Python regex not working when all expressions work on regex101.com when trying to turn 8 " W -&-CodePudding

I have some measurements of objects using W, L, H an D for width, length, etc. I want to move inch and feet symbols (", ') immediately next to the number and move those letters immediately next to the inch or feet symbols:

Example input:

09 ' 09 " 09 ' 09 "Test 09 " W a 09" W a

09 ' 09" 09 ' 09 "Test 09 " L a 09" w a

09 ' 09 " 09 ' 09 "Test 09 " L a 09" h a

09 ' 09 " 09 ' 09 "Test 09 " l a 09" d a

09 ' 09 " 09 ' 09 "Test 09 "word

Desired output:

09' 09" 09' 09" Test 09"W a 09"W a

09' 09" 09' 09" Test 09"L a 09"w a

09' 09" 09' 09" Test 09"L a 09"h a

09' 09" 09' 09" Test 09"l a 09"d a

09' 09" 09' 09" Test 09" word

I have tried multiple variations of this expression, every single one of which worked on regex101.com but not on actual Python:

Pattern: r'(?<=[0-9])( )?("|')((( )?([wWlLhHdD])( ))|( )?)(.)?'

Replacement: \2\6 \9

Actual input:

3/8 "W x 20" H -8 "Cortador de haste para madeira

Actual output:

3/8"W X 20"H -8 "Cortador de Haste p/ Madeira

Desired output:

3/8"W X 20"H -8" Cortador de Haste p/ Madeira

Notice how it almost works except that the very last " is not separated from the word and brought together with the number. What am I doing wrong? Since I am learning regex, I am as interested in my error lies as in an alternative, possibly more efficient solution.

Code:

new_title = re.sub(regex_prefix   keyword_to_replace, keyword_to_replace_with, new_title, re.IGNORECASE)

CodePudding user response：

I am not sure which exact command you tried, but I would likely use the following:

# input
s='3/8 "W x 20" H -8 "Cortador de haste para madeira'

import re

re.sub(r'(?<=\d)\s*([\'"])\s*([wWlLhHdD]\b)?\s*', r'\1\2 ', s)

Output: '3/8"W X 20"H -8" Cortador de Haste p/ Madeira'