The latest direction of this PR is to avoid two 'options' parameters and have the very FindManyOptions class(es) extend CommandOptions to capture all options.

This required a bit of machinery between the various classes involved, which should be considered as there's the potential for mistakes such as forgetting to add a new CommandOptions field in one of the 'extraction' and clone methods:

• No explicit constructors, so this should be ok and futureproof
• The *FindManyOptions' Clone method now lists all the CmdOpts properties explicitly
• Introduced a .CommandOptions() method that extracts that part, and a .PayloadOptions() method that extracts that other part. These seem necessary to feed the right thing to the right parts of the logic to construct the find Command's
  • well, not strictly necessary maybe but I thuoght it cleaner and centralized.
  • indeed, the [JsonIgnore] had to be added to the CommandOptions members to avoid them leaking into the find payload! (which does not seem to have adverse consequences anywhere). We don't want e.g. the token to leak by mistake or something like that, and creating a specific class for the "find payload options" can always be done later.

I'm completing this PR with the conflation of CommandOptions into InsertManyOptions as well.

• removed some InsertMany[Async] signatures. Now it's only (docs) and (docs,options) (i.e. docs/rows)

• methods' xmldocs made uniform (which tags appear where, etc)

• methods' xmldocs phrasing slightly reworked

• (renamed the static const MaxConcurrency to DefaultConcurrency in the options)

• removed the guard that prevented inserting an empty list of docs/rows

• added test (coll+table) about passing a generic 'command options' setting

(@skedwards88 this is possibly reflected somewhere in the docs?)