用于在 Python 中检测公司名称的正则表达式答案

【问题标题】：Regex for detection company names in Python用于在 Python 中检测公司名称的正则表达式
【发布时间】：2021-07-04 20:59:54
【问题描述】：

我想使用 Python 用正则表达式检测公司名称。

这是我的想法：

公司名称应包含 1 到 3 个单词
公司名称的第一个单词应大写
公司名称中的一个单词可以有 .com 或 .co (Amazon.com Inc)
公司名称的最后一个单词（第四个单词）应为 Inc., Ltd, GmbH, AG, GmbH, Group, Holding 等。
名称的最后一个单词和 Inc. , Ltd, GmbH, AG 之间有时可以是 ',' 或 ', '

我尝试过类似的方法，但它不起作用：

address_1 = 'I work in Amazon.com Inc.'
address_2 = 'Company named Swiss Medic Holding invested in vaccine'
address_3 = 'what do you think about Abercrombie & Fitch Co. ?'
address_4 = 'do you work in Delta Group?'
address_5 = 'I have worked in CocaCola Gmbh'

regex_company = '([A-Z][\w]+[ -]+){1,3}(Ltd|ltd|LTD|llc|LLC|Inc|inc|INC|plc|Corp|Group)'
found = re.search(regex_company, address)

我想打印检测到的公司的结果我使用相同的正则表达式逻辑来查找街道地址并且效果很好，但对于公司名称却没有。这是我使用的正则表达式：

regex_street = "(\d{0,6})(?:\w)\s([A-Z][\w]+[ -]+){1,3}(Street|St|Road|Rd)

正则表达式逻辑：数字 + 1-3 个单词 + street/st/road/rd

【问题讨论】：

这可能是正则表达式的可能性范围太大。是否有其他方法可以获取公司名称？
I want to 你面临什么问题？什么是“不起作用”？
条件 3 使得它变得困难，因为正则表达式不是为“仅一个”而制作的。匹配每个单词以.com/.co结尾的模式是否可以？
@peer 很好的回答是肯定的和否定的，我没有这样的情况，我有 2 个或 3 个单词的名字，只有一个 .com，它只有一个以 .com 结尾的单词

标签： python regex

【解决方案1】：

使用

\b[A-Z]\w+(?:\.com?)?(?:[ -]+(?:&[ -]+)?[A-Z]\w+(?:\.com?)?){0,2}[,\s]+(?i:ltd|llc|inc|plc|co(?:rp)?|group|holding|gmbh)\b

见regex proof。

解释

--------------------------------------------------------------------------------
  \b                       the boundary between a word char (\w) and
                           something that is not a word char
--------------------------------------------------------------------------------
  [A-Z]                    any character of: 'A' to 'Z'
--------------------------------------------------------------------------------
  \w+                      word characters (a-z, A-Z, 0-9, _) (1 or
                           more times (matching the most amount
                           possible))
--------------------------------------------------------------------------------
  (?:                      group, but do not capture (optional
                           (matching the most amount possible)):
--------------------------------------------------------------------------------
    \.                       '.'
--------------------------------------------------------------------------------
    co                       'co'
--------------------------------------------------------------------------------
    m?                       'm' (optional (matching the most amount
                             possible))
--------------------------------------------------------------------------------
  )?                       end of grouping
--------------------------------------------------------------------------------
  (?:                      group, but do not capture (between 0 and 2
                           times (matching the most amount
                           possible)):
--------------------------------------------------------------------------------
    [ -]+                    any character of: ' ', '-' (1 or more
                             times (matching the most amount
                             possible))
--------------------------------------------------------------------------------
    (?:                      group, but do not capture (optional
                             (matching the most amount possible)):
--------------------------------------------------------------------------------
      &                        '&'
--------------------------------------------------------------------------------
      [ -]+                    any character of: ' ', '-' (1 or more
                               times (matching the most amount
                               possible))
--------------------------------------------------------------------------------
    )?                       end of grouping
--------------------------------------------------------------------------------
    [A-Z]                    any character of: 'A' to 'Z'
--------------------------------------------------------------------------------
    \w+                      word characters (a-z, A-Z, 0-9, _) (1 or
                             more times (matching the most amount
                             possible))
--------------------------------------------------------------------------------
    (?:                      group, but do not capture (optional
                             (matching the most amount possible)):
--------------------------------------------------------------------------------
      \.                       '.'
--------------------------------------------------------------------------------
      co                       'co'
--------------------------------------------------------------------------------
      m?                       'm' (optional (matching the most
                               amount possible))
--------------------------------------------------------------------------------
    )?                       end of grouping
--------------------------------------------------------------------------------
  ){0,2}                   end of grouping
--------------------------------------------------------------------------------
  [,\s]+                   any character of: ',', whitespace (\n, \r,
                           \t, \f, and " ") (1 or more times
                           (matching the most amount possible))
--------------------------------------------------------------------------------
  (?i:                     group, but do not capture (case-
                           insensitive) (with ^ and $ matching
                           normally) (with . not matching \n)
                           (matching whitespace and # normally):
--------------------------------------------------------------------------------
    ltd                      'ltd'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    llc                      'llc'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    inc                      'inc'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    plc                      'plc'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    co                       'co'
--------------------------------------------------------------------------------
    (?:                      group, but do not capture (optional
                             (matching the most amount possible)):
--------------------------------------------------------------------------------
      rp                       'rp'
--------------------------------------------------------------------------------
    )?                       end of grouping
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    group                    'group'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    holding                  'holding'
--------------------------------------------------------------------------------
   |                        OR
--------------------------------------------------------------------------------
    gmbh                     'gmbh'
--------------------------------------------------------------------------------
  )                        end of grouping
--------------------------------------------------------------------------------
  \b                       the boundary between a word char (\w) and
                           something that is not a word char

Python code：

import re

regex = r"\b[A-Z]\w+(?:\.com?)?(?:[ -]+(?:&[ -]+)?[A-Z]\w+(?:\.com?)?){0,2}[,\s]+(?i:ltd|llc|inc|plc|co(?:rp)?|group|holding|gmbh)\b"

test_str = ("I work in Amazon.com Inc.\n"
    "Company named Swiss Medic Holding invested in vaccine\n"
    "what do you think about Abercrombie & Fitch Co. ?\n"
    "do you work in Delta Group?\n"
    "I have worked in CocaCola Gmbh")

print(re.findall(regex, test_str))

结果：['Amazon.com Inc', 'Swiss Medic Holding', 'Abercrombie & Fitch Co', 'Delta Group', 'CocaCola Gmbh']

【讨论】：

似乎无法正常工作，因为它返回非大写单词（即工作、命名、关于...）。
@JoanLaraGanau 不，它工作正常。我提供了代码和结果。
它现在可以正常工作，因为您对其进行了编辑。无论如何，干得好，这是一个非常棘手的正则表达式。编辑：也许你只添加了r"...
但公司名称后面必须有ltd、group、plc、holding等
@taga 正是我的解决方案的作用。在最后找到它们。

【解决方案2】：

使用https://regex101.com 来测试正则表达式，这很棒。对于您的具体示例，这是您想要的正则表达式。在此示例中，我认为不需要测试可选的 .com。

regex_company = '[A-Z]([^ ]*[ &]*){0,2}(Inc\.|Ltd|GmbH|AG|Gmbh|Group|Holding|Co\.)'

for address in [address_1, address_2, address_3, address_4, address_5]:
    found = re.search(regex_company, address)
    if found:
        print(found)

# prints:
# <regex.Match object; span=(10, 25), match='Amazon.com Inc.'>
# <regex.Match object; span=(14, 33), match='Swiss Medic Holding'>
# <regex.Match object; span=(24, 47), match='Abercrombie & Fitch Co.'>
# <regex.Match object; span=(15, 26), match='Delta Group'>
# <regex.Match object; span=(17, 30), match='CocaCola Gmbh'>

【讨论】：

请查看我用于街道检测的正则表达式的类型，您可以为公司名称复制类似的内容吗？