正则短语，正数或负数/带单位的十进制

Question

我正试图得到一个字符串中的数字。然而，这个数字在句子中必须遵循一个非常具体的模式。我们可以假装我在寻找包裹的最大重量。例如，The 3rd box marked 4 cannot exceed 7.34kg but if its exact 8.42kg its okay too。在这种情况下，我想要匹配数字7.34。我的英语模式是：

<Starts with phrase which is in ('must not exceed', 'cannot exceed', 'limited by')><Can be any characters of any length (0 length to infinite) as long as its not any alphabet characters><a positive or negative int or decimal which may or may not be comma separated (ie. 1,032.43kg)><Can be any characters of any length (0 length to infinite) as long as its not any alphabet characters><ends with the characters which are in ('kg', 'k.g', 'k/g')

我拥有的是：

(must not exceed|cannot exceed|limited by).*?[0-9]+ ?(kg|k\.g|k\/g)

但是，我不能做的主要事情是能够匹配<Can be any characters of any length (0 length to infinite) as long as its not any alphabet characters>和`

一些例子

The 3 boxes with red on them must not exceed -23.4435kg and don't pick them up.
Parcels that can be sent are limited by 1,402kg and its okay to send
The 2 boxes on the shelf must not exceed: 
102 kg
Do not pick up 18 boxes at a time and make sure they cannot exceed,: 92302 k.g
Do not pick up boxes that weight 56.23 kg
Boxes cannot exceed -23 k/g

我知道我可能需要双倍匹配。因此，我首先对句子进行匹配(即must not exceed -23.4435kg然后regex匹配数字，这就是我目前在代码中所做的。)我的问题本质上是如何与字符串的正确部分匹配。

Answer 1

我建议用

\b((?:must\s+|can)not\s+exceed|limited\s+by)\W*?(-?\d+(?:,\d+)*(?:\.\d+)?)\s*(kg|k\.g|k/g)\b

见regex演示。详细信息

• \b -一个单词边界
• ((?:must\s+|can)not\s+exceed|limited\s+by) -第1组：must not exceed，cannot exceed或limited by，在单词之间有空格
• \W*? -任何零或更多，但尽可能少，非字字符
• (-?\d+(?:,\d+)*(?:\.\d+)?) -第2组:数字模式，一个可选的-，然后是一个或多个数字，然后是零或多个,序列和一个或多个数字，然后是可选的.和一个或多个数字。
• \s* -零或多个空白空间
• (kg|k\.g|k/g) -第3组：kg，k.g还是k/g
• \b -一个单词边界

见Python演示

import re
texts = ['The 3rd box marked 4 cannot exceed 7.34kg but if its exact 8.42kg its okay too',
    'The 3 boxes with red on them must not exceed -23.4435kg and don\'t pick them up.',
    'Parcels that can be sent are limited by 1,402kg and its okay to send',
    'The 2 boxes on the shelf must not exceed: \n102 kg',
    'Do not pick up 18 boxes at a time and make sure they cannot exceed,: 92302 k.g',
    'Do not pick up boxes that weight 56.23 kg',
    'Boxes cannot exceed -23 k/g',
    'must not exceed -23.4435kg']

rx = re.compile(r'\b((?:must\s+|can)not\s+exceed|limited\s+by)\W*?(-?\d+(?:,\d+)*(?:\.\d+)?)\s*(kg|k\.g|k/g)\b')
for text in texts:
    print("----", text,"----")
    m = rx.search(text)
    if m:
        print(f"Phrase: {m.group(1)}")
        print(f"Number: {m.group(2)}")
        print(f"UOM: {m.group(3)}")
    else:
        print("Not matched!")

输出：

---- The 3rd box marked 4 cannot exceed 7.34kg but if its exact 8.42kg its okay too ----
Phrase: cannot exceed
Number: 7.34
UOM: kg
---- The 3 boxes with red on them must not exceed -23.4435kg and don't pick them up. ----
Phrase: must not exceed
Number: -23.4435
UOM: kg
---- Parcels that can be sent are limited by 1,402kg and its okay to send ----
Phrase: limited by
Number: 1,402
UOM: kg
---- The 2 boxes on the shelf must not exceed: 
102 kg ----
Phrase: must not exceed
Number: 102
UOM: kg
---- Do not pick up 18 boxes at a time and make sure they cannot exceed,: 92302 k.g ----
Phrase: cannot exceed
Number: 92302
UOM: k.g
---- Do not pick up boxes that weight 56.23 kg ----
Not matched!
---- Boxes cannot exceed -23 k/g ----
Phrase: cannot exceed
Number: -23
UOM: k/g
---- must not exceed -23.4435kg ----
Phrase: must not exceed
Number: -23.4435
UOM: kg

在Oracle中，您需要丢弃单词边界，并将所有非捕获组替换为捕获组：

 REGEXP_SUBSTR(
   col,
   '((must\s+|can)not\s+exceed|limited\s+by)\W*?(-?\d+(,\d+)*(\.\d+)?)\s*(kg|k\.g|k/g)',
   1,1, NULL, 3)