BIND 10 #2383: add dns::Name constructor to be usable from MasterLoader

[BIND 10 Development]

#2383: add dns::Name constructor to be usable from MasterLoader
-------------------------------------+-------------------------------------
                   Reporter:         |                 Owner:  vorner
  jinmei                             |                Status:  reviewing
                       Type:  task   |             Milestone:
                   Priority:         |  Sprint-20121106
  medium                             |            Resolution:
                  Component:         |             Sensitive:  0
  libdns++                           |           Sub-Project:  DNS
                   Keywords:         |  Estimated Difficulty:  4
            Defect Severity:  N/A    |           Total Hours:  0
Feature Depending on Ticket:         |
  loadzone-ng                        |
        Add Hours to Ticket:  0      |
                  Internal?:  0      |
-------------------------------------+-------------------------------------
Changes (by jelte):

 * owner:  jelte => vorner


Comment:

 Replying to [comment:8 vorner]:
 > Hello
 >
 > Replying to [comment:5 jelte]:
 > > - ft_at in name.cc:130 has a comment that it is not used, but it is
 used (at least in the code below)
 >
 > Well, it is checked to see if the current state is set to the value. But
 it is
 > never set to in in the code, so it's effectively unused. I removed it
 (after
 > checking the bind9 code and seeing there's no deep stuff hidden behind
 it). We
 > don't need it, we have the `is_root` variable.
 >

 ok that looks good

 > > - regarding @, should input like '@foo' be accepted (how lenient do we
 want to be)? and if not, do we fail here or in the lexer? (looks like the
 current implementation would drop all text after the @-sign, if it gets
 here), this may also be related to requirements section 3.7.3 (soa rname
 with @)
 >
 > I don't see a reason why it should not be accepted. And our
 implementation
 > actually doesn't drop it, it (IMO correctly) takes @ as any other
 character if
 > there's something other as well. I added some tests to verify it.
 >

 ah, yes, my initial experiment was wrong then :) in that case behaviour is
 good

 > I don't see how it relates to the SOA with `@` as RNAME). But I think
 the Name
 > class is too low-level to care about reporting warnings to users, this
 should
 > be done at a higher level.
 >

 in the original rfc, rules were actually quite strict on what characters
 were allowed in the first place (almost none). This has been massively
 ignored by every implementer ever :) I do believe some implementation
 automatically replace the @ in RNAMEs, but as said, this should at the
 lowest be done on the SOA level, not the Name.

 > > - wild idea (and totally offtopic, should maybe have though of this
 earlier): perhaps we'll want to warn if the name is not absolute, but does
 appear to end with the origin. This is like the second most often made
 mistake in editing zone files (forgetting the ., the first being not
 updating soa serial). Of course, apart from whether we want to do this in
 the first place, name.cc is probably not the place to do so.
 >
 > The same as above.

 yes just mentioning it since it is related to the subject

 OK all good, go ahead and merge.

-- 
Ticket URL: <http://bind10.isc.org/ticket/2383#comment:9>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development