在windows中使用python根据时间戳将多个日志文件按排序顺序合并为单个日志文件？

Question

有人能帮我解决以下问题吗

输入文件1:

abc.exe TryEndHand [520] 30-4-2020 8:8:52.786  [3636] Handshake value
Executing end handlier

abc.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:51.583  [3760] Create the general message 
Error Occured!! 30-4-2020 8:9:29.93  [2932] WARNING cannot remove qid
def.exe SharedCreateNamed [488] 30-5-2020 8:8:51.584  [3636] Create the general different message 

输入文件2:

abc1.exe TryEndHand [520] 30-5-2020 8:8:51.786  [3636] Handshake value 
abc1.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:52.58  [3760] Create the general message 
def1.exe SharedCreateNamed [488] 30-5-2020 8:8:51.53  [3636] Create the general different message

同样地

输入文件N: ...........

输出文件（1,2，---N）：

abc.exe TryEndHand [520] 30-4-2020 8:8:52.786  [3636] Handshake value 
Executing end handlier
Error Occured!! 30-4-2020 8:9:29.93  [2932] WARNING cannot remove qid
abc.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:51.583  [3760] Create the general message 
abc1.exe TryEndHand [520] 30-5-2020 8:8:51.786  [3636] Handshake value
def1.exe SharedCreateNamed [488] 30-5-2020 8:8:51.53  [3636] Create the general different message 
def.exe SharedCreateNamed [488] 30-5-2020 8:8:51.584  [3636] Create the general different message 
abc1.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:52.58  [3760] Create the general message 

请帮助根据输出文件中以粗体字母显示的日期和时间戳生成排序日志。 请注意，输入文件包含空格和不带时间戳的语句，还包含错误案例。

Answer 1

使用datefinder提取日期时间，使用pandas按日期时间排序

import os
# pip install datefinder
import datefinder
# pip install pandas
import pandas as pd

PATH = './input_files/'

# a dictonary with key - datetime and values - substrings
df_dict = dict()
for in_file in os.listdir(PATH):
    # read input files one by one
    input_file = open(os.path.join(PATH,in_file)).read()
    # to store prev datetime to handle lines with no datetime
    prev_dt = 0
    for line in input_file.splitlines():
        # parse lines one by one
        if line.strip():
            pre, dt, post, raw_dt = '', 0, '', 0
            # using datefinder to extract datetime
            for match in datefinder.find_dates(line, index=True, source=True):
                # change valid year conditions accoring to use_case
                if match[0].year == 2020:
                    dt, pre, raw_dt, post  = match[1], line[:match[2][0]], match[1], line[match[2][1]:]
            if dt:
                prev_dt = dt
                df_dict[dt] = [pre, raw_dt, post]
            else:
                df_dict[prev_dt].append(line)

df = pd.DataFrame.from_dict(dt_dict, orient='index')
df.index = pd.to_datetime(df.index)
next_lines = df.pop(3)
df = pd.concat([df,next_lines]).dropna(how='all').fillna('').sort_index()

output =df[df.columns].apply(lambda x: ' '.join(x), axis=1)
output.to_csv('output.txt', header=False, index=None)

输出：

abc.exe TryEndHand [520] 30-4-2020 8:8:52.786 [3636] Handshake value
Executing end handlier  
Error Occured!! 30-4-2020 8:9:29.93 [2932] WARNING cannot remove qid
def1.exe SharedCreateNamed [488] 30-5-2020 8:8:51.53 [3636] Create the general different message
abc.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:51.583 [3760] Create the general message 
def.exe SharedCreateNamed [488] 30-5-2020 8:8:51.584 [3636] Create the general different message
abc1.exe TryEndHand [520] 30-5-2020 8:8:51.786 [3636] Handshake value 
abc1.exe QueueSharedCreateNamed [488] 30-5-2020 8:8:52.58 [3760] Create the general message