fix(datepicker): Datepicker only accept text that matches format #3152
Conversation
Added parseStrict param in parseDate function, which preserves use of new Date() fallback in that function, but allows developer to set parsing to strictly conform to dateFormat, otherwise returns undefined. fixes issue #2886
|
HI. I just found this PR. Few days ago I prepared mine, very similar one (#3294). But I haven't added additional parameter strict-date. For now I would say it's redundant one because is there really a situation when we have some date format defined (and actually we always have, don't we?) but don't want date to be scrict-parsed? I just can't imagine such requirement. |
|
@clockworked247 - This fails in the travis build, can you update the PR? |
|
@karianna - Sure. Will do. |
|
@jmayday - I believe the changes are quite similar. I only added the param to preserve existing functionality, so if you think it's not necessary, I'm all for your change. Looking at the code, you're removing the fallback and relying on dateParser.parse(). That code appears to effectively do the same thing in returning undefined if the date is not valid. So that looks good to me. |
|
@clockworked247 The build is still failing. Can you please fix this PR? |
|
@rvanbaalen My understanding was that my PR was a duplicate of #3294, so I'm OK to just close this and let the issue be hashed out there. It seems it's under debate though. I'm not really following why #3417 conflicts with the change. Considering we need to validate a date entered manually in the associated text field, it only makes sense to add this kind of change. Otherwise, the Date() fallback will just change the invalid date to some odd number that has nothing to do with what the user entered. Super confusing... We also wouldn't be able to do custom handling of invalid dates without it returning undefined for invalid dates. Closing this, as it's a duplicate. |
|
Thanks for detecting the duplicate @clockworked247 👍 |
fixes issue #2886