-
Notifications
You must be signed in to change notification settings - Fork 154
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cookbook: Parsing localized timestamps #2004
Comments
We have mostly considered parsing (other than deserialization of ISO 8601-like strings) to be in the domain of business logic and outside the scope of Temporal (since everybody knows their own use case for parsing other formats better than Temporal possibly can, and to avoid repeating the mistakes of Here's some discussion about how it might be supported in the future: js-temporal/proposal-temporal-v2#2 What I'd recommend depends on your use case, and how robust it needs to be. I think for most cases I'd recommend parsing by splitting strings or using regular expressions, tailored to the format that you are parsing, and then constructing one of the wall-clock-aware objects (e.g. PlainDate, PlainDateTime) using number values that you obtain from the string. Then you can convert that to Instant using the appropriate time zone, to obtain an exact time. By the way, I'm assuming that once Temporal is more widely adopted, libraries will start appearing that will provide this functionality. I hope that answers your question! |
Thanks for answering. I understand your arguments. Let me just explain why I think that the domain of business logic for me is not a good place: I'd prefer a world with ISO strings, UTC, unified calendars and without daylight savings time but unfortunately that is not in my choice ;-) (the issue could be closed from my side) |
I see, so your question is not about users inputting data in one locale-specific format, but in any locale-specific format? That might be more appropriate for the I think it might be nice to document in the Temporal cookbook how to parse a particular expected format as I described above, so maybe it's a good idea to keep this issue open in order to do that. |
Thanks for that link, very interesting. I can only underline the points made in this comment tc39/ecma402#342 (comment) but I also understand arguments agains string parsing. I think it very much depends on the target user group. In our case it's a limited group of international power users. These guys work with timestamps and measured values all day. Selecting a specific date and time with one of the common date and time pickers needs up to 10 clicks (at least when it's located in another year and month)! When you have to do that 30 times a day you start getting mad at the software developer. And that's exactly where the strict parser that reverses the output of some format function comes into play. These power users are happy to stick to a given (localized) format an accept an error message now and then when they have mistyped the timestamp string. I could force them to use ISO format but then they would complain "Why do I have to use that strange format? My local date format works in Excel!!!". I agree with others that this usage may probably not be a good idea for public web sites for everyone. It would just be nice if the develop has a choice. And the idea that every developer (usually not an expert in Date parsing) comes up with her own solution sounds terrible to me. OK, let's keep the issue open. Am curious to see how it turns out. Apart from that working with Temporal is such a great joy! |
What is the general strategy for parsing timestamps that a user entered in localized format (non ISO 8601)?
F.e. in a German locale a user may enter '22.01.2022 13:45:00'.
Assuming the software knows the time zone, the calendar and the locale, I would like to end up with a (UTC) Instance for internal arithmetics.
What do you recommend for parsing user input? Is there already any documentation about this?
The text was updated successfully, but these errors were encountered: