Issue/#116 - Removes the fraction part based on the SerializationFeature WRITE_TIMESTAMP_WITHOUT_FRACTION

[sashikanthr]

This fixes #116

We have added a new feature to in SerializationFeature in the databind package to avoid the fraction part. This has a separate pull request.

[sashikanthr]

Since this is updating the databind module, the build will fail. We couldnt find information on how to update changes in multiple modules.

[sashikanthr]

@kupci May I know if we can address your comments and update the PR? Will it be still relevant?

[kupci]

@sashikanthr I think the issue is still relevant, but it might be easier to create a new PR. I think your original PR highlighted the need for a better overall solution for handling issues like these.

One thought (refreshing myself on the issue) was to use the Shape.* but the difficulty there was both milliseconds and epoch seconds would be Shape.NUMBER_INT.

However, on that subject @cowtowncoder describes some initial thoughts / Idea-lets here:

would it make sense to add unit (or timeUnit) property for @jsonformat?

On one hand, it is bit datatype-specific, but there is one big benefit: ability to set format defaults, and overrides (both programmatically and via annotations) would make implementation relatively easy

https://github.com/FasterXML/jackson-future-ideas/wiki/JSTEP-5

One question I have is whether we can simply add more some sub-types, i.e. we already have Shape.NUMBER, Shape.NUMBER_FLOAT and Shape.NUMBER_INT, so could we have the new types following the Java convention, say Shape.NUMBER_INT_MILLISECONDS? And do we need to distinguish between epoch seconds, or would Shape.NUMBER_INT_SECONDS work, following the Java convention here:
https://docs.oracle.com/javase/8/docs/api/java/util/concurrent/TimeUnit.html

updated: See the newly added CoercionConfig - wondering if that can be used as DateTimeConfig, or does this use case not fit?

@cowtowncoder - not sure how to update the JSTEP wiki. On the jackson-databind/wiki there is an 'Edit' button, and a 'New Page', but I don't see that for the corresponding wiki page on 'Future Ideas', specifically the JSTEP-5 page: https://github.com/FasterXML/jackson-future-ideas/wiki/JSTEP-5]

[cowtowncoder]

So, CoercionConfig determines reading-side only (deserialization), on whether to accept alternate, secondary input shapes: like whether boolean property can be read from (JSON) integral number value.
It probably will not be quite as useful here because of this, unfortunately.

As to possible Shape.NUMBER_INT_MILLISECONDS, I don't think that works either unfortunately -- shapes, as a concept, refer to type in input format, and there is no qualifier in JSON Number (or most other dataformats) to really make such distinction.

Finally, on wiki: @kupci should now have access via "wiki" team (you may need to accept it or something, requested your addition) so hopefully can now edit that wiki page.

[kupci]

Thanks @cowtowncoder , that worked: I can get see the "Edit" button on the new-ideas wiki.

[kupci]

Closing this one as per comments above, the direction for changes like this is to implement via a more long-term solution. Details here: https://github.com/FasterXML/jackson-future-ideas/wiki/JSTEP-5