migrate-users.py

gistfile1.txt

import argparse
import collections
import csv
import escapism
import string
import sys


def escape_email(email):
    return escapism.escape(email, safe=safe_chars, escape_char="-").lower()


if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("names", type=argparse.FileType("r"))
    parser.add_argument(
        "-r", "--rate", type=int, default=500, help="Links to create/remove per second"
    )
    args = parser.parse_args()

    reader = csv.DictReader(args.names)

    safe_chars = set(string.ascii_lowercase + string.digits)

    entries = []
    seen = []

    for row in reader:
        src = escape_email(row["jupyter_email"])
        dst = row["sis_user_id"]

        seen.append(src)
        seen.append(dst)

        entries.append((src, dst))

    seen_counts = collections.Counter(seen)
    print(len(entries))

    with open("create-links.sh", "w") as cf, open("remove-links.sh", "w") as rf, open("report-links.sh", "w") as rpf:
        for f in rf, cf, rpf:
            f.write("#!/usr/bin/env bash\n")
            f.write("set -eux\n")

        rpf.write("set +x\n")

        n_written = 0
        for src, dst in entries:
            if seen_counts[src] > 1 or seen_counts[dst] > 1 or not dst:
                print("skipping", src, dst, file=sys.stderr, flush=True)
                continue

            cf.write(f"test -L '{dst}' || ln -s '{src}' '{dst}'\n")
            rf.write(f"test -L '{dst}' && rm '{dst}'\n")
            rpf.write(f"test -L '{dst}' && echo '{dst} exists!'\n")
            n_written += 1

            if not n_written % args.rate:
                sleep_command = "sleep 1"
                cf.write(f"{sleep_command}\n")
                rf.write(f"{sleep_command}\n")