Canonicalize and ensure uniqueness of indices

[tcsalameh]

Prior to this commit, adding a entryUUID or entryCSN
index conflicted with those being added by default
when the syncprov overlay is enabled.

This avoids that behavior by canonicalizing and uniquing
the indices before adding them into OpenLDAP.

[bodgit]

I'm trying to understand this change.

1. Do you have a test case or example for this? What other indices do you need for entryUUID/entryCSN?
2. Your new function appears to be something supported by Puppet 4 only. I don't want to drop Puppet 3 support; writing this in terms of a parser function would work on both.

There will also need to be tests written for this.

[tcsalameh]

Hi Matt,

Thanks for responding so quickly - I'm happy to clarify. I've cc'd Jesse
Hathaway, who I'm working with.

1. The idea is to make the behavior more lenient in the case where someone
   specifies indices that are already included when syncprov is set to true.
   The example would be a user who sets syncprov => true and also adds indices
   for entryUUID and entryCSN up front. These indices are added again in
   config.pp, which causes LDAP to reject the duplicate indices and the puppet
   apply to fail. It seems more user-friendly to just reformat and unique the
   index entries before adding them to LDAP, rather than completely fail in
   the case that someone accidentally adds duplicate indices.
2. Good point - I've rewritten it as a parser function.

We had some trouble getting the tests to run yesterday, but got them
working today. I added several test cases for the new function, and had to
modify a few existing test cases to account for the different format for
the olcDbIndex attributes. I've amended the pull request to include these
changes.

Best,
Tom Christiansen-Salameh

On Wed, Jul 20, 2016 at 4:49 PM, Matt Dainty notifications@github.com
wrote:

I'm trying to understand this change.

1. Do you have a test case or example for this? What other indices do
   you need for entryUUID/entryCSN?
2. Your new function appears to be something supported by Puppet 4
   only. I don't want to drop Puppet 3 support; writing this in terms of a
   parser function would work on both.

There will also need to be tests written for this.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#28 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AIXiUN-V1gUa0UfGWWI_w-qHKsVdkqm7ks5qXpfigaJpZM4JQ3Sm
.

[coveralls]

Coverage increased (+0.7%) to 53.527% when pulling 9a7bddd on tcsalameh:unique-indices into 9a14f51 on bodgit:master.

[coveralls]

Coverage increased (+0.7%) to 53.527% when pulling 91f7c71 on tcsalameh:unique-indices into 9a14f51 on bodgit:master.

[bodgit]

Ok, the change looks good to me. Tests all pass. 👍

Thanks for your contribution!

[tcsalameh]

Thank you! We're using this module as part of an OpenLDAP implementation project, so I may be submitting a couple more pull requests in the next few weeks.

[bodgit]

Thank you! We're using this module as part of an OpenLDAP implementation project, so I may be submitting a couple more pull requests in the next few weeks.

No problem. I've pushed v1.1.4 to the forge.