Deeper matches should win

[mcdonnelldean]

@mcollina The following test is failing, with the smaller patter being matched:

test('deeper matches should win - #', function (t) {
  t.plan(1)

  var instance = bloomrun()

  instance.add({ role: 'auth', cmd: 'login' }, 42)
  instance.add({ role: 'auth', cmd: 'login', retry: 'true' }, 24)

  t.equal(instance.lookup({role: 'auth', cmd: 'login', retry: 'true'}), 24)
})

If you reverse the patterns the test passes. This is because the code in deep match is only validating the params on the pattern. If a smaller pattern fully matches a larger pattern and it was inserted first it wins because our check if for full match on pattern. See https://github.com/mcollina/bloomrun/blob/master/lib/deepMatch.js#L16-L18

Any suggestions? The only way I can see right now is to sort the matching buckets.data by length, that way we are always looking at the larger matches first. This of course would require a major bump because we are going against insertion order.

Any other suggestions?

[mcollina]

I'd say put an option on the constructor, and make it non default to havr the resorting on add.

What do you think? Il giorno mer 18 nov 2015 alle 11:29 Dean McDonnell < notifications@github.com> ha scritto:

@mcollina https://github.com/mcollina The following test is failing, with the smaller patter being matched:

test('deeper matches should win - #', function (t) { t.plan(1)

var instance = bloomrun()

instance.add({ role: 'auth', cmd: 'login' }, 42) instance.add({ role: 'auth', cmd: 'login', retry: 'true' }, 24)

t.equal(instance.lookup({role: 'auth', cmd: 'login', retry: 'true'}), 24) })

If you reverse the patterns the test passes. This is because the code in deep match is only validating the params on the pattern. If a smaller pattern fully matches a larger pattern and it was inserted first it wins because our check if for full match on pattern. See https://github.com/mcollina/bloomrun/blob/master/lib/deepMatch.js#L16-L18

Any suggestions? The only way I can see right now is to sort the matching buckets.data by length, that way we are always looking at the larger matches first. This of course would require a major bump because we are going against insertion order.

Any other suggestions?

— Reply to this email directly or view it on GitHub https://github.com/mcollina/bloomrun/issues/23.

[mcdonnelldean]

Yup, I'm happy with that. I'll get it done tonight.