use count for +, - in Species and remove tryparse

[mattsignorelli]

No description provided.

[DavidSagan]

Also make sure that something like "He+-" is caught.

[rot4te]

Working on it, though I'm more certain of how to catch "+-" than how to use count; since the name needs to catch charge markers eg +, ++, +3, +4, etc. tryparse is good because it just converts the string to the int.
I'm thinking about adding a reference statement to the species construnctor where it tells you the string name of what it's building.

[DavidSagan]

Try this: After the first + or - is encountered, check whether the next character is a +, -, digit, or none of the above. Then you can do the right thing (notice that something like +79 has more than one digit. count is not good in this case since you need to make sure that multiple + or - are consecutive.

[mattsignorelli]

Why do the +- need to be consecutive?

[DavidSagan]

Likely a typo then. Nobody writes He+-+.

[mattsignorelli]

Just because nobody writes it like that doesn't mean the code needs to impose such a restriction, especially if it lets us use count instead of having to write our own code.

[DavidSagan]

Catching typos is important. And parsing the species string is so simple I don't see the loss of count as of any importance.

[DavidSagan]

Also allowing a non-standard syntax can possibly hinder portability with other codes.

[DavidSagan]

Actually you can use count. Just make sure that + and - are not both present.

[dtabell]

I agree wholeheartedly with the importance of catching typos! Having worked on a code (which shall remain nameless) that sometimes failed to catch even the simplest human “errors”, and thereby made it possible for many scientists (including me) to waste *days* of precious time, I firmly believe in limiting the amount of rope we give our users. They’ll find plenty of ways to hang themselves, but let’s not make it easy. Dan

…

On Sat, Oct 5, 2024 at 05:51 David Sagan ***@***.***> wrote: Actually you can use count. Just make sure that + and - are not both present. — Reply to this email directly, view it on GitHub <#64 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACXTUHCLIR6YRCXRPI7XOV3ZZ7HD7AVCNFSM6AAAAABO3QYNC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJVGAZTAOBRGY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.*** com>

[rot4te]

See PR #126