BIND 10 #2512: RR type implementation: CAA

[BIND 10 Development]

#2512: RR type implementation: CAA
-------------------------------------+-------------------------------------
            Reporter:  shane         |                        Owner:
                Type:  enhancement   |  stephen
            Priority:  high          |                       Status:
           Component:  libdns++      |  reviewing
            Keywords:                |                    Milestone:
           Sensitive:  0             |  bind10-1.2-release-freeze
         Sub-Project:  DNS           |                   Resolution:
Estimated Difficulty:  5             |                 CVSS Scoring:
         Total Hours:  0             |              Defect Severity:  N/A
                                     |  Feature Depending on Ticket:
                                     |          Add Hours to Ticket:  0
                                     |                    Internal?:  0
-------------------------------------+-------------------------------------
Changes (by muks):

 * owner:  muks => stephen


Comment:

 Hi Stephen

 Replying to [comment:12 stephen]:
 > '''src/lib/dns/rdata/generic/caa_257.{cc,h}'''
 > All the specialised methods - including the trivial getFlags() etc -
 > should have Doxygen headers. (These are usually in the .h file, but
 > it's not worth changing it now.)

 These are added now.

 > The documentation for getValue() should note that the reference
 > returned is only valid for as long as the relevant CAA object is in
 > existence.

 Done.

 > CAA::CAA(!InputBuffer&, size_t): tag_vec is created then immediately
 > resized to tag_length.  Suggest that the vector be created with the
 > correct size (by passing tag_length as the constructor argument).

 Done.

 > '''src/lib/dns/rdata/generic/detail/char_string.{cc,h}'''
 > stringToCharStringData: I would prefix the assertion with a
 > comment. The fact that there are three digits following an escape is
 > checked in decimalToNumber(), so the only reason an assert() is needed
 > is to catch the case where the check is removed from that
 > function. (Also, it took a few seconds to work out why the assert was
 > for n >=3 when the following line subtracts 2 from n.)

 Comments were added for these cases.

 > '''src/lib/dns/tests/rdata_char_string_data_unittest.cc'''
 > charStringDataToString test: rather than use an assert(), why not an
 > ASSERT_GE? That way if the assertion does fail, the test fails in a
 > graceful way rather than the the test program terminate.

 Done.

-- 
Ticket URL: <https://bind10.isc.org/ticket/2512#comment:15>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development