pulls code out of carpentries episodes - not the most reliable!

extract_code.py

import sys
import os
import re

path = os.path.dirname(sys.argv[1])
path = "." if path == "" else path
episode = os.path.splitext(os.path.basename(sys.argv[1]))[0]

with open(f"{path}/{episode}.md") as f:
    content = f.read()

content = re.sub(r"^(>\s)*", "", content, flags=re.MULTILINE)  # strip all "> "
content = re.sub("(\%matplotlib widget)", "", content)  # strip notebook magics
matches = re.findall(
    "(\~\~\~\n)(?!{)([\w\W]*?)(\~\~\~\n)(.*)",
    content,
)  # find all code blocks

code = ""

for m in matches:
    if "python" in m[3]:  # python code blocks only
        code += m[1]

print(code)

with open(f"{path}/{episode}.py", "w") as f:
    f.write(code)