components/registration/utils.py

import json
import logging
import re

import pandas as pd

from about import REGISTRATION_REPO
from components.registration.config import PARTICIPANTS_FILE_NAME, PARTICIPANT_COLUMNS
from components.utils import load_data_from_dataset

logger = logging.getLogger(__name__)


def validate_email_domain(email: str) -> bool:
    """Return True if email belongs to an academic or registered company domain."""
    academic_tlds = [
        ".edu",
        ".ac.uk",
        ".ac.",
        "university",
        "univ.",
        "institute",
        ".gov",
    ]
    return any(token in email.lower() for token in academic_tlds)


def validate_email_format(email: str) -> bool:
    """
    Validate the email format.

    :param email:    The email address to validate.
    :return:         True if the email is valid, False otherwise.
    """
    return bool(re.match(r"^[^@\s]+@[^@\s]+\.[^@\s]+$", email))


def is_institutional_email(email: str) -> bool:
    """Check if an email address is an institutional email."""
    patterns = [".edu", ".ac.", ".gov", "university", "univ.", "institute", "research"]
    return any(p in email.lower() for p in patterns)


def get_user_teams(username: str) -> pd.DataFrame:
    """
    Return the teams associated with a given username.

    :param username:The unique hf username to query.
    :return:        The teams associated with a given username as a DataFrame.
    """
    participants_df = load_data_from_dataset(
        REGISTRATION_REPO, PARTICIPANTS_FILE_NAME, PARTICIPANT_COLUMNS
    )
    if participants_df.empty:
        return pd.DataFrame(columns=["team_name", "role"])

    existing_participant = participants_df[participants_df["username"] == username]

    if existing_participant.empty:
        return pd.DataFrame(columns=["team_name", "role"])

    return existing_participant[["team_name", "is_leader"]]