285: Pass interval id to extensions

[j6s]

Passes the id of an interval to extensions. Now, stdin of an extension looks like this:

{"id":99,"start":"20191219T183718Z","end":"20191219T185724Z","annotation":"project:gaming"},
{"id":98,"start":"20191219T211035Z","end":"20191219T220034Z","tags":["project:gaming"]},
{"id":97,"start":"20200107T090000Z","end":"20200107T113235Z","tags":["tempo:ABC-477"]},
{"id":96,"start":"20200107T125607Z","end":"20200107T152439Z","tags":["tempo:ABC-477"]},
{"id":95,"start":"20200107T165000Z","end":"20200107T180803Z","tags":["tempo:ABC-477"]},
{"id":94,"start":"20200108T081022Z","end":"20200108T083142Z","tags":["tempo:ABC-477"]},

Resolves #285

[sruffell]

The only other thought I have is that I wonder if this should only be enabled when the :ids hint is specified on the command line? This would be consistent with timew summary but most consumers of the JSON should probably only be looking for specific keys and would ignore the id tag by default.

[sruffell]

Here too you'll want to add curly braces.

[lauft]

I know that there rarely is an interval without a start, but I would prefer to keep lines 109 - 110 for consistency

[j6s]

As I understand lines 109-110 were adding a comma in front only if the start existed because if a start existed, the start would be the first item in the json object but if a start did not exist, the end time would be the first item in the object.

With this change, there is always a defined first item in the object: the id.

[lauft]

@j6s you are definitely right here 😅 - however I found that there is the case of an empty interval which then gets serialized as {"id":0} instead of {}.
And also tests dom.t and interval.t are broken. They appear green on Travis (should be fixed with dcfe0d1c), but if you inspect the individual runs you see the error reports (e.g. https://travis-ci.org/GothenburgBitFactory/timewarrior/jobs/651153570)

[lauft]

Empty interval case needs to be addressed
Please check/adapt tests dom.t and interval.t

[lauft]

I discovered that this feature comes handy for tests, also in context of #294. I will merge it and handle the open issues mentioned in the review. Thanks @j6s nevertheless for this contribution!

[j6s]

Sorry for not finishing this - I had genuinely planned to have a look at it on the weekend.

[lauft]

@j6s Don't be. You can take a look at the documentation and see whether (or better where) this needs to be updated. And, also important, consider to publish your addon on github/... and tell us about it so we can mention it to other users 😉