正则表达式：包含模式的字符数

Question

I have a dataframe with the following structure:

Desc_ORF	ORF
beta-glucosidase	tb512
succinate-semialdehyde dehydrogenase	tb111
probable epoxide hydrolase	tb045

I am using this function to filter the dataframe:

df.set_index('Desc_ORF').filter(regex=pattern, axis=0)

It is working perfectly fine with other patterns that I'am trying, but I cannot obtain a regex pattern that filters the rows where the Desc_ORF contains hydro into a word of 13 characters.

For example: My code should keep the row succinate-semialdehyde dehydrogenase because it contains dehydrogenase that has 13 characters and contains the pattern hydro . On the other hand, the filter must discard probable epoxide hydrolase because, although it contains hydro , the word hydrolase is not of 13 characters.

Desc_ORF	ORF
succinate-semialdehyde dehydrogenase	tb111

I have tried different patterns and my last try has been: ^(?={13}$)(\b\S hydro\S \b). With this pattern I am only filtering by words that contain hydro, but I cannot obtain words that contain hydro with a length of 13 characters.

Answer 1

One option to match the word in the second line could be:

(?<!\S)(?=\S{13}(?!\S))\S*hydro\S*

• (?<!\S) Assert a whitespace bounadary on the left
• (?=\S{13}(?!\S)) Assert 13 non whitespace chars followed by a whitespace boundary
• \S*hydro\S* Match hydro between optional non whitespace chars

Regex demo

pattern=r"(?<!\S)(?=\S{13}(?!\S))\S*hydro\S*"
df = df.set_index('Desc_ORF').filter(regex=pattern, axis=0)