kevin-toolbox-dev 1.4.5__py3-none-any.whl → 1.4.7__py3-none-any.whl
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.
- kevin_toolbox/__init__.py +2 -2
- kevin_toolbox/computer_science/algorithm/cache_manager/cache_manager.py +6 -0
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/__init__.py +1 -0
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/multi_process_execute.py +109 -0
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/multi_thread_execute.py +50 -29
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/utils/__init__.py +15 -0
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/utils/wrapper_with_timeout_1.py +69 -0
- kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/utils/wrapper_with_timeout_2.py +76 -0
- kevin_toolbox/computer_science/data_structure/executor.py +2 -2
- kevin_toolbox/data_flow/file/excel/__init__.py +1 -0
- kevin_toolbox/data_flow/file/excel/write_excel_with_matrix.py +105 -0
- kevin_toolbox/data_flow/file/json_/read_json.py +1 -0
- kevin_toolbox/data_flow/file/json_/write_json.py +36 -3
- kevin_toolbox/env_info/__init__.py +2 -1
- kevin_toolbox/env_info/check_validity_and_uninstall.py +41 -21
- kevin_toolbox/env_info/check_version_and_update.py +70 -49
- kevin_toolbox/env_info/test/test_check_.py +52 -0
- kevin_toolbox/nested_dict_list/serializer/backends/_json_.py +2 -2
- kevin_toolbox/patches/for_matplotlib/common_charts/plot_lines.py +5 -1
- kevin_toolbox/patches/for_os/find_files_in_dir.py +22 -17
- kevin_toolbox/patches/for_os/organize/__init__.py +1 -0
- kevin_toolbox/patches/for_os/organize/group_files_by_timestamp.py +90 -0
- kevin_toolbox_dev-1.4.7.dist-info/METADATA +69 -0
- {kevin_toolbox_dev-1.4.5.dist-info → kevin_toolbox_dev-1.4.7.dist-info}/RECORD +26 -17
- kevin_toolbox_dev-1.4.5.dist-info/METADATA +0 -73
- {kevin_toolbox_dev-1.4.5.dist-info → kevin_toolbox_dev-1.4.7.dist-info}/WHEEL +0 -0
- {kevin_toolbox_dev-1.4.5.dist-info → kevin_toolbox_dev-1.4.7.dist-info}/top_level.txt +0 -0
kevin_toolbox/__init__.py
CHANGED
@@ -1,4 +1,4 @@
|
|
1
|
-
__version__ = "1.4.
|
1
|
+
__version__ = "1.4.7"
|
2
2
|
|
3
3
|
|
4
4
|
import os
|
@@ -12,5 +12,5 @@ os.system(
|
|
12
12
|
os.system(
|
13
13
|
f'python {os.path.split(__file__)[0]}/env_info/check_validity_and_uninstall.py '
|
14
14
|
f'--package_name kevin-toolbox-dev '
|
15
|
-
f'--expiration_timestamp
|
15
|
+
f'--expiration_timestamp 1755525740 --verbose 0'
|
16
16
|
)
|
@@ -44,6 +44,12 @@ class Cache_Manager:
|
|
44
44
|
当设置为 float 时表示占 upper_bound 的比例
|
45
45
|
默认为 0.5
|
46
46
|
strategy: <str/dict/Strategy_Base> 管理策略
|
47
|
+
目前支持以下策略:
|
48
|
+
- ":by_initial_time:FIFO" 删除最后一次访问时间最久远的部分
|
49
|
+
- ":by_counts:LFU" 删除访问频率最低的部分
|
50
|
+
- ":by_last_time:LRU" 删除最后一次访问时间最久远的部分
|
51
|
+
- ":by_survival_time:LST" 删除访问频率最低的部分
|
52
|
+
默认使用 LRU 策略
|
47
53
|
cache: <str/dict/Cache_Base> 缓存种类
|
48
54
|
"""
|
49
55
|
# 默认参数
|
@@ -0,0 +1,109 @@
|
|
1
|
+
import pickle
|
2
|
+
import concurrent.futures
|
3
|
+
from multiprocessing import Manager
|
4
|
+
from kevin_toolbox.computer_science.data_structure import Executor
|
5
|
+
from kevin_toolbox.computer_science.algorithm.parallel_and_concurrent.utils import wrapper_for_mp as wrapper
|
6
|
+
from kevin_toolbox.computer_science.algorithm.parallel_and_concurrent.utils import DEFAULT_PROCESS_NUMS
|
7
|
+
|
8
|
+
|
9
|
+
def multi_process_execute(executors, worker_nums=DEFAULT_PROCESS_NUMS, b_display_progress=True, timeout=None,
|
10
|
+
_hook_for_debug=None):
|
11
|
+
"""
|
12
|
+
多进程执行
|
13
|
+
|
14
|
+
参数:
|
15
|
+
executors: <list/generator/iterator of Executor> 执行器序列
|
16
|
+
worker_nums: <int> 进程数
|
17
|
+
b_display_progress: <boolean> 是否显示进度条
|
18
|
+
timeout: <int/float> 每个进程的最大等待时间,单位是s
|
19
|
+
默认为 None,表示允许等待无限长的时间
|
20
|
+
_hook_for_debug: <dict/None> 当设置为非 None 值时,将保存中间的执行信息。
|
21
|
+
包括:
|
22
|
+
- "execution_orders": 执行顺序
|
23
|
+
- "completion_orders": 完成顺序
|
24
|
+
这些信息与最终结果无关,仅面向更底层的调试需求,任何人都不应依赖该特性
|
25
|
+
返回:
|
26
|
+
res_ls, failed_idx_ls
|
27
|
+
执行结果列表,以及执行失败的执行器索引列表
|
28
|
+
"""
|
29
|
+
executor_ls = []
|
30
|
+
for i in executors:
|
31
|
+
assert isinstance(i, (Executor,))
|
32
|
+
try:
|
33
|
+
pickle.dumps(i)
|
34
|
+
except:
|
35
|
+
raise AttributeError(
|
36
|
+
f'非法任务。因为进程池中的任务必须要能被pickle化。\n对象 {i} 无法被 pickle,请检查其中是否使用了闭包内定义的函数')
|
37
|
+
executor_ls.append(i)
|
38
|
+
if b_display_progress:
|
39
|
+
from tqdm import tqdm
|
40
|
+
p_bar = tqdm(total=len(executor_ls))
|
41
|
+
else:
|
42
|
+
p_bar = None
|
43
|
+
|
44
|
+
if isinstance(_hook_for_debug, dict):
|
45
|
+
_execution_orders, _completion_orders = Manager().list(), Manager().list()
|
46
|
+
else:
|
47
|
+
_execution_orders, _completion_orders = None, None
|
48
|
+
|
49
|
+
res_ls = [None] * len(executor_ls)
|
50
|
+
failed_idx_ls = []
|
51
|
+
with concurrent.futures.ProcessPoolExecutor(max_workers=worker_nums) as process_pool:
|
52
|
+
# 提交任务并添加进度回调
|
53
|
+
futures = []
|
54
|
+
for i, executor in enumerate(executor_ls):
|
55
|
+
future = process_pool.submit(wrapper, executor, timeout, i, _execution_orders, _completion_orders)
|
56
|
+
if b_display_progress:
|
57
|
+
future.add_done_callback(lambda _: p_bar.update())
|
58
|
+
futures.append(future)
|
59
|
+
|
60
|
+
# 收集结果
|
61
|
+
for i, future in enumerate(futures):
|
62
|
+
try:
|
63
|
+
res, b_success = future.result()
|
64
|
+
except:
|
65
|
+
b_success = False
|
66
|
+
if b_success:
|
67
|
+
res_ls[i] = res
|
68
|
+
else:
|
69
|
+
failed_idx_ls.append(i)
|
70
|
+
|
71
|
+
if b_display_progress:
|
72
|
+
p_bar.close()
|
73
|
+
|
74
|
+
if isinstance(_hook_for_debug, (dict,)):
|
75
|
+
_hook_for_debug.update({
|
76
|
+
"execution_orders": list(_execution_orders),
|
77
|
+
"completion_orders": list(_completion_orders)
|
78
|
+
})
|
79
|
+
|
80
|
+
return res_ls, failed_idx_ls
|
81
|
+
|
82
|
+
|
83
|
+
if __name__ == '__main__':
|
84
|
+
import time
|
85
|
+
|
86
|
+
|
87
|
+
def func_(i):
|
88
|
+
# 模拟部分任务长时间运行,部分任务正常结束
|
89
|
+
if i in [2, 3, 7]:
|
90
|
+
time.sleep(100)
|
91
|
+
else:
|
92
|
+
time.sleep(0.01)
|
93
|
+
print(f"任务 {i} 执行完成")
|
94
|
+
return i * 2
|
95
|
+
|
96
|
+
|
97
|
+
hook_for_debug = dict()
|
98
|
+
a = time.time()
|
99
|
+
results, failed = multi_process_execute(
|
100
|
+
executors=[Executor(func=func_, args=(i,)) for i in range(10)],
|
101
|
+
worker_nums=10,
|
102
|
+
timeout=0.2,
|
103
|
+
_hook_for_debug=hook_for_debug
|
104
|
+
)
|
105
|
+
gap = time.time() - a
|
106
|
+
print("执行结果:", results)
|
107
|
+
print("超时失败的任务索引:", failed)
|
108
|
+
print("调试信息:", hook_for_debug)
|
109
|
+
print("总耗时:", gap)
|
@@ -1,14 +1,18 @@
|
|
1
1
|
import concurrent.futures
|
2
|
+
from multiprocessing import Manager
|
2
3
|
from kevin_toolbox.computer_science.data_structure import Executor
|
4
|
+
from kevin_toolbox.computer_science.algorithm.parallel_and_concurrent.utils import wrapper_for_mt as wrapper
|
5
|
+
from kevin_toolbox.computer_science.algorithm.parallel_and_concurrent.utils import DEFAULT_THREAD_NUMS
|
3
6
|
|
4
7
|
|
5
|
-
def multi_thread_execute(executors,
|
8
|
+
def multi_thread_execute(executors, worker_nums=DEFAULT_THREAD_NUMS, b_display_progress=True, timeout=None,
|
9
|
+
_hook_for_debug=None):
|
6
10
|
"""
|
7
11
|
多线程执行
|
8
12
|
|
9
13
|
参数:
|
10
14
|
executors: <list/generator/iterator of Executor> 执行器序列
|
11
|
-
|
15
|
+
worker_nums: <int> 线程数
|
12
16
|
b_display_progress: <boolean> 是否显示进度条
|
13
17
|
timeout: <int> 每个线程的最大等待时间,单位是s
|
14
18
|
默认为 None,表示允许等待无限长的时间
|
@@ -30,37 +34,42 @@ def multi_thread_execute(executors, thread_nums=50, b_display_progress=True, tim
|
|
30
34
|
p_bar = tqdm(total=len(executor_ls))
|
31
35
|
else:
|
32
36
|
p_bar = None
|
33
|
-
_execution_orders, _completion_orders = [], []
|
34
37
|
|
35
|
-
|
36
|
-
|
37
|
-
|
38
|
-
|
39
|
-
_completion_orders.append(idx)
|
40
|
-
if p_bar is not None:
|
41
|
-
p_bar.update()
|
42
|
-
|
43
|
-
return res
|
38
|
+
if isinstance(_hook_for_debug, dict):
|
39
|
+
_execution_orders, _completion_orders = Manager().list(), Manager().list()
|
40
|
+
else:
|
41
|
+
_execution_orders, _completion_orders = None, None
|
44
42
|
|
45
|
-
res_ls
|
46
|
-
|
43
|
+
res_ls = [None] * len(executor_ls)
|
44
|
+
failed_idx_ls = []
|
45
|
+
with concurrent.futures.ThreadPoolExecutor(max_workers=worker_nums) as thread_pool:
|
47
46
|
# 提交任务
|
48
|
-
futures = [
|
49
|
-
|
50
|
-
|
51
|
-
|
47
|
+
futures = []
|
48
|
+
for i, executor in enumerate(executor_ls):
|
49
|
+
future = thread_pool.submit(wrapper, executor, timeout, i, _execution_orders, _completion_orders)
|
50
|
+
if b_display_progress:
|
51
|
+
future.add_done_callback(lambda _: p_bar.update())
|
52
|
+
futures.append(future)
|
53
|
+
|
54
|
+
# 收集结果
|
52
55
|
for i, future in enumerate(futures):
|
53
|
-
|
54
|
-
|
56
|
+
try:
|
57
|
+
res, b_success = future.result()
|
58
|
+
except:
|
59
|
+
b_success = False
|
60
|
+
if b_success:
|
61
|
+
res_ls[i] = res
|
55
62
|
else:
|
56
|
-
res_ls.append(None)
|
57
63
|
failed_idx_ls.append(i)
|
64
|
+
|
58
65
|
if b_display_progress:
|
59
66
|
p_bar.close()
|
60
67
|
|
61
|
-
#
|
62
68
|
if isinstance(_hook_for_debug, (dict,)):
|
63
|
-
_hook_for_debug.update(
|
69
|
+
_hook_for_debug.update({
|
70
|
+
"execution_orders": list(_execution_orders),
|
71
|
+
"completion_orders": list(_completion_orders)
|
72
|
+
})
|
64
73
|
|
65
74
|
return res_ls, failed_idx_ls
|
66
75
|
|
@@ -70,15 +79,27 @@ if __name__ == '__main__':
|
|
70
79
|
|
71
80
|
|
72
81
|
def func_(i):
|
82
|
+
# 模拟部分任务长时间运行,部分任务正常结束
|
73
83
|
if i in [2, 3, 7]:
|
74
|
-
time.sleep(
|
84
|
+
time.sleep(100)
|
85
|
+
elif i in [4, 5, 6]:
|
86
|
+
time.sleep(0.01)
|
75
87
|
else:
|
76
|
-
time.sleep(
|
77
|
-
print(i)
|
88
|
+
time.sleep(0.05)
|
89
|
+
print(f"任务 {i} 执行完成")
|
78
90
|
return i * 2
|
79
91
|
|
80
92
|
|
81
93
|
hook_for_debug = dict()
|
82
|
-
|
83
|
-
|
84
|
-
|
94
|
+
a = time.time()
|
95
|
+
results, failed = multi_thread_execute(
|
96
|
+
executors=[Executor(func=func_, args=(i,)) for i in range(10)],
|
97
|
+
worker_nums=5,
|
98
|
+
timeout=0.2,
|
99
|
+
_hook_for_debug=hook_for_debug
|
100
|
+
)
|
101
|
+
gap = time.time() - a
|
102
|
+
print("执行结果:", results)
|
103
|
+
print("超时失败的任务索引:", failed)
|
104
|
+
print("调试信息:", hook_for_debug)
|
105
|
+
print("总耗时:", gap)
|
@@ -0,0 +1,15 @@
|
|
1
|
+
from .wrapper_with_timeout_1 import wrapper_with_timeout_1
|
2
|
+
from .wrapper_with_timeout_2 import wrapper_with_timeout_2
|
3
|
+
|
4
|
+
import signal
|
5
|
+
import multiprocessing
|
6
|
+
|
7
|
+
if callable(getattr(signal, "setitimer", None)):
|
8
|
+
wrapper_for_mp = wrapper_with_timeout_1 # 效率更高,优先选择
|
9
|
+
else:
|
10
|
+
wrapper_for_mp = wrapper_with_timeout_2
|
11
|
+
|
12
|
+
wrapper_for_mt = wrapper_with_timeout_2
|
13
|
+
|
14
|
+
DEFAULT_PROCESS_NUMS = multiprocessing.cpu_count() + 2
|
15
|
+
DEFAULT_THREAD_NUMS = DEFAULT_PROCESS_NUMS * 2
|
kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/utils/wrapper_with_timeout_1.py
ADDED
@@ -0,0 +1,69 @@
|
|
1
|
+
import signal
|
2
|
+
|
3
|
+
|
4
|
+
# 定义超时异常
|
5
|
+
class TimeoutException(Exception):
|
6
|
+
pass
|
7
|
+
|
8
|
+
|
9
|
+
# 定时器信号处理函数
|
10
|
+
def __alarm_handler(*args, **kwargs):
|
11
|
+
raise TimeoutException("任务超时")
|
12
|
+
|
13
|
+
|
14
|
+
def wrapper_with_timeout_1(executor, timeout=None, idx=-1, _execution_orders=None, _completion_orders=None):
|
15
|
+
"""
|
16
|
+
限制执行时间,使用 multiprocessing.Process 强制终止超时任务
|
17
|
+
该函数仅适用于多进程以及 unix 操作系统
|
18
|
+
|
19
|
+
参数:
|
20
|
+
executor: <Executor>执行器,需实现 run() 方法
|
21
|
+
idx: <int> 任务索引(用于调试)
|
22
|
+
timeout: <int/float>最大等待时间(单位:秒,支持 float)
|
23
|
+
_execution_orders, _completion_orders: 用于记录调试信息的 Manager.list
|
24
|
+
返回:
|
25
|
+
(result, b_success) 若超时或异常则 b_success 为 False
|
26
|
+
"""
|
27
|
+
if _execution_orders is not None:
|
28
|
+
_execution_orders.append(idx)
|
29
|
+
|
30
|
+
# 定时器
|
31
|
+
if timeout is not None:
|
32
|
+
signal.signal(signal.SIGALRM, __alarm_handler)
|
33
|
+
signal.setitimer(signal.ITIMER_REAL, timeout)
|
34
|
+
|
35
|
+
# 执行
|
36
|
+
res, b_success = None, True
|
37
|
+
try:
|
38
|
+
res = executor.run()
|
39
|
+
if _completion_orders is not None:
|
40
|
+
_completion_orders.append(idx)
|
41
|
+
except TimeoutException:
|
42
|
+
b_success = False
|
43
|
+
finally:
|
44
|
+
signal.alarm(0) # 取消定时器
|
45
|
+
return res, b_success
|
46
|
+
|
47
|
+
|
48
|
+
if __name__ == '__main__':
|
49
|
+
import time
|
50
|
+
|
51
|
+
|
52
|
+
def func_(i):
|
53
|
+
if i in [2, 3, 7]:
|
54
|
+
time.sleep(300)
|
55
|
+
else:
|
56
|
+
time.sleep(0.5)
|
57
|
+
return i * 2
|
58
|
+
|
59
|
+
|
60
|
+
from kevin_toolbox.computer_science.data_structure import Executor
|
61
|
+
|
62
|
+
print(wrapper_with_timeout_1(Executor(func=func_, args=(2,)), timeout=1))
|
63
|
+
print(wrapper_with_timeout_1(Executor(func=func_, args=(1,)), timeout=1))
|
64
|
+
|
65
|
+
execution_orders = []
|
66
|
+
completion_orders = []
|
67
|
+
print(wrapper_with_timeout_1(Executor(func=func_, args=(2,)), timeout=1, _execution_orders=execution_orders,
|
68
|
+
_completion_orders=completion_orders))
|
69
|
+
print(execution_orders, completion_orders)
|
kevin_toolbox/computer_science/algorithm/parallel_and_concurrent/utils/wrapper_with_timeout_2.py
ADDED
@@ -0,0 +1,76 @@
|
|
1
|
+
from multiprocessing import Process, Queue
|
2
|
+
|
3
|
+
|
4
|
+
def __inner_wrapper(q, executor):
|
5
|
+
try:
|
6
|
+
res = executor.run()
|
7
|
+
q.put((res, True))
|
8
|
+
except:
|
9
|
+
q.put((None, False))
|
10
|
+
|
11
|
+
|
12
|
+
def wrapper_with_timeout_2(executor, timeout=None, idx=-1, _execution_orders=None, _completion_orders=None):
|
13
|
+
"""
|
14
|
+
限制执行时间,使用 multiprocessing.Process 强制终止超时任务
|
15
|
+
该函数适用于多线程、多进程以及所有操作系统,但是效率相较于 wrapper_with_timeout_1 较差
|
16
|
+
|
17
|
+
参数:
|
18
|
+
executor: <Executor>执行器,需实现 run() 方法
|
19
|
+
idx: <int> 任务索引(用于调试)
|
20
|
+
timeout: <int/float>最大等待时间(单位:秒,支持 float)
|
21
|
+
_execution_orders, _completion_orders: 用于记录调试信息的 Manager.list
|
22
|
+
返回:
|
23
|
+
(result, b_success) 若超时或异常则 b_success 为 False
|
24
|
+
"""
|
25
|
+
if _execution_orders is not None:
|
26
|
+
_execution_orders.append(idx)
|
27
|
+
|
28
|
+
res, b_success = None, False
|
29
|
+
if timeout is not None:
|
30
|
+
q = Queue()
|
31
|
+
p = Process(target=__inner_wrapper, args=(q, executor))
|
32
|
+
p.start()
|
33
|
+
p.join(timeout) # 最多等待 timeout 秒
|
34
|
+
|
35
|
+
if q.qsize():
|
36
|
+
try:
|
37
|
+
res, b_success = q.get_nowait()
|
38
|
+
except:
|
39
|
+
pass
|
40
|
+
if p.is_alive():
|
41
|
+
p.terminate()
|
42
|
+
p.join()
|
43
|
+
else:
|
44
|
+
try:
|
45
|
+
res, b_success = executor.run(), True
|
46
|
+
except:
|
47
|
+
pass
|
48
|
+
|
49
|
+
if b_success:
|
50
|
+
if _completion_orders is not None:
|
51
|
+
_completion_orders.append(idx)
|
52
|
+
return res, b_success
|
53
|
+
|
54
|
+
|
55
|
+
if __name__ == '__main__':
|
56
|
+
import time
|
57
|
+
|
58
|
+
|
59
|
+
def func_(i):
|
60
|
+
if i in [2, 3, 7]:
|
61
|
+
time.sleep(300)
|
62
|
+
else:
|
63
|
+
time.sleep(0.5)
|
64
|
+
return i * 2
|
65
|
+
|
66
|
+
|
67
|
+
from kevin_toolbox.computer_science.data_structure import Executor
|
68
|
+
|
69
|
+
print(wrapper_with_timeout_2(Executor(func=func_, args=(2,)), timeout=1))
|
70
|
+
print(wrapper_with_timeout_2(Executor(func=func_, args=(1,)), timeout=1))
|
71
|
+
|
72
|
+
execution_orders = []
|
73
|
+
completion_orders = []
|
74
|
+
print(wrapper_with_timeout_2(Executor(func=func_, args=(2,)), timeout=1, _execution_orders=execution_orders,
|
75
|
+
_completion_orders=completion_orders))
|
76
|
+
print(execution_orders, completion_orders)
|
@@ -83,7 +83,7 @@ class Executor:
|
|
83
83
|
# 校验参数
|
84
84
|
# func
|
85
85
|
assert paras["func"] is None or callable(paras["func"]), \
|
86
|
-
f
|
86
|
+
f'func should be callable, but get a {type(paras["func"])}'
|
87
87
|
# args
|
88
88
|
assert isinstance(paras["args"], (list, tuple,)) and isinstance(paras["f_args"], (list, tuple,))
|
89
89
|
for i, f in enumerate(paras["f_args"]):
|
@@ -93,7 +93,7 @@ class Executor:
|
|
93
93
|
assert isinstance(paras["kwargs"], (dict,)) and isinstance(paras["f_kwargs"], (dict,))
|
94
94
|
for k, v in paras["f_kwargs"].items():
|
95
95
|
assert callable(v) and isinstance(k, (str,)), \
|
96
|
-
f"item {
|
96
|
+
f"item {k} in f_kwargs should be (str, callable) pairs, but get a ({type(k)}, {type(v)})"
|
97
97
|
|
98
98
|
# update paras
|
99
99
|
self.paras = paras
|
@@ -0,0 +1 @@
|
|
1
|
+
from .write_excel_with_matrix import write_excel_with_matrix as write_with_matrix
|
@@ -0,0 +1,105 @@
|
|
1
|
+
import os
|
2
|
+
import openpyxl
|
3
|
+
from openpyxl.styles import Alignment, Font, PatternFill
|
4
|
+
from kevin_toolbox.patches.for_os import remove
|
5
|
+
|
6
|
+
|
7
|
+
# excel
|
8
|
+
def write_excel_with_matrix(matrix, file_path=None, file_obj=None, sheet_name="matrix",
|
9
|
+
column_label_ls=None, row_label_ls=None, column_title="", row_title="", main_title=""):
|
10
|
+
"""
|
11
|
+
将矩阵写入到 excel 文件中
|
12
|
+
|
13
|
+
参数:
|
14
|
+
file_path: 要写入到哪个文件
|
15
|
+
file_obj: <openpyxl.Workbook> 文件对象
|
16
|
+
注意!!以上两个参数指定其一即可,同时指定时候,以后者为准。
|
17
|
+
sheet_name: 要写入到哪个sheet页面
|
18
|
+
matrix: <np.array or np.matrix> 矩阵
|
19
|
+
column_label_ls, row_label_ls: 行列标签
|
20
|
+
column_title, row_title: 行列标题
|
21
|
+
main_title: 总标题
|
22
|
+
"""
|
23
|
+
assert file_path is not None or file_obj is not None
|
24
|
+
|
25
|
+
if file_obj is None:
|
26
|
+
file_path = os.path.abspath(os.path.expanduser(file_path))
|
27
|
+
os.makedirs(os.path.dirname(file_path), exist_ok=True)
|
28
|
+
# 判断文件是否存在,不存在则新建,否则读取文件
|
29
|
+
if not os.path.isfile(file_path):
|
30
|
+
wb = openpyxl.Workbook() # 创建文件对象
|
31
|
+
# wb对象创建后,默认含有一个默认的名为 Sheet 的 页面,将其删除
|
32
|
+
ws_ = wb.active
|
33
|
+
wb.remove(ws_)
|
34
|
+
else:
|
35
|
+
wb = openpyxl.load_workbook(file_path)
|
36
|
+
else:
|
37
|
+
assert isinstance(file_obj, openpyxl.Workbook)
|
38
|
+
wb = file_obj
|
39
|
+
# 判断sheet是否存在,不存在则建立,否则先删除再建立
|
40
|
+
if sheet_name in wb.sheetnames:
|
41
|
+
ws = wb[sheet_name]
|
42
|
+
wb.remove(ws)
|
43
|
+
ws = wb.create_sheet(sheet_name)
|
44
|
+
|
45
|
+
# 开始写
|
46
|
+
matrix_r_offset, matrix_c_offset = 1, 1 # 矩阵的起始位置
|
47
|
+
r_offset, c_offset = 1, 1 # 目前的写入位置
|
48
|
+
for i in [main_title, column_title, column_label_ls]:
|
49
|
+
if i:
|
50
|
+
matrix_r_offset += 1
|
51
|
+
for j in [row_title, row_label_ls]:
|
52
|
+
if j:
|
53
|
+
matrix_c_offset += 1
|
54
|
+
# print(matrix)
|
55
|
+
matrix_row_num, matrix_column_num = matrix.shape[0], matrix.shape[1]
|
56
|
+
# 标题
|
57
|
+
alignment = Alignment(horizontal="center", vertical="center", wrap_text=True)
|
58
|
+
if main_title:
|
59
|
+
ws.merge_cells(start_row=r_offset, start_column=1, end_row=r_offset,
|
60
|
+
end_column=matrix_c_offset + matrix_column_num - 1)
|
61
|
+
ws.cell(row=r_offset, column=1).value = main_title
|
62
|
+
ws.cell(row=r_offset, column=1).alignment = alignment
|
63
|
+
ws.cell(row=r_offset, column=1).font = Font(size=10, bold=True, name='微软雅黑', color="000000")
|
64
|
+
r_offset += 1
|
65
|
+
if column_title:
|
66
|
+
ws.merge_cells(start_row=r_offset, start_column=matrix_c_offset, end_row=r_offset,
|
67
|
+
end_column=matrix_c_offset + matrix_column_num - 1)
|
68
|
+
ws.cell(row=r_offset, column=matrix_c_offset).value = column_title
|
69
|
+
ws.cell(row=r_offset, column=matrix_c_offset).alignment = alignment
|
70
|
+
ws.cell(row=r_offset, column=matrix_c_offset).font = Font(size=10, bold=True, name='微软雅黑', color="000000")
|
71
|
+
r_offset += 1
|
72
|
+
if row_title:
|
73
|
+
ws.merge_cells(start_row=matrix_r_offset, start_column=1, end_row=matrix_r_offset + matrix_row_num - 1,
|
74
|
+
end_column=1)
|
75
|
+
ws.cell(row=matrix_r_offset, column=1).value = row_title
|
76
|
+
ws.cell(row=matrix_r_offset, column=1).alignment = alignment
|
77
|
+
ws.cell(row=matrix_r_offset, column=1).font = Font(size=10, bold=True, name='微软雅黑', color="000000")
|
78
|
+
c_offset += 1
|
79
|
+
# 标签
|
80
|
+
if column_label_ls:
|
81
|
+
for i in range(matrix_column_num):
|
82
|
+
ws.cell(row=r_offset, column=matrix_c_offset + i).value = column_label_ls[i]
|
83
|
+
ws.cell(row=r_offset, column=matrix_c_offset + i).alignment = alignment
|
84
|
+
ws.cell(row=r_offset, column=matrix_c_offset + i).fill = PatternFill(patternType="solid",
|
85
|
+
start_color="33CCFF")
|
86
|
+
r_offset += 1
|
87
|
+
if row_label_ls:
|
88
|
+
for i in range(matrix_row_num):
|
89
|
+
ws.cell(row=matrix_r_offset + i, column=c_offset).value = row_label_ls[i]
|
90
|
+
ws.cell(row=matrix_r_offset + i, column=c_offset).alignment = alignment
|
91
|
+
ws.cell(row=matrix_r_offset + i, column=c_offset).fill = PatternFill(patternType="solid",
|
92
|
+
start_color="33CCFF")
|
93
|
+
c_offset += 1
|
94
|
+
# 校验,可省略
|
95
|
+
if not (c_offset == matrix_c_offset and r_offset == matrix_r_offset):
|
96
|
+
print("wrong here")
|
97
|
+
for r_ in range(matrix_row_num):
|
98
|
+
for c_ in range(matrix_column_num):
|
99
|
+
ws.cell(row=matrix_r_offset + r_, column=matrix_c_offset + c_).value = matrix[r_][c_]
|
100
|
+
|
101
|
+
if file_path is not None:
|
102
|
+
remove(file_path, ignore_errors=True)
|
103
|
+
wb.save(file_path)
|
104
|
+
|
105
|
+
return wb
|
@@ -23,6 +23,7 @@ def read_json(file_path=None, file_obj=None, converters=None, b_use_suggested_co
|
|
23
23
|
"""
|
24
24
|
assert file_path is not None or file_obj is not None
|
25
25
|
if file_path is not None:
|
26
|
+
file_path = os.path.abspath(os.path.expanduser(file_path))
|
26
27
|
assert os.path.isfile(file_path), f'file {file_path} not found'
|
27
28
|
file_obj = open(file_path, 'r')
|
28
29
|
elif isinstance(file_obj, (BytesIO,)):
|
@@ -4,8 +4,13 @@ import copy
|
|
4
4
|
from kevin_toolbox.data_flow.file.json_.converter import integrate, escape_tuple_and_set, escape_non_str_dict_key
|
5
5
|
from kevin_toolbox.nested_dict_list import traverse
|
6
6
|
|
7
|
+
format_s = {
|
8
|
+
"pretty_printed": dict(indent=4, ensure_ascii=False, sort_keys=False),
|
9
|
+
"minified": dict(indent=None, ensure_ascii=False, sort_keys=False, separators=(',', ':'))
|
10
|
+
}
|
7
11
|
|
8
|
-
|
12
|
+
|
13
|
+
def write_json(content, file_path, converters=None, b_use_suggested_converter=False, output_format="pretty_printed"):
|
9
14
|
"""
|
10
15
|
写入 json file
|
11
16
|
|
@@ -13,7 +18,6 @@ def write_json(content, file_path, sort_keys=False, converters=None, b_use_sugge
|
|
13
18
|
content: 待写入内容
|
14
19
|
file_path: <path or None> 写入路径
|
15
20
|
当设置为 None 时,将直接把(经converters处理后的)待写入内容作为结果返回,而不进行实际的写入
|
16
|
-
sort_keys
|
17
21
|
converters: <list of converters> 对写入内容中每个节点的处理方式
|
18
22
|
转换器 converter 应该是一个形如 def(x): ... ; return x 的函数,具体可以参考
|
19
23
|
json_.converter 中已实现的转换器
|
@@ -22,8 +26,27 @@ def write_json(content, file_path, sort_keys=False, converters=None, b_use_sugge
|
|
22
26
|
可以避免因 json 的读取/写入而丢失部分信息。
|
23
27
|
默认为 False。
|
24
28
|
注意:当 converters 非 None,此参数失效,以 converters 中的具体设置为准
|
29
|
+
output_format: <str/dict/tuple> json的输出格式
|
30
|
+
对于 str 目前支持以下取值:
|
31
|
+
- "pretty_printed": 通过添加大量的空格和换行符来格式化输出,使输出更易读
|
32
|
+
- "minified": 删除所有空格和换行符,使输出更紧凑
|
33
|
+
默认为 pretty_printed。
|
34
|
+
对于 dict,将允许使用更加细致的格式设定,比如:
|
35
|
+
{"indent": 2, ensure_ascii=True}
|
36
|
+
如果需要基于已有格式进行微调可以使用以下方式:
|
37
|
+
("pretty_printed", {"indent": 2, ensure_ascii=True})
|
25
38
|
"""
|
39
|
+
global format_s
|
26
40
|
assert isinstance(file_path, (str, type(None)))
|
41
|
+
if isinstance(output_format, (str,)):
|
42
|
+
output_format = format_s[output_format]
|
43
|
+
elif isinstance(output_format, (tuple,)):
|
44
|
+
output_format = format_s[output_format[0]]
|
45
|
+
output_format.update(output_format[1])
|
46
|
+
elif isinstance(output_format, (dict,)):
|
47
|
+
pass
|
48
|
+
else:
|
49
|
+
raise ValueError(f'Unsupported output_format: {output_format}.')
|
27
50
|
|
28
51
|
if converters is None and b_use_suggested_converter:
|
29
52
|
converters = [escape_tuple_and_set, escape_non_str_dict_key]
|
@@ -35,7 +58,7 @@ def write_json(content, file_path, sort_keys=False, converters=None, b_use_sugge
|
|
35
58
|
converter=lambda _, x: converter(x),
|
36
59
|
b_traverse_matched_element=True)[0]
|
37
60
|
|
38
|
-
content = json.dumps(content,
|
61
|
+
content = json.dumps(content, **output_format)
|
39
62
|
|
40
63
|
if file_path is not None:
|
41
64
|
file_path = os.path.abspath(os.path.expanduser(file_path))
|
@@ -44,3 +67,13 @@ def write_json(content, file_path, sort_keys=False, converters=None, b_use_sugge
|
|
44
67
|
f.write(content)
|
45
68
|
else:
|
46
69
|
return content
|
70
|
+
|
71
|
+
|
72
|
+
if __name__ == "__main__":
|
73
|
+
a = {'rect': {'l:eft': [0, 1, 2], 'top': 67, 'right': 286, 'bottom': 332}}
|
74
|
+
res_0 = write_json(a, file_path=None, output_format="pretty_printed")
|
75
|
+
print(res_0)
|
76
|
+
print(len(res_0))
|
77
|
+
res_1 = write_json(a, file_path=None, output_format="minified")
|
78
|
+
print(res_1)
|
79
|
+
print(len(res_1))
|
@@ -1 +1,2 @@
|
|
1
|
-
|
1
|
+
from .check_version_and_update import check_version_and_update
|
2
|
+
from .check_validity_and_uninstall import check_validity_and_uninstall
|