用正则表达式修改TXT中的行的问题

Question

我在“发展”一个清理TXT行的脚本方面遇到了问题，附带的TXT示例如下：

Fri Oct 14 22:27:49.100 EDT

Interface          Status      Protocol    Description
--------------------------------------------------------------------------------
Lo0                up          up          Loopback0 interface configured by Netmiko
Lo55               up          up          
Lo100              up          up          ***MERGE LOOPBACK 100****
Lo111              up          up          Configured by NETCONF
Nu0                up          up          
Mg0/RP0/CPU0/0     up          up          DO NOT TOUCH THIS !
Gi0/0/0/0          admin-down  admin-down  ANSIBLE NXOS TEST
Gi0/0/0/1          admin-down  admin-down  test
Gi0/0/0/1.100      admin-down  admin-down  
Gi0/0/0/2          admin-down  admin-down  Link to P2 configured by Netmiko
Gi0/0/0/3          up          up          Configured by Ansible !!!!!!!!
Gi0/0/0/4          up          up          Updated by Ansible using Jinja Template
Gi0/0/0/5          up          up          Configured by Ansible !!!!!!
Gi0/0/0/6          admin-down  admin-down  Updated by Ansible using Jinja Template
Gi0/0/0/6.11       admin-down  admin-down
Lo20               admin-down  admin-down  
Lo22               up          up          Loopback para pruebas
[K --More--           [KLo69               admin-down  admin-down  
Gi0/3/3/4          up          up          A SDH 
Gi0/3/3/4.852      up          up          TMU a Red BIT
[K --More--           [KGi0/3/3/4.853      up          up          Configured by Ansible !!!!!!
Gi0/3/4/2.256      up          up          Frontera Cliente A
Gi0/3/4/2.257      up          up          Frontera Cliente B
[K --More--           [KGi0/3/4/2.261      up          up          Frontera Cliente C
Te0/7/0/3          admin-down  admin-down  
Mg0/RP0/CPU0/0     down        down        
Mg0/RP1/CPU0/0     admin-down  admin-down  
[KRP/0/RP0/CPU0:ROUTER1# 

剧本如下：

list_txt = [ruta/"prueba.txt"]

for txt in list_txt:

  with open(txt, "r") as f:

    lines = f.readlines()

  with open(txt, "w") as fw:
    for line in lines:

      if not re.match("-{5}|\s+|([A-Za-z0-9]+( [A-Za-z0-9]+)+)", line):
        fw.write(line)

有了这个脚本，我可以删除所有上面日期的行、空行和它们是纯连字符的行，问题是我试图添加两样东西：

1-在regex中添加，如果其中包含"CPU“一词，则行将被删除：

Mg0/RP0/CPU0/0     down        down        
Mg0/RP1/CPU0/0     admin-down  admin-down  
[KRP/0/RP0/CPU0:ROUTER1#   

2.另一方面，我需要删除一些行中添加的奇怪的内容，例如：

[K --More--           [KLo69               admin-down  admin-down

让它像这样干净：

Lo69               admin-down  admin-down

最后一个我尝试通过txt.lstrip("[K")来实现它，但是没有效果，我做的不对，它不工作，Regex我也没有按键，我也不能添加CPU这个词，我不太清楚如何生成Regex。

理想情况下，我希望你能把所有的东西都添加到现有的脚本中，这样就不会让事情变得那么复杂，你能帮我一下吗？

Answer 1

我认为正则表达式在这里并没有什么区别，只是可能解析标题，这样您就可以知道某一列的数据将从哪个索引开始：

spans = []
result = []
with open(txt, "r") as f:
    it = iter(f.readlines())    
    # Skip lines until headings are found
    for line in it:
        if "  " in line:
            break
    # Get start/end indices of each column
    spans = [
         m.span()
         for m in re.finditer(r"\S+\s*", line)
    ]
    spans[-1] = (spans[-1][0], -1)
    next(it) # Skip line with just hyphens
    # Iterate the real data part of the input
    for line in it:
        if line.startswith("[K --More--"):
            # Get content part of this line
            line = line[line.index("[K", 10) + 2:]
        elif line.startswith("[K") or "/CPU" in line:
            continue  # Not interested
        result.append(tuple(line[start:end].rstrip() for start, end in spans))

# Output what was collected:
for record in result:
    print(record)

给定示例文件的输出是：

('Lo0', 'up', 'up', 'Loopback0 interface configured by Netmiko')
('Lo55', 'up', 'up', '')
('Lo100', 'up', 'up', '***MERGE LOOPBACK 100****')
('Lo111', 'up', 'up', 'Configured by NETCONF')
('Nu0', 'up', 'up', '')
('Gi0/0/0/0', 'admin-down', 'admin-down', 'ANSIBLE NXOS TEST')
('Gi0/0/0/1', 'admin-down', 'admin-down', 'test')
('Gi0/0/0/1.100', 'admin-down', 'admin-down', '')
('Gi0/0/0/2', 'admin-down', 'admin-down', 'Link to P2 configured by Netmiko')
('Gi0/0/0/3', 'up', 'up', 'Configured by Ansible !!!!!!!!')
('Gi0/0/0/4', 'up', 'up', 'Updated by Ansible using Jinja Template')
('Gi0/0/0/5', 'up', 'up', 'Configured by Ansible !!!!!!')
('Gi0/0/0/6', 'admin-down', 'admin-down', 'Updated by Ansible using Jinja Template')
('Gi0/0/0/6.11', 'admin-down', 'admin-down', '')
('Lo20', 'admin-down', 'admin-down', '')
('Lo22', 'up', 'up', 'Loopback para pruebas')
('Lo69', 'admin-down', 'admin-down', '')
('Gi0/3/3/4', 'up', 'up', 'A SDH')
('Gi0/3/3/4.852', 'up', 'up', 'TMU a Red BIT')
('Gi0/3/3/4.853', 'up', 'up', 'Configured by Ansible !!!!!!')
('Gi0/3/4/2.256', 'up', 'up', 'Frontera Cliente A')
('Gi0/3/4/2.257', 'up', 'up', 'Frontera Cliente B')
('Gi0/3/4/2.261', 'up', 'up', 'Frontera Cliente C')
('Te0/7/0/3', 'admin-down', 'admin-down', '')

我可能错过了您的一些需求，但我认为使用这种方法很容易适应。

Answer 2

我补充了我能够以尽可能简化的方式给出的最后解决办法：

for txt in list_txt:

  with open(txt, "r") as f:

    lines = f.readlines()

  with open(txt, "w") as fw:
    for line in lines:

      if re.match("-{5}|\s+|([A-Za-z0-9]+( [A-Za-z0-9]+)+)", line) or re.search("CPU", line):
        pass
      elif "[K" in line:
        line2 = line.rindex("[K")
        line3 = line[line2+2:]
        fw.write(line3)
      else:
        fw.write(line)