Skip to content

Instantly share code, notes, and snippets.

What would you like to do?
A script to generate lesson files for gtypist. Currently generates workman-layout lessons.
import re
import itertools
import collections
import textwrap
import random
import shelve
import contextlib
workman_lessons = [
('tn', '', 'home row-first finger', 'ngrams'),
('he', '', 'home row-second finger', 'ngrams'),
('he', 'tn', 'home row-two fingers', 'words'),
# ('then', '', 'home row-two fingers', 'ngrams'),
('so', '', 'home row-third finger', 'ngrams'),
('so', 'then', 'home row-three fingers', 'words'),
('ai', '', 'home row-fourth finger', 'ngrams'),
('asht', '', 'home row-left hand', 'words'),
('noei', '', 'home row-right hand', 'words'),
('ashtneoi', '', 'home row-both hands', 'ngrams'),
('ashtneoi', '', 'home row-both hands', 'words'),
('lr', '', 'lr', 'ngrams'),
('lr', 'ashtneoi', 'homerow + lr', 'ngrams'),
('lr', 'ashtneoi', 'homerow + lr', 'words'),
('cudp', '', 'cudp', 'ngrams'),
('cudp', 'ashtneoi', 'cudp + homerow', 'ngrams'),
('cudp', 'ashtneoi', 'cudp + homerow', 'words'),
('gy', '', 'gy', 'ngrams'),
('gy', 'ashtneoi', 'gy + homerow', 'ngrams'),
('gy', 'ashtneoi', 'gy + homerow', 'words'),
('mfw', '', 'mfw', 'ngrams'),
('mfw', 'ashtneoi', 'mfw + homerow', 'ngrams'),
('mfw', 'ashtneoi', 'mfw + homerow', 'words'),
('kb', '', 'kb', 'ngrams'),
('kb', 'ashtneoi', 'kb + homerow', 'ngrams'),
('kb', 'ashtneoi', 'kb + homerow', 'words'),
('cudp', 'lrashtneoi', 'cumulative including cudp', 'words'),
('gy', 'cudplrashtneoi', 'cumulative including gy', 'words'),
('mfw', 'cudplrgyashtneoi', 'cumulative including mfw', 'words'),
('kb', 'mfwcudplrgyashtneoi', 'cumulative including kb', 'words'),
given some letters to drill on:
- drill on ngrams
- drill on words
- 3x each then random order 3x
- if there's too many words, do a random sample.
def ngrams(letters, extraletters, dictionary):
"""from the given set of letters, discover the frequency for which each
possible permutation appears within dictionary words. create a drill
according to this frequency.
if naive=True, don't calculate frequency, just determine if the
combination appears at all in any dictionary words. this is faster.
possible_re = re.compile(r'[%s]{3,}' % (letters+extraletters))
valid_re = re.compile(r'[%s]' % letters)
memo = collections.Counter()
for word in dictionary:
for match in possible_re.findall(word):
memo[match] += 1
for ngram in memo:
yield (memo[ngram], ngram)
def find_words(letters, extraletters, dictionary, min_length=2):
"""from the given set of letters, discover which words may be
created from them.
word_re = re.compile(r'^[%s]{%s,}$' % (letters+extraletters, min_length))
valid_re = re.compile(r'.*[%s]' % letters)
return (word for word in dictionary if word_re.match(word) and valid_re.match(word))
def wordlesson(words):
"""generate a lesson from the given words:
- for each word, repeat 3x
- all words, random order, repeat 3x
words = list(words)
for word in words:
yield " ".join([word] * 3)
for x in range(3):
yield " ".join(words)
def gtypist_lesson(label, words, title=None, instructions=None):
"""Given a title, instructions, and word list, generate lines in
gtypist format.
tw = textwrap.TextWrapper(subsequent_indent=" :")
yield "*:%s" % label
if title:
yield tw.fill(title)
if instructions:
yield tw.fill(instructions)
yield tw.fill(" ".join(words))
def menu(labels):
yield "*:MENU"
yield "B:Workman Lessons"
yield 'M: UP=_EXIT "These are lessons for learning the workman keyboard layout."'
for label, title in labels:
yield ' :%s "%s"' % (label, title)
def full_lesson(dictionary):
wordsused = set()
labels = []
for n, (letters, extraletters, description, lessontype) in enumerate(workman_lessons):
if lessontype == 'ngrams':
words = ngrams(letters, extraletters, dictionary)
# omit 0-frequency ngrams
words = set(ng for ng in words if ng[0])
# take 20 most-frequent ngrams
lesson_words = set(w for f, w in sorted(words, reverse=True)[:10])
for x in letters:
lesson_words = wordlesson(lesson_words)
instructions = " ".join(x.strip() for x in """
a drill on word pieces including the letters
%s. (%d found)""".strip().splitlines()) % (", ".join(letters), len(words))
title = "%s %s drill" % (description, lessontype)
label = "lesson_%d" % n
labels.append((label, title))
yield gtypist_lesson(
elif lessontype == 'words':
words = set(find_words(letters, extraletters, dictionary))
# don't use words from previous lessons
words -= wordsused
wordsused |= words
# if we have a lot of words, make a few lessons.
for lesson_number in range(min((len(words)-1)/10+1, 3)):
if len(words) >= 10:
lesson_words = set(random.sample(words, 10))
lesson_words = set(words)
words -= lesson_words
lesson_words = wordlesson(lesson_words)
instructions = " ".join(x.strip() for x in """
a drill on words that include the letters %s. (Set
%d) (%d remaining in dictionary)""".strip().splitlines()) % (
", ".join(letters), lesson_number,
title = "%s %s drill #%d" % (description, lessontype, lesson_number + 1)
label = "lesson_%d_%d" % (n, lesson_number)
labels.append((label, title))
yield gtypist_lesson(
raise ValueError("unknown lesson type", lessontype)
yield menu(labels)
def main(argv0):
dictionary_filename = '/usr/share/dict/american-english-huge'
dictionary_filename = '/usr/share/dict/words'
dictionary = set(word.strip() for word in open(dictionary_filename))
print "G:MENU"
for lesson in full_lesson(dictionary):
print "\n".join(lesson)
if __name__=="__main__":
import sys
raise SystemExit(main(*sys.argv))
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment