/util.py

"""This module holds utility functions used by some modules and the bot itself."""
import config
import requests_cache
import logging
import re
from bs4 import BeautifulSoup
from urllib.request import Request, urlopen
from urllib.error import HTTPError, URLError
import stopit
import shlex

logger = logging.getLogger(__name__)

idregex = re.compile(r'\@([A-Z0-9]{9})', re.UNICODE)
channelregex = re.compile(r'\#([A-Z0-9]{9})', re.UNICODE)
urlregex = re.compile(r'\<(.*?)\>')


def getUserdataForId(userId):
    """
    Get the json response from the slack api for the specified user id.

    API endpoint: https://api.slack.com/methods/users.info

    Arguments: userId, slack user id to fetch. Required.

    Returns: the json response for the id
    """
    logging.info("Fetching user data for id %s", userId)
    s = requests_cache.CachedSession(cache_name='namelookup_cache', backend='sqlite', expire_after=3600)
    payload = {"token": config.SLACK_TOKEN, "user": userId}
    req = s.get("https://slack.com/api/users.info", params=payload)
    logging.debug("User data for id %s was cached=%s", userId, req.from_cache)
    return req.json()


def __replaceUserId(userId):
    """
    Get the username for the specified id.

    Will return the raw id if username could not be found.
    """
    logging.debug("found id %s", userId.group(1))
    data = getUserdataForId(userId.group(1))
    if not data["ok"]:
        logging.warning("%s for id %s", data["error"], userId.group(1))
        return userId.group(0)
    else:
        return "@" + data["user"]["name"]


def getChanneldataForId(channelId):
    """
    Get the json response from the slack api for the specified channel id.

    API endpoint: https://api.slack.com/methods/channels.info

    Arguments: channelId, slack channel id to fetch. Required.

    Returns: the json response for the id
    """
    logging.info("Fetching channel data for id %s", channelId)
    s = requests_cache.CachedSession(cache_name='namelookup_cache', backend='sqlite', expire_after=3600)
    payload = {"token": config.SLACK_TOKEN, "channel": channelId}
    req = s.get("https://slack.com/api/channels.info", params=payload)
    logging.debug("Channel data for id %s was cached=%s", channelId, req.from_cache)
    return req.json()


def __replaceChannelId(channelId):
    """
    Gets the channel name for the specified id.

    Will return the raw id if channel could not be found.
    """
    logging.debug("found id %s", channelId.group(1))
    data = getChanneldataForId(channelId.group(1))
    if not data["ok"]:
        logging.warning("%s for id %s", data["error"], channelId.group(1))
        return channelId.group(0)
    else:
        return "#" + data["channel"]["name"]


def replaceIdsInMessage(message):
    """
    Replaces occurences of slack id's in messages with their human readable counterparts.
    """
    message = re.sub(channelregex, __replaceChannelId, message)
    return re.sub(idregex, __replaceUserId, message)


@stopit.threading_timeoutable(default=None)
def getTitle(url):
    """
    Fetch the <title> tag of the specified url.

    Returns a string if <title> could be found, else returns None.
    """
    try:
        headers = {'User-Agent': 'SlackBot %s V1.0' % config.BOT_NAME}
        req = Request(url, None, headers)
        response = urlopen(req)
        html = response.read()
    except HTTPError as e:
        print('The server couldn\'t fulfill the request.')
        print(('Error code: ', e.code))
    except URLError as e:
        print('We failed to reach a server.')
        print(('Reason: ', e.reason))
    except Exception as e:
        print(e)
    else:
        soup = BeautifulSoup(html, "lxml")
        if soup.title is not None:
            return soup.title.text.strip()

    return None


def dict_of_key_value_pairs(arg):
    """ parse KEY=val,KEY2=val2 into {'KEY':'val', 'KEY2':'val2'}
        Quotes can be used to allow commas in the value
        This piece of code has been taken from:
        https://github.com/Supervisor/supervisor/blob/e4f3f0cad1697fdb76632dfe8a3935ae4540fbae/supervisor/datatypes.py#L80
    """
    lexer = shlex.shlex(arg)
    lexer.wordchars += '/.+-():'

    tokens = list(lexer)
    tokens_len = len(tokens)

    D = {}
    i = 0
    while i < tokens_len:
        k_eq_v = tokens[i:i+3]
        if len(k_eq_v) != 3 or k_eq_v[1] != '=':
            raise ValueError("Unexpected end of key/value pairs")
        D[k_eq_v[0]] = k_eq_v[2].strip('\'"')
        i += 4
    return D


def replaceUrlTags(message):
    """Strip slacks url formatting out of the message"""
    matches = re.findall(urlregex, message)
    if matches:
        for match in matches:
            message = urlregex.sub(match.split("|")[0], message, count=1)
    return message