I am attempting to implement a conditional statement within regex, applied via the pandas.Series.str.extractall method. Given the reading I've done here, this seems like a pretty easy problem to solve, but I am still getting stuck...

I have the following regex in the Pythex tester:

(a)(?(1)b|c)

As I understand it, (a) is my first test group. The conditional block (?(1)b|c) should attempt to match "b" if my first test group is a match, or else it will attempt to match "c". The results I am hoping for are as follows:

1. "b" = No Match
2. "ab" = Match
3. "c" = Match
4. "ac" = No Match

The (a)(?(1)b|c) statement achieves 1, 2, and 4, but it misses 3... Any tips?

Thank you!

CodePudding user response：

I am not seeing how your code matches up with the docs. It supplies id 1, fine.

Will try to match with yes-pattern if the group with given id ... exists, ...

For example, (<)?(\w @\w (?:\.\w ) )(?(1)>|$)

Notice how the example's group 1 is optional, it matches or it doesn't and then we move on for more matching.

In your (a) expression the "a" is non-optional. I would expect that group 1 always exists at the point that we're evaluating the conditional. Which makes it not especially conditional.

Start with (a)? to improve matters.

CodePudding user response：

To get the matches, you don't need a conditional.

If a, then also match b.. else match c, can be written as:

\b(?:ab|c)\b

Regex demo