Enhance docs of Statement constructor and add alternate constructors

[JeroenDeDauw]

No description provided.

[JeroenDeDauw]

@Benestar this is a follow up to the work you did. I've found that there are many places where you want to construct a Statement given just its claims main snak, since all else is optional. So a new shortcut constructor method has been added for this.

[JeroenDeDauw]

@Benestar Your changes to the constructor are also not really reflected in things such as StatementList::addNewStatement. I'd hate to also go change these - they've just been added and are good what they where meant for.

[Benestar]

This method has the same signature as the constructure and is not helpful in my opinion.

[JeroenDeDauw]

This is a named constructor. Also, one can use this now, without getting any IDE warnings due to parameter mismatch.

[Benestar]

@Benestar this is a follow up to the work you did. I've found that there are many places where you want to construct a Statement given just its claims main snak, since all else is optional. So a new shortcut constructor method has been added for this.

Yeah, I think moving this into a static method is the right way.

@Benestar Your changes to the constructor are also not really reflected in things such as StatementList::addNewStatement. I'd hate to also go change these - they've just been added and are good what they where meant for.

I'm totally fine with keeping those shortcut methods as they are. However, might there also be situations where you have a Claim and want to create a new Statement in the StatementList? This would become a bit nasty but perhaps those situations are quite rare anyway.

[thiemowmde]

This syntax means that all parameters are optional now. But the first should not be optional, as it was before.

[JeroenDeDauw]

The first parameter is not necessarily a claim. So putting that in there is deceptive.

[thiemowmde]

So? $snakOrClaim. The first parameter was and still is required, right? Making it optional does not help.

[JeroenDeDauw]

It is not optional. Though an explicit check for it would indeed be good

[Benestar]

Since the next release is 2.0 anyways we could also break the constructor signature as it is quite easy to adopt the new form.

[JeroenDeDauw]

@Benestar if someone can already make the needed changes in WB.git, then let's do that yes.