TopDownHockey-Scraper 4.2.1__tar.gz → 4.3.1__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.


This version of TopDownHockey-Scraper might be problematic. Click here for more details.

@@ -1,6 +1,6 @@
1
- Metadata-Version: 2.2
1
+ Metadata-Version: 2.4
2
2
  Name: TopDownHockey_Scraper
3
- Version: 4.2.1
3
+ Version: 4.3.1
4
4
  Summary: The TopDownHockey Scraper
5
5
  Home-page: https://github.com/TopDownHockey/TopDownHockey_Scraper
6
6
  Author: Patrick Bacon
@@ -29,6 +29,7 @@ Dynamic: description
29
29
  Dynamic: description-content-type
30
30
  Dynamic: home-page
31
31
  Dynamic: license
32
+ Dynamic: license-file
32
33
  Dynamic: project-url
33
34
  Dynamic: requires-dist
34
35
  Dynamic: requires-python
@@ -1,6 +1,6 @@
1
1
  [metadata]
2
2
  name = TopDownHockey_Scraper
3
- version = 4.2.1
3
+ version = 4.3
4
4
  author = Patrick Bacon
5
5
  author_email = patrick.s.bacon@gmail.com
6
6
  description = A package built for scraping hockey data from EliteProspects, the NHL's HTML/API reports, and ESPN's XML reports.
@@ -9,7 +9,7 @@ with open("README.md", "r", encoding="utf-8") as fh:
9
9
 
10
10
  setup(
11
11
  name="TopDownHockey_Scraper", # Replace with your own username
12
- version="4.2.1",
12
+ version="4.3.1",
13
13
  author="Patrick Bacon",
14
14
  author_email="patrick.s.bacon@gmail.com",
15
15
  description="The TopDownHockey Scraper",
@@ -503,10 +503,13 @@ def scrape_html_roster(season, game_id):
503
503
  (np.where(roster_df['Name']== "JAKE LUCCHINI" , "JACOB LUCCHINI",
504
504
  (np.where(roster_df['Name']== "EMIL LILLEBERG" , "EMIL MARTINSEN LILLEBERG",
505
505
  (np.where(roster_df['Name']== "CAMERON ATKINSON" , "CAM ATKINSON",
506
- roster_df['Name']))))))))))))))))))))))))))))))))
506
+ (np.where(roster_df['Name']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY",
507
+ roster_df['Name']))))))))))))))))))))))))))))))))))
507
508
 
508
509
  roster_df['Name'] = roster_df['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8').str.upper()
509
510
 
511
+ roster_df['Name'] = np.where(roster_df['Name']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY", roster_df['Name']) # Need to do this after normalization, only then he becomes Slafkovska?
512
+
510
513
  return roster_df
511
514
 
512
515
  def scrape_html_shifts(season, game_id):
@@ -1224,9 +1227,12 @@ def scrape_html_shifts(season, game_id):
1224
1227
  (np.where(all_shifts['name']== "JAKE LUCCHINI" , "JACOB LUCCHINI",
1225
1228
  (np.where(all_shifts['name']== "EMIL LILLEBERG" , "EMIL MARTINSEN LILLEBERG",
1226
1229
  (np.where(all_shifts['name']== "CAMERON ATKINSON" , "CAM ATKINSON",
1227
- all_shifts['name']))))))))))))))))))))))))))))))))
1230
+ (np.where(all_shifts['name']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY",
1231
+ all_shifts['name']))))))))))))))))))))))))))))))))))
1228
1232
 
1229
1233
  all_shifts['name'] = all_shifts['name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8').str.upper()
1234
+
1235
+ all_shifts['name'] = np.where(all_shifts['name']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY", all_shifts['name']) # Need to do this after normalization, only then he becomes Slafkovska?
1230
1236
 
1231
1237
 
1232
1238
  all_shifts = all_shifts.assign(end_time = np.where(pd.to_datetime(all_shifts.start_time).dt.time > pd.to_datetime(all_shifts.end_time).dt.time, '20:00', all_shifts.end_time),
@@ -1545,13 +1551,16 @@ def scrape_api_events(game_id, drop_description = True, shift_to_espn = False):
1545
1551
  (np.where(api_events['ep1_name']== "JAKE LUCCHINI" , "JACOB LUCCHINI",
1546
1552
  (np.where(api_events['ep1_name']== "EMIL LILLEBERG" , "EMIL MARTINSEN LILLEBERG",
1547
1553
  (np.where(api_events['ep1_name']== "CAMERON ATKINSON" , "CAM ATKINSON",
1548
- api_events['ep1_name']))))))))))))))))))))))))))))))))
1554
+ (np.where(api_events['ep1_name']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY",
1555
+ api_events['ep1_name']))))))))))))))))))))))))))))))))))
1549
1556
 
1550
1557
 
1551
1558
  # 21-22 CHANGES
1552
1559
 
1553
1560
  api_events['ep1_name'] = api_events['ep1_name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8').str.upper()
1554
1561
 
1562
+ api_events['ep1_name'] = np.where(api_events['ep1_name'] == "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY", api_events['ep1_name']) # Need to do this after normalization, only then he becomes Slafkovska?
1563
+
1555
1564
  api_events = api_events.assign(ep1_name = np.where(api_events.ep1_name=='ALEX BARRÉ-BOULET', 'ALEX BARRE_BOULET', api_events.ep1_name))
1556
1565
 
1557
1566
  if drop_description == True:
@@ -1930,7 +1939,8 @@ def scrape_espn_events(espn_game_id, drop_description = True):
1930
1939
  (np.where(espn_events['event_player_1']== "JAKE LUCCHINI" , "JACOB LUCCHINI",
1931
1940
  (np.where(espn_events['event_player_1']== "EMIL LILLEBERG" , "EMIL MARTINSEN LILLEBERG",
1932
1941
  (np.where(espn_events['event_player_1']== "CAMERON ATKINSON" , "CAM ATKINSON",
1933
- espn_events['event_player_1']))))))))))))))))))))))))))))))))
1942
+ (np.where(espn_events['event_player_1']== "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY",
1943
+ espn_events['event_player_1']))))))))))))))))))))))))))))))))))
1934
1944
 
1935
1945
 
1936
1946
  espn_events = espn_events.assign(version =
@@ -1963,6 +1973,8 @@ def scrape_espn_events(espn_game_id, drop_description = True):
1963
1973
  espn_events['event_player_1'] = espn_events['event_player_1'].str.strip()
1964
1974
 
1965
1975
  espn_events['event_player_1'] = espn_events['event_player_1'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8').str.upper()
1976
+
1977
+ espn_events['event_player_1'] = np.where(espn_events['event_player_1'] == "JURAJ SLAFKOVSKA" , "JURAJ SLAFKOVSKY", espn_events['event_player_1'])
1966
1978
 
1967
1979
  #espn_events = espn_events.assign(event_player_1 = np.where(
1968
1980
  #espn_events.event_player_1=='ALEX BURROWS', 'ALEXANDRE BURROWS', espn_events.event_player_1))
@@ -1,6 +1,6 @@
1
- Metadata-Version: 2.2
1
+ Metadata-Version: 2.4
2
2
  Name: TopDownHockey_Scraper
3
- Version: 4.2.1
3
+ Version: 4.3.1
4
4
  Summary: The TopDownHockey Scraper
5
5
  Home-page: https://github.com/TopDownHockey/TopDownHockey_Scraper
6
6
  Author: Patrick Bacon
@@ -29,6 +29,7 @@ Dynamic: description
29
29
  Dynamic: description-content-type
30
30
  Dynamic: home-page
31
31
  Dynamic: license
32
+ Dynamic: license-file
32
33
  Dynamic: project-url
33
34
  Dynamic: requires-dist
34
35
  Dynamic: requires-python