Error parsing user metadata

[SimonMolinsky]

Problem

There is a problem with parsing contributors' metadata in package reviews. The problem is probably related to this line of the code:

    def _get_line_meta(self, line_item: list[str]) -> dict[str, object]:
        ...
        if self._contains_keyword(line_item[0]):
            if line_item[0].startswith("All current maintainers"):
                ...
            else:
                names = line_item[1].split("(", 1)  # THIS LINE HERE!
               ...
        return meta

It probably leads to the parsing error here: https://github.com/pyOpenSci/pyopensci.github.io/actions/runs/6521128026/job/17709405005

Reason

I think splitting the line by round bracket ) is why the error occurs. One reviewer's name here: pyOpenSci/software-submission#115 has this character.

Solution

There are two options:

1. Restrict characters in names (update Editors Guide),
2. Change the parsing method to detect the @ character.

I can do both, so you can assign me to this issue 🚀

[cmarmo]

@SimonMolinsky I'm having the same issue when updating the website for astartes. It would be wonderful if you can fix it!
As far as I'm concerned I'd rather have the @ detected.

[lwasser]

@SimonMolinsky if you can work on this i'd greatly appreciate it!! i must have broken the build with a recent update!! i'll assign you now.

[lwasser]

also agree with @cmarmo on - let's parse for the @. i haven't look at that code in a bit of time but i DO recall i used the ghusername as a unique identifier in the dictionaries (i'm pretty sure each dict contains a pydantic "personmodel" that i created and validates accordingly. But it is the case that sometimes in our issues, people add their names in ways that i didn't expect. then it breaks the build.

i'm wondering if that is the case here. but regardless appreciate your looking into this - thank you!!

[SimonMolinsky]

@lwasser It was relatively easy to dive into your code. It is self-explanatory. I will work on it then. Thanks for the assignment!

@cmarmo I will check your case too :)

[lwasser]

ok this has been merged!! thank you BOTH for being invested in fixing this. i'm going to see if the new packages add to our site now. And then followup with a small cleanup fix so there is no markdown in package names!!

i'm relieved to hear the code here was not too bad to navigate! i have a tendency to over complicate workflows altho i can say pydantic helped me really clean things up here!!!