[Python #8] create e41.py
This commit is contained in:
parent
155b48f5bd
commit
da6e94a5e7
2 changed files with 89 additions and 0 deletions
87
kjg/python-the-hard-way/e41.py
Normal file
87
kjg/python-the-hard-way/e41.py
Normal file
|
@ -0,0 +1,87 @@
|
|||
import random
|
||||
from urllib.request import urlopen
|
||||
import sys
|
||||
|
||||
WORD_URL = "http://learncodethehardway.org/words.txt"
|
||||
WORDS = []
|
||||
|
||||
PHRASES = {
|
||||
"class %%%(%%%):":
|
||||
"Make a class named %%% that is-a %%%.",
|
||||
"class %%%(object):\n\tdef __init__(self, ***)" :
|
||||
"class %%% has-a __init__ that takes self and *** params.",
|
||||
"class %%%(object):\n\tdef ***(self, @@@)":
|
||||
"class %%% has-a function *** that takes self and @@@ params.",
|
||||
"*** = %%%()":
|
||||
"Set *** to an instance of class %%%.",
|
||||
"***.***(@@@)":
|
||||
"From *** get the *** function, call it with params self, @@@.",
|
||||
"***.*** = '***'":
|
||||
"From *** get the *** attribute and set it to '***'."
|
||||
}
|
||||
|
||||
# do they want to drill phrases first
|
||||
if len(sys.argv) == 2 and sys.argv[1] == "english":
|
||||
PHRASE_FIRST = True
|
||||
else:
|
||||
PHRASE_FIRST = False
|
||||
|
||||
|
||||
# load up the words from the website
|
||||
for word in urlopen(WORD_URL).readlines():
|
||||
WORDS.append(str(word.strip(), encoding="utf-8"))
|
||||
|
||||
|
||||
def convert(snippet, phrase):
|
||||
class_names = [w.capitalize() for w in
|
||||
random.sample(WORDS, snippet.count("%%%"))]
|
||||
other_names = random.sample(WORDS, snippet.count("***"))
|
||||
results = []
|
||||
param_names = []
|
||||
|
||||
for i in range(0, snippet.count("@@@")):
|
||||
param_count = random.randint(1,3)
|
||||
param_names.append(', '.join(
|
||||
random.sample(WORDS, param_count)))
|
||||
|
||||
for sentence in snippet, phrase:
|
||||
# this is how you duplicate a list or string
|
||||
result = sentence[:]
|
||||
|
||||
# fake class names
|
||||
for word in class_names:
|
||||
result = result.replace("%%%", word, 1)
|
||||
|
||||
# fake other names
|
||||
for word in other_names:
|
||||
result = result.replace("***", word, 1)
|
||||
|
||||
# fake parameter lists
|
||||
for word in param_names:
|
||||
result = result.replace("@@@", word, 1)
|
||||
|
||||
results.append(result)
|
||||
return results
|
||||
|
||||
|
||||
# keep going until they hit CTRL-D
|
||||
try:
|
||||
while True:
|
||||
snippets = list(PHRASES.keys())
|
||||
random.shuffle(snippets)
|
||||
|
||||
for snippet in snippets:
|
||||
phrase = PHRASES[snippet]
|
||||
question, answer = convert(snippet, phrase)
|
||||
if PHRASE_FIRST:
|
||||
question, answer = answer, question
|
||||
|
||||
print(question)
|
||||
input("> ")
|
||||
print(f"ANSWER: {answer}\n\n")
|
||||
|
||||
except EOFError:
|
||||
print("\nBye")
|
||||
|
||||
|
||||
|
|
@ -145,3 +145,5 @@ How to use dictionary
|
|||
*** Python #8
|
||||
**** ex40
|
||||
How to use class
|
||||
**** ex41
|
||||
practice class
|
||||
|
|
Loading…
Reference in a new issue