generated from IntensiveCoLearning/template
-
Notifications
You must be signed in to change notification settings - Fork 41
/
sync_status_readme.py
445 lines (379 loc) · 17.5 KB
/
sync_status_readme.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
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
import os
import subprocess
import re
import requests
from datetime import datetime, timedelta
import pytz
import logging
# Constants
START_DATE = datetime.fromisoformat(os.environ.get(
'START_DATE', '2024-06-24T00:00:00+00:00')).replace(tzinfo=pytz.UTC)
END_DATE = datetime.fromisoformat(os.environ.get(
'END_DATE', '2024-07-14T23:59:59+00:00')).replace(tzinfo=pytz.UTC)
DEFAULT_TIMEZONE = 'Asia/Shanghai'
FILE_SUFFIX = os.environ.get('FILE_SUFFIX', '_EICL1st.md')
README_FILE = 'README.md'
FIELD_NAME = os.environ.get('FIELD_NAME', 'EICL1st· Name')
Content_START_MARKER = "<!-- Content_START -->"
Content_END_MARKER = "<!-- Content_END -->"
TABLE_START_MARKER = "<!-- START_COMMIT_TABLE -->"
TABLE_END_MARKER = "<!-- END_COMMIT_TABLE -->"
GITHUB_REPOSITORY_OWNER = os.environ.get('GITHUB_REPOSITORY_OWNER')
GITHUB_REPOSITORY = os.environ.get('GITHUB_REPOSITORY')
# Configure logging
logging.basicConfig(level=logging.INFO,
format='%(asctime)s - %(levelname)s - %(message)s')
def print_env():
print(f"""
START_DATE: {START_DATE}
END_DATE: {END_DATE}
DEFAULT_TIMEZONE: {DEFAULT_TIMEZONE}
FILE_SUFFIX: {FILE_SUFFIX}
README_FILE: {README_FILE}
FIELD_NAME: {FIELD_NAME}
Content_START_MARKER: {Content_START_MARKER}
Content_END_MARKER: {Content_END_MARKER}
TABLE_START_MARKER: {TABLE_START_MARKER}
TABLE_END_MARKER: {TABLE_END_MARKER}
""")
def print_variables(*args, **kwargs):
def format_value(value):
if isinstance(value, str) and ('\n' in value or '\r' in value):
return f'"""\n{value}\n"""'
return repr(value)
variables = {}
# 处理位置参数
for arg in args:
if isinstance(arg, dict):
variables.update(arg)
else:
variables[arg] = eval(arg)
# 处理关键字参数
variables.update(kwargs)
# 打印变量
for name, value in variables.items():
print(f"{name}: {format_value(value)}")
def get_date_range():
return [START_DATE + timedelta(days=x) for x in range((END_DATE - START_DATE).days + 1)]
def get_user_timezone(file_content):
yaml_match = re.search(r'---\s*\ntimezone:\s*(\S+)\s*\n---', file_content)
if yaml_match:
try:
return pytz.timezone(yaml_match.group(1))
except pytz.exceptions.UnknownTimeZoneError:
logging.warning(
f"Unknown timezone: {yaml_match.group(1)}. Using default {DEFAULT_TIMEZONE}.")
return pytz.timezone(DEFAULT_TIMEZONE)
def extract_content_between_markers(file_content):
start_index = file_content.find(Content_START_MARKER)
end_index = file_content.find(Content_END_MARKER)
if start_index == -1 or end_index == -1:
logging.warning("Content_START_MARKER markers not found in the file")
return ""
return file_content[start_index + len(Content_START_MARKER):end_index].strip()
def find_date_in_content(content, local_date):
date_patterns = [
r'###\s*' + local_date.strftime("%Y.%m.%d"),
r'###\s*' + local_date.strftime("%Y.%m.%d").replace('.0', '.'),
r'###\s*' +
local_date.strftime("%m.%d").lstrip('0').replace('.0', '.'),
r'###\s*' + local_date.strftime("%Y/%m/%d"),
r'###\s*' +
local_date.strftime("%m/%d").lstrip('0').replace('/0', '/'),
r'###\s*' + local_date.strftime("%m.%d").zfill(5)
]
combined_pattern = '|'.join(date_patterns)
return re.search(combined_pattern, content)
def get_content_for_date(content, start_pos):
next_date_pattern = r'###\s*(\d{4}\.)?(\d{1,2}[\.\/]\d{1,2})'
next_date_match = re.search(next_date_pattern, content[start_pos:])
if next_date_match:
return content[start_pos:start_pos + next_date_match.start()]
return content[start_pos:]
def check_md_content(file_content, date, user_tz):
try:
content = extract_content_between_markers(file_content)
local_date = date.astimezone(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0)
current_date_match = find_date_in_content(content, local_date)
if not current_date_match:
logging.info(
f"No match found for date {local_date.strftime('%Y-%m-%d')}")
return False
date_content = get_content_for_date(content, current_date_match.end())
date_content = re.sub(r'\s', '', date_content)
logging.info(
f"Content length for {local_date.strftime('%Y-%m-%d')}: {len(date_content)}")
return len(date_content) > 10
except Exception as e:
logging.error(f"Error in check_md_content: {str(e)}")
return False
def get_user_study_status(nickname):
user_status = {}
file_name = f"{nickname}{FILE_SUFFIX}"
try:
with open(file_name, 'r', encoding='utf-8') as file:
file_content = file.read()
user_tz = get_user_timezone(file_content)
logging.info(
f"File content length for {nickname}: {len(file_content)} user_tz: {user_tz}")
current_date = datetime.now(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0) # - timedelta(days=1)
for date in get_date_range():
local_date = date.astimezone(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0)
if date.day == current_date.day:
user_status[date] = "✅" if check_md_content(
file_content, date, pytz.UTC) else " "
elif date > current_date:
user_status[date] = " "
else:
user_status[date] = "✅" if check_md_content(
file_content, date, pytz.UTC) else "⭕️"
logging.info(f"Successfully processed file for user: {nickname}")
except FileNotFoundError:
logging.error(f"Error: Could not find file {file_name}")
user_status = {date: "⭕️" for date in get_date_range()}
except Exception as e:
logging.error(
f"Unexpected error processing file for {nickname}: {str(e)}")
user_status = {date: "⭕️" for date in get_date_range()}
return user_status
def check_weekly_status(user_status, date, user_tz):
try:
local_date = date.astimezone(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0)
week_start = (local_date - timedelta(days=local_date.weekday()))
week_dates = [week_start + timedelta(days=x) for x in range(7)]
current_date = datetime.now(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0)
week_dates = [d for d in week_dates if d.astimezone(pytz.UTC).date() in [
date.date() for date in get_date_range()] and d <= min(local_date, current_date)]
missing_days = sum(1 for d in week_dates if user_status.get(datetime.combine(
d.astimezone(pytz.UTC).date(), datetime.min.time()).replace(tzinfo=pytz.UTC), "⭕️") == "⭕️")
if local_date == current_date and missing_days > 2:
return "❌"
elif local_date < current_date and missing_days > 2:
return "❌"
elif local_date > current_date:
return " "
else:
return user_status.get(datetime.combine(date.date(), datetime.min.time()).replace(tzinfo=pytz.UTC), "⭕️")
except Exception as e:
logging.error(f"Error in check_weekly_status: {str(e)}")
return "⭕️"
def get_all_user_files():
exclude_prefixes = ('template', 'readme')
return [f[:-len(FILE_SUFFIX)] for f in os.listdir('.')
if f.lower().endswith(FILE_SUFFIX.lower())
and not f.lower().startswith(exclude_prefixes)]
def update_readme(content):
try:
start_index = content.find(TABLE_START_MARKER)
end_index = content.find(TABLE_END_MARKER)
if start_index == -1 or end_index == -1:
logging.error(
"Error: Couldn't find the table markers in README.md")
return content
new_table = [
f'{TABLE_START_MARKER}\n',
f'| {FIELD_NAME} | ' +
' | '.join(date.strftime("%m.%d").lstrip('0')
for date in get_date_range()) + ' |\n',
'| ------------- | ' +
' | '.join(['----' for _ in get_date_range()]) + ' |\n'
]
existing_users = set()
table_rows = content[start_index +
len(TABLE_START_MARKER):end_index].strip().split('\n')[2:]
for row in table_rows:
match = re.match(r'\|\s*([^|]+)\s*\|', row)
if match:
display_name = match.group(1).strip()
if display_name: # 检查 display_name 是否为非空
existing_users.add(display_name)
new_table.append(generate_user_row(display_name))
else:
logging.warning(
f"Skipping empty display name in row: {row}")
else:
logging.warning(f"Skipping invalid row: {row}")
new_users = set(get_all_user_files()) - existing_users
for user in new_users:
if user.strip(): # 确保用户名不是空的或只包含空格
new_table.append(generate_user_row(user))
logging.info(f"Added new user: {user}")
else:
logging.warning(f"Skipping empty user: '{user}'")
new_table.append(f'{TABLE_END_MARKER}\n')
return content[:start_index] + ''.join(new_table) + content[end_index + len(TABLE_END_MARKER):]
except Exception as e:
logging.error(f"Error in update_readme: {str(e)}")
return content
def generate_user_row(user):
user_status = get_user_study_status(user)
with open(f"{user}{FILE_SUFFIX}", 'r', encoding='utf-8') as file:
file_content = file.read()
user_tz = get_user_timezone(file_content)
new_row = f"| {user} |"
is_eliminated = False
absent_count = 0
current_week = None
user_current_day = datetime.now(user_tz).replace(
hour=0, minute=0, second=0, microsecond=0)
for date in get_date_range():
# 获取用户时区和当地时间进行比较,如果用户打卡时间大于当地时间,则不显示- timedelta(days=1)
user_datetime = date.astimezone(pytz.UTC).replace(
hour=0, minute=0, second=0, microsecond=0)
if is_eliminated or (user_datetime > user_current_day and user_datetime.day > user_current_day.day):
new_row += " |"
else:
user_date = user_datetime
# 检查是否是新的一周
week = user_date.isocalendar()[1] # 获取ISO日历周数
if week != current_week:
current_week = week
absent_count = 0 # 重置缺勤计数
status = user_status.get(user_date, "")
if status == "⭕️":
absent_count += 1
if absent_count > 2:
is_eliminated = True
new_row += " ❌ |"
else:
new_row += " ⭕️ |"
else:
new_row += f" {status} |"
return new_row + '\n'
def get_repo_info():
if 'GITHUB_REPOSITORY' in os.environ:
# 在GitHub Actions环境中
full_repo = os.environ['GITHUB_REPOSITORY']
owner, repo = full_repo.split('/')
else:
# 在本地环境中
try:
remote_url = subprocess.check_output(
['git', 'config', '--get', 'remote.origin.url']).decode('utf-8').strip()
if remote_url.startswith('https://github.com/'):
owner, repo = remote_url.split('/')[-2:]
elif remote_url.startswith('git@github.com:'):
owner, repo = remote_url.split(':')[-1].split('/')
else:
raise ValueError("Unsupported remote URL format")
repo = re.sub(r'\.git$', '', repo)
except subprocess.CalledProcessError:
logging.error(
"Failed to get repository information from git config")
return None, None
return owner, repo
def get_fork_count():
owner, repo = get_repo_info()
if not owner or not repo:
logging.error("Failed to get repository information")
return None
api_url = f"https://api.github.com/repos/{owner}/{repo}"
try:
response = requests.get(api_url)
response.raise_for_status()
repo_data = response.json()
return repo_data['forks_count']
except requests.RequestException as e:
logging.error(f"Error fetching fork count: {e}")
return None
def calculate_statistics(content):
start_index = content.find(TABLE_START_MARKER)
end_index = content.find(TABLE_END_MARKER)
if start_index == -1 or end_index == -1:
logging.error("Error: Couldn't find the table markers in README.md")
return None
table_content = content[start_index +
len(TABLE_START_MARKER):end_index].strip()
rows = table_content.split('\n')[2:] # Skip header and separator rows
total_participants = len(rows)
eliminated_participants = 0
completed_participants = 0
perfect_attendance_users = []
for row in rows:
user_name = row.split('|')[1].strip()
# Exclude first and last empty elements
statuses = [status.strip() for status in row.split('|')[2:-1]]
if '❌' in statuses:
eliminated_participants += 1
elif all(status == '✅' for status in statuses):
completed_participants += 1
perfect_attendance_users.append(user_name)
elif all(status in ['✅', '⭕️', ' '] for status in statuses):
completed_participants += 1
elimination_rate = (eliminated_participants /
total_participants) * 100 if total_participants > 0 else 0
fork_count = get_fork_count()
return {
'total_participants': total_participants,
'completed_participants': completed_participants,
'eliminated_participants': eliminated_participants,
'elimination_rate': elimination_rate,
'fork_count': fork_count,
'perfect_attendance_users': perfect_attendance_users
}
def main():
try:
print_variables(
'START_DATE', 'END_DATE', 'DEFAULT_TIMEZONE',
GITHUB_REPOSITORY_OWNER=GITHUB_REPOSITORY,
GITHUB_REPOSITORY=GITHUB_REPOSITORY,
FILE_SUFFIX=FILE_SUFFIX,
README_FILE=README_FILE,
FIELD_NAME=FIELD_NAME,
Content_START_MARKER=Content_START_MARKER,
Content_END_MARKER=Content_END_MARKER,
TABLE_START_MARKER=TABLE_START_MARKER,
TABLE_END_MARKER=TABLE_END_MARKER
)
with open(README_FILE, 'r', encoding='utf-8') as file:
content = file.read()
new_content = update_readme(content)
current_date = datetime.now(pytz.UTC)
if current_date > END_DATE:
stats = calculate_statistics(new_content)
if stats:
stats_content = f"\n\n## 统计数据\n\n"
stats_content += f"- 总参与人数: {stats['total_participants']}\n"
stats_content += f"- 完成人数: {stats['completed_participants']}\n"
stats_content += f"- 全勤用户: {', '.join(stats['perfect_attendance_users'])}\n"
stats_content += f"- 淘汰人数: {stats['eliminated_participants']}\n"
stats_content += f"- 淘汰率: {stats['elimination_rate']:.2f}%\n"
stats_content += f"- Fork人数: {stats['fork_count']}\n"
# 将统计数据添加到文件末尾
# 在<!-- END_COMMIT_TABLE -->标记后插入统计数据
stats_start = new_content.find(
"<!-- STATISTICALDATA_START -->")
stats_end = new_content.find("<!-- STATISTICALDATA_END -->")
if stats_start != -1 and stats_end != -1:
# Replace existing statistical data
new_content = new_content[:stats_start] + "<!-- STATISTICALDATA_START -->\n" + stats_content + \
"<!-- STATISTICALDATA_END -->" + \
new_content[stats_end +
len("<!-- STATISTICALDATA_END -->"):]
else:
# Add new statistical data after <!-- END_COMMIT_TABLE -->
end_table_marker = "<!-- END_COMMIT_TABLE -->"
end_table_index = new_content.find(end_table_marker)
if end_table_index != -1:
insert_position = end_table_index + \
len(end_table_marker)
new_content = new_content[:insert_position] + "\n\n<!-- STATISTICALDATA_START -->\n" + \
stats_content + "<!-- STATISTICALDATA_END -->" + \
new_content[insert_position:]
else:
logging.warning(
"<!-- END_COMMIT_TABLE --> marker not found. Appending stats to the end.")
new_content += "\n\n<!-- STATISTICALDATA_START -->\n" + \
stats_content + "<!-- STATISTICALDATA_END -->"
with open(README_FILE, 'w', encoding='utf-8') as file:
file.write(new_content)
logging.info("README.md has been successfully updated.")
except Exception as e:
logging.error(f"An error occurred in main function: {str(e)}")
if __name__ == "__main__":
main()