[Tim Wood]

We're possibly going to use this in our fledgling web-thing, and others (dansays, QS, etc.) might find this useful...

As of earlier today there is a 'complete' script command in OmniFocus. This uses exactly the same logic as in the completion cells but presents it in a script-y interface.

Say you have two contexts, "abc" and "car", you can now do:

tell application "OmniFocus"
	tell default document
		complete "a" as context
	end tell
end tell

and get:

{{xml:"<span class=\"match\">a</span>bc", score:15, id:"bxMoRcqDm0g", name:"abc"},
{xml:"c<span class=\"match\">a</span>r", score:0, id:"krLrtHo7Pbe", name:"car"}}

That is, you get a score-sorted array of match records, each with:

• id: the identifier of the matching object
• name: the name of the matching object (currently the full hierarchical name, but we could add both this and the local name)
• score: the numerical score of the match; bigger is better. The result is already sorted, but this might be useful. It isn't scaled to any max value, so a relevance indicator would be hard to do.
• xml: a little XML blob of the 'name' marked up with <span> tags. The default span class name is "match". You can pass an optional 'span class' argument to the command to change this to whatever you want.

Hopefully you all will find this useful; let us know if there are other scripting hooks that'd be helpful too.