYTRAIN
PRETTYSURE
PLOTTWIST

I've found this script very useful for crossword construction and anything that involves filtering through a dictionary or other wordlist file.