【问题标题】:Regex to match company names from copyright statements under yet more conditions正则表达式在更多条件下匹配版权声明中的公司名称
【发布时间】:2019-03-21 20:07:53
【问题描述】:

一段时间以来,我一直在尝试寻找一个强大的正则表达式来从版权声明中提取公司名称(并且对正则表达式了解不多)。

在这个问题中:Regex to match company names from copyright statements under several conditions

我得到了正则表达式:

(?i)(?:©(?:\s*Copyright)?|Copyright(?:\s*©)?)\s*\d+(?:\s*-\s*\d+)?\s*(.*?(?=\W*All\s+rights\s+reserved)|[^.]*(?=\.)|.*)

但当我尝试更多示例时,我发现这还不够。我想对其进行更改,使其也符合以下条件,同时仍适用于所有以前的情况:

  1. 考虑到在“©”或“copyright”(以最后出现的为准)之前可能会出现其他任何内容并忽略它。

例子:

602-226-2389 ©2019 Endurance International Group.
Copyright 1999 — 2019 © Iflexion. All rights reserved.
  1. 请注意,在“©”或“版权”之后可能不是年份,而是公司名称。

例子:

ISO 9001:2008, ISO/ IEC 27001:2005 © Mobikasa 2019
  1. 考虑到年份可能出现在“版权”或“©”一词之前(我认为条件 1 也满足这一点)

例子:

© 2019 Copyright arcadia.io.
2018 © Power Tools LLC
  1. 如果有 |匹配到那里,然后忽略其余部分:

例子:

Copyright 2019 ComputerEase Construction Software | 1-800-544-2530

【问题讨论】:

  • 那么,考虑到新规则 4,All Comments © Copyright 2017 Kroger | The Kroger Co. All Rights Reserved 的预期输出是什么?
  • @WiktorStribiżew 应该是“克罗格”
  • 有一个规则是在点之前停止,但是© 2019 Copyright arcadia.io.呢?
  • 是的,这很棘手。但在那种情况下,我们可以只保留阿卡迪亚
  • 然后检查regex101.com/r/n1TBiu/1

标签: python regex


【解决方案1】:

你可以使用

(?i)(?:©(?:\s*(?:\d{4}(?:\s*[-—–]\s*\d{4})?)?\s*Copyright)?|Copyright(?:\s*(?:\d{4}(?:\s*[-—–]\s*\d{4})?)?\s*©)?)(?:\s*\d{4}(?:\s*[-—–]\s*\d{4})?)?\s*(.*?(?=\s*[.|]|\W*All\s+rights\s+reserved)|.*\b)

regex demo

Python 代码:

import re
s = "Copyright © 2019 Apple Inc. All rights reserved.\r\n© 2019 Quid, Inc. All Rights Reserved.\r\n© 2009 Database Designs \r\n© 2019 Rediker Software, All Rights Reserved\r\n©2019 EVOSUS, INC. ALL RIGHTS RESERVED\r\n© 2019 Walmart. All Rights Reserved.\r\n© Copyright 2003-2019 Exxon Mobil Corporation. All Rights Reserved.\r\nCopyright © 1978-2019 Berkshire Hathaway Inc.\r\n© 2019 McKesson Corporation\r\n© 2019 UnitedHealth Group. All rights reserved.\r\n© Copyright 1999 - 2019 CVS Health\r\nCopyright 2019 General Motors. All Rights Reserved.\r\n© 2019 Ford Motor Company\r\n©2019 AT&T Intellectual Property. All rights reserved.\r\n© 2019 GENERAL ELECTRIC\r\nCopyright ©2019 AmerisourceBergen Corporation. All Rights Reserved.\r\n© 2019 Verizon\r\n© 2019 Fannie Mae\r\nCopyright © 2018 Jonas Construction Software Inc. All rights reserved.\r\nAll Comments © Copyright 2017 Kroger | The Kroger Co. All Rights Reserved\r\n© 2019 Express Scripts Holding Company. All Rights Reserved. 1 Express Way, St. Louis, MO 63121\r\n© 2019 JPMorgan Chase & Co.\r\nCopyright © 1995 - 2018 Boeing. All Rights Reserved.\r\n© 2019 Bank of America Corporation. All rights reserved.\r\n© 1999 - 2019 Wells Fargo. All rights reserved. NMLSR ID 399801\r\n©2019 Cardinal Health. All rights reserved.\r\n© 2019 Quid, Inc All Rights Reserved.\r\n602-226-2389 ©2019 Endurance International Group.\r\nCopyright 1999 — 2019 © Iflexion. All rights reserved.\r\nISO 9001:2008, ISO/ IEC 27001:2005 © Mobikasa 2019\r\n© 2019 Copyright arcadia.io.\r\n2018 © Power Tools LLC\r\nCopyright 2019 ComputerEase Construction Software | 1-800-544-2530\r\n© 2019 3M. 3M Health Information Systems Privacy Policy"
rx = r'''(?xi)
(?:©                                        # Start of a group: © symbol
(?:\s*                                      #  Start of optional group: 0+ whitespaces
  (?:\d{4}                                  #   Start of optional group: 4 digits
    (?:\s*[-—–]\s*\d{4})?                   #     0+ spaces, dashes, spaces, 4 digits
  )?                                        #   End of group
  \s*Copyright                              #  Spaces and Copyright
)?                                          #  End of group 
|                                           #  OR 
Copyright                                   
 (?:\s*                                     #  Start of optional group: 0+ whitespaces
   (?:\d{4}                                 #   Start of optional group: 4 digits
     (?:\s*[-—–]\s*\d{4})?                  #     0+ spaces, dashes, spaces, 4 digits
   )?\s*©                                   #   End of group, 0+ spaces, ©
 )?                                         #  End of group
)                                           # End of group
(?:\s*\d{4}(?:\s*[-—–]\s*\d{4})?)?          # Optional group, 9999 optionally followed with dash enclosed with whitespaces and then 9999
\s*                                         # 0+ whitespaces
(                                           # Start of a capturing group:
   .*?                                      # any 0+ chars other than linebreak chars, as few as possible, up to...
    (?=\s*[.|]|                             # 0+ spaces and then | or ., or
        \W*All\s+rights\s+reserved)         # All rights reserved with any 0+ non-word chars before it
  |                                         # or
   .*\b                                     # any 0+ chars other than linebreak chars, as many as possible
)'''

for m in re.findall(rx, s):
    print(m)

请参阅Python demo。输出:

Apple Inc
Quid, Inc
Database Designs
Rediker Software
EVOSUS, INC
Walmart
Exxon Mobil Corporation
Berkshire Hathaway Inc
McKesson Corporation
UnitedHealth Group
CVS Health
General Motors
Ford Motor Company
AT&T Intellectual Property
GENERAL ELECTRIC
AmerisourceBergen Corporation
Verizon
Fannie Mae
Jonas Construction Software Inc
Kroger
Express Scripts Holding Company
JPMorgan Chase & Co
Boeing
Bank of America Corporation
Wells Fargo
Cardinal Health
Quid, Inc
Endurance International Group
Iflexion
Mobikasa 2019
arcadia
Power Tools LLC
ComputerEase Construction Software
3M

【讨论】:

  • 最后一个:如果公司名称中有数字怎么办?例如“© 2019 3M。3M 健康信息系统隐私政策”
  • @FilipeAleixo 预期的结果是什么? 3M?
  • “3M”。正如您在第一个答案中所说,我尝试将 {4} 添加到 \d ,但未能使其正常工作
【解决方案2】:

我相信this regex 可以满足您的需求。解释如下:

(?i)                                # make the regex case insensitive
(?:Copyright\s*©?|©\s*(Copyright)?) # Look for Copyright and/or © to get us started
([\d\s—-]+)?                        # There might be some digits, spaces, and dashes, but not necessarily
(©|Copyright)?\s*                   # Copyright or © could be separated by dates, so look for them again
(.+?)                               # This is the sugar we're looking for
(?=All rights reserved|\||$)        # If you find "All rights reserved" a | or end of string, stop capturing the text

【讨论】:

    【解决方案3】:

    我知道它的老问题,但想发布更好的解决方案。 我训练了 spacy 模型,该模型在 5k+ 版权文本样本上进行了训练。 这里是模特和作品Repo Link

    【讨论】:

      猜你喜欢
      • 2019-08-11
      • 1970-01-01
      • 1970-01-01
      • 2019-08-09
      • 1970-01-01
      • 1970-01-01
      • 1970-01-01
      • 1970-01-01
      • 2013-01-28
      相关资源
      最近更新 更多