-
Notifications
You must be signed in to change notification settings - Fork 8
/
espn.py
234 lines (200 loc) · 7.93 KB
/
espn.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
#!/usr/bin/env python
"""
This is my simple library to scrape NBA and NCB play-by-play information from
ESPN. I needed a small project to test out BeautifulSoup Alpha 4, so I felt
this was a good opportunity to refine some earlier code I had written.
In order to use this, you will need to download bs4 and also lxml (bs4 is
modular and can use both lxml and html5lib as back-end parsers).
The output is in dictionary format and contains the following information for
each individual play: quarter, quarter_time, overall_time, home_score,
away_score, home_play, away_play, and official_play (for timeouts, starts of
quarters, etc). I've found the majority of games normally have around 400 to
460 plays.
"""
import urllib2
import re
import datetime
from urlparse import urlparse
from bs4 import BeautifulSoup as bs
def daterange(start, end):
"""Generator for days between two specific days."""
for n in range((end - start).days):
yield start + datetime.timedelta(n)
def _format_scoreboard_url(day, league='nba'):
"""Format ESPN scoreboard link to scrape individual box scores from."""
league = league.lower()
link = [league + '/scoreboard?date=']
if isinstance(day, datetime.date):
link.append(day.strftime('%Y%m%d'))
else:
link.append(day)
if league == 'ncb':
link.append('&confId=50')
scoreboard_link = ''.join(['http://scores.espn.go.com/', ''.join(link)])
return scoreboard_link
def scrape_links(espn_scoreboard):
"""Scrape ESPN's scoreboard for Play-By-Play links."""
url = urllib2.urlopen(espn_scoreboard)
print url.geturl()
soup = bs(url.read(), ['fast', 'lxml'])
div = soup.find('div', {'class': 'span-4'})
links = (a['href'] for a in div.findAll('a') if re.match('Play.*',
a.contents[0]))
queries = [urlparse(link).query for link in links]
return queries
def adjust_game(plays, league='nba'):
"""
Takes plays from parse_plays (generator of lists) and the league which
it is parsing (used for correct quarter/halve time). It returns a list
of plays in dictionary format -- which is more convenient for lookups.
The dictionary contains the following information: quarter, quarter_time,
overall_time, home_score, away_score, home_play, away_play, and
official_play (for timeouts, starts of quarters, etc).
"""
# TODO: Maybe 'period' instead of 'quarter'? NCB uses 'halves'.
game = []
quarter = 1
end_of_quarter = False
for play in plays:
new_play = _play_as_dict(play)
time = play[0]
time_dict, quarter, end_of_quarter = _adjust_time(time,
quarter, end_of_quarter, league)
new_play.update(time_dict)
try:
scores = play[2]
except IndexError:
if len(game) > 0:
# Official Play without score (new quarter, etc.)
last_play = game[-1]
new_play['away_score'] = last_play['away_score']
new_play['home_score'] = last_play['home_score']
else:
# Start of game
new_play['away_score'] = 0
new_play['home_score'] = 0
else:
away_score, home_score = scores.split('-')
new_play['away_score'] = int(away_score)
new_play['home_score'] = int(home_score)
game.append(new_play)
return game
def _adjust_time(time, quarter, end_of_quarter, league):
"""
Takes the time logic out of adjust_game.
Returns a dict, quarter, and end_of_quarter.
"""
new_time = re.split(':', time)
minutes = int(new_time[0])
seconds = int(new_time[1])
if minutes is 0 and not end_of_quarter:
end_of_quarter = True
elif end_of_quarter and minutes > 1:
quarter += 1
end_of_quarter = False
overall_time = _calc_overall_time(seconds, minutes, quarter, league)
time_dict = {}
time_dict['overall_time'] = overall_time
time_dict['quarter_time'] = time
time_dict['quarter'] = quarter
return time_dict, quarter, end_of_quarter
def _league_time(league):
"""
Return league specific game info -- number of quarters, regulation time
limit, regular quarter length.
"""
if league is 'nba':
num_quarters = 4
regulation_time = 48
regular_quarter = 12
else:
num_quarters = 2
regulation_time = 40
regular_quarter = 20
return num_quarters, regulation_time, regular_quarter
def _calc_overall_time(seconds, minutes, quarter, league):
"""
Calculate the overall time that's elapsed for a given game. I'm not a fan
of four arguments, but it's necessary unfortunately.
"""
num_quarters, regulation_time, regular_quarter = _league_time(league)
if quarter > num_quarters:
# We're in overtime.
quarter_length = 5
overtimes = quarter - num_quarters
previous_time = datetime.timedelta(minutes=(regulation_time +
5 * (overtimes - 1)))
else:
quarter_length = regular_quarter
previous_time = datetime.timedelta(minutes=(quarter_length *
(quarter - 1)))
mins = datetime.timedelta(minutes=quarter_length) -\
datetime.timedelta(minutes=minutes, seconds=seconds)
overall_time = str(mins + previous_time)
return overall_time
def _play_as_dict(play):
"""
Give it a play in list/tuple format, get back a dict containing
official_play, home_play, and away_play data.
Really only for internal use with adjust_game.
"""
# TODO: Play can be ' ' or u'\xa0', so I put len < 10.
# Should probably change to something more explicit in the future.
new_play = {}
if len(play) is 2:
new_play['official_play'] = play[1]
new_play['home_play'] = None
new_play['away_play'] = None
else:
new_play['official_play'] = None
away_play = play[1]
home_play = play[3]
if len(away_play) < 10:
new_play['away_play'] = None
new_play['home_play'] = home_play
elif len(home_play) < 10:
new_play['away_play'] = away_play
new_play['home_play'] = None
return new_play
def parse_plays(game_id, league='nba'):
"""Parse a game's Play-By-Play page on ESPN."""
league = league.lower()
espn = 'http://scores.espn.go.com/' + league + '/playbyplay?' +\
game_id + '&period=0'
url = urllib2.urlopen(espn)
print url.geturl()
soup = bs(url.read(), ['fast', 'lxml'])
table = soup.find('table', {'class': 'mod-data'})
thead = [thead.extract() for thead in table.findAll('thead')]
rows = (list(tr(text=True)) for tr in table.findAll('tr'))
game = adjust_game(rows, league)
teams = thead[0].findChildren('th', {'width':'40%'})
away_team, home_team = [team.string.title() for team in teams]
print len(game), away_team, home_team
return away_team, home_team, game
def get_games(day, league='nba', iterable=False):
"""
Get the games and play-by-play data from ESPN for a date.
The date can be in `datetime.date` format or be a YYYYMMDD string.
By default it only looks for NBA games on a date. You can modify
this by passing it a league='ncb' argument to scrape the NCAA men's
basketball games that have Play-By-Play data (NOTE: not all will).
You can pass it an optional `iterable=True` argument in order to receive
back a generator instead of the default list.
"""
espn_scoreboard = _format_scoreboard_url(day, league=league)
all_games = scrape_links(espn_scoreboard)
if not iterable:
games = [parse_plays(game, league=league) for game in all_games]
else:
games = (parse_plays(game, league=league) for game in all_games)
return games
def main():
yesterday = datetime.date.today() - datetime.timedelta(1)
for game in get_games(yesterday, iterable=True):
print game
if __name__ == '__main__':
import time
start = time.time()
main()
print time.time() - start, 'seconds'