def get_espn_game(date, home_team, away_team):
"""
Gets the ESPN pbp feed
Ex: http://www.espn.com/nhl/gamecast/data/masterFeed?lang=en&isAll=true&gameId=400885300
:param date: date of the game
:param home_team: home team
:param away_team: away team
:return: raw xml
"""
game_id = get_espn_game_id(date, home_team.upper(), away_team.upper())
file_info = {
"url":
'http://www.espn.com/nhl/gamecast/data/masterFeed?lang=en&isAll=true&gameId={}'
.format(game_id),
"name":
game_id,
"type":
"espn_pbp",
"season":
shared.get_season(date),
}
response = shared.get_file(file_info)
if response is None:
raise Exception
return response
def get_espn_date(date):
"""
Get the page that contains all the games for that day
:param date: YYYY-MM-DD
:return: response
"""
page_info = {
"url":
'http://www.espn.com/nhl/scoreboard?date={}'.format(
date.replace('-', '')),
"name":
date,
"type":
"espn_scoreboard",
"season":
shared.get_season(date),
}
response = shared.get_file(page_info)
# If can't get or not there throw an exception
if not response:
raise Exception
else:
return response
def scrape_shifts(game_id, players, date):
"""
Scrape the Shift charts (or TOI tables)
:param game_id: json game id
:param players: dict of players with numbers and id's
:param date: date of game
:return: DataFrame with info or None if it fails
"""
shifts_df = None
# Control for fact that shift json is only available from 2010 onwards
if shared.get_season(date) >= 2010:
shifts_df = json_shifts.scrape_game(game_id)
if shifts_df is None:
shifts_df = html_shifts.scrape_game(game_id, players)
if shifts_df is None:
shared.print_warning("Unable to scrape shifts for game" + game_id)
broken_shifts_games.extend([[game_id, date]])
return None # Both failed so just return nothing
shifts_df['Date'] = date
return shifts_df
def get_espn_game(date, home_team, away_team, game_id=None):
"""
Gets the ESPN pbp feed
Ex: http://www.espn.com/nhl/gamecast/data/masterFeed?lang=en&isAll=true&gameId=400885300
:param date: date of the game
:param home_team: home team
:param away_team: away team
:param game_id: Game id of we already have it - for live scraping. None if not there
:return: raw xml
"""
# Get if not provided
if not game_id:
game_id = get_espn_game_id(date, home_team.upper(), away_team.upper())
file_info = {
"url": 'http://www.espn.com/nhl/gamecast/data/masterFeed?lang=en&isAll=true&gameId={}'.format(game_id),
"name": game_id,
"type": "espn_pbp",
"season": shared.get_season(date),
}
response = shared.get_file(file_info)
if response is None:
raise Exception
return response
def get_schedule(date_from, date_to):
"""
Scrapes games in date range
Ex: https://statsapi.web.nhl.com/api/v1/schedule?startDate=2010-10-03&endDate=2011-06-20
:param date_from: scrape from this date
:param date_to: scrape until this date
:return: raw json of schedule of date range
"""
page_info = {
"url": 'https://statsapi.web.nhl.com/api/v1/schedule?startDate={a}&endDate={b}'.format(a=date_from, b=date_to),
"name": date_from + "_" + date_to,
"type": "json_schedule",
"season": shared.get_season(date_from),
}
return json.loads(shared.get_file(page_info))
def parse_event(event, score, teams, date, game_id, players):
"""
Parses a single event when the info is in a json format
:param event: json of event
:param score: Current score of the game
:param teams: Teams dict (id -> name)
:param date: date of the game
:param game_id: game id for game
:param players: Dict of player ids to player names
:return: dictionary with the info
"""
play = dict()
# Basic shit
play['play_index'] = event['play_index']
play['date'] = date
play['game_id'] = game_id
play['season'] = shared.get_season(date)
play['period'] = event['time_interval']
play['seconds_elapsed'] = shared.convert_to_seconds(event['clock_time_string']) if event['clock_time_string'] else None
play['home_score'], play['away_score'] = score['home'], score['away']
# If shootout go with 'play_by_play_string' field -> more descriptive
play['event'] = event['play_type'] if event['play_type'] != "Shootout" else event['play_by_play_string'].strip()
# Teams
play['home_team'], play['away_team'] = teams['home']['name'], teams['away']['name']
if event['play_summary']['off_team_id'] == teams['home']['id']:
play['ev_team'] = teams['home']['name']
else:
play['ev_team'] = teams['away']['name']
# Player Id
play['p1_id'] = event.get('primary_player_id')
play['away_goalie_id'] = event['play_actions'][0].get('away_team_goalie')
play['home_goalie_id'] = event['play_actions'][0].get('home_team_goalie')
play['away_goalie'] = players.get(int(play['away_goalie_id']) if play['away_goalie_id'] not in ['', None] else 0)
play['home_goalie'] = players.get(int(play['home_goalie_id']) if play['home_goalie_id'] not in ['', None] else 0)
# Event specific stuff
if event['play_type'] == 'Faceoff':
play['p2_id'] = event['play_summary'].get("loser_id")
elif event['play_type'] == 'Penalty':
# TODO: Format better?
play['details'] = ",".join([str(event['play_summary'].get("infraction_type", " ")),
str(event['play_summary'].get("penalty_type", " ")),
str(event['play_summary'].get("penalty_minutes", " "))])
elif event['play_type'] == "Goal":
get_goal_players(play, event, players)
play['p2_id'] = event['play_summary'].get("assist_1_id")
play['p3_id'] = event['play_summary'].get("assist_2_id")
# Update Score
if event['play_summary']['off_team_id'] == teams['home']['id']:
score['home'] += 1
else:
score['away'] += 1
# Player Id's --> Player Names
for num in range(1, 4):
player_id = play.get('p{num}_id'.format(num=num), 0)
# Control for None
player_id = player_id if player_id else 0
play['p{num}_name'.format(num=num)] = players.get(int(player_id))
# Coords
play['xC'] = event['play_summary'].get('x_coord')
play['yC'] = event['play_summary'].get('y_coord')
return play
def test_get_season():
""" Tests that this function returns the correct season for a given date"""
assert shared.get_season("2017-10-01") == 2017
assert shared.get_season("2016-06-01") == 2015