正则表达式从 pandas 列中的字符串中提取数据

Question

help me to write a regex to process the string in rawDim to extract the height, width and the depth (as float64 integers).

Bonus: Is there a single regex for all 5 examples?

import pandas as pd
dim_df = pd.read_csv("dim_df_correct.csv")
dim_df

    rawDim          height  width   depth
0   19×52cm         19.0    52.0    NaN
1   50 x 66,4 cm    50.0    66.4    NaN
2   168.9 x 274.3 x 3.8 cm (66 1/2 x 108 x 1 1/2 in.)   168.9   274.3   3.8
3   Sheet: 16 1/4 × 12 1/4 in. (41.3 × 31.1 cm) Im...   35.6    25.1    NaN
4   5 by 5in    12.7    12.7    NaN


import re
import pandas as pd

Answer 1

You can use '(?P<height>[\d.]+)\s*(?:[x×]|by)\s*(?P<width>[\d.]+)\s*(?:[x×]\s*(?P<depth>[\d.]+))?' :

df[['rawDim']].join(
   df['rawDim'].str.replace(r'(\d+),', r'\1.', regex=True)
   .str.extract(r'(?P<height>[\d.]+)\s*(?:[x×]|by)\s*(?P<width>[\d.]+)\s*(?:[x×]\s*(?P<depth>[\d.]+))?')
   .astype(float)
   )

output:

                                              rawDim  height  width  depth
0                                            19×52cm    19.0   52.0    NaN
1                                       50 x 66,4 cm    50.0   66.4    NaN
2  168.9 x 274.3 x 3.8 cm (66 1/2 x 108 x 1 1/2 in.)   168.9  274.3    3.8
3  Sheet: 16 1/4 × 12 1/4 in. (41.3 × 31.1 cm) Im...     4.0   12.0    NaN
4                                           5 by 5in     5.0    5.0    NaN

NB. add \s*cm\b to ensure cm only