[augeas-devel] Re: [Augeas] #56: last()+1 doesn't behave very logically

Augeas trac at fedorahosted.org
Wed Apr 1 18:12:41 UTC 2009


#56: last()+1 doesn't behave very logically
----------------------+-----------------------------------------------------
  Reporter:  raphink  |       Owner:  lutter
      Type:  defect   |      Status:  closed
  Priority:  minor    |   Milestone:  next  
 Component:  Augeas   |     Version:  devel 
Resolution:  wontfix  |    Keywords:        
----------------------+-----------------------------------------------------
Changes (by lutter):

  * status:  new => closed
  * resolution:  => wontfix

Comment:

 The problem is that both behaviors are reasonable - the path expression
 simply doesn't have enough enough information to control where a missing
 node should be created; I therefore think that always creating them as the
 last child of their parent is at least consistent.

 As you note, you need to use aug_insert if you need more control over
 where nodes should be created - in general, creating from just a path
 expression isn't well-defined, anyway; for example, what should happen for
 "set /a/b/c[../d = 'foo'] bar" if there is no d sibling ?

 I agree that it would be nice to embed instructions on what to create when
 a path expression does not match; but I have no good ideas on how to
 address that

-- 
Ticket URL: <https://fedorahosted.org/augeas/ticket/56#comment:1>
Augeas <http://augeas.net/>
a configuration API




More information about the augeas-devel mailing list