BIND 10 #2853: Python wrapper of data source extensions

[BIND 10 Development]

#2853: Python wrapper of data source extensions
-------------------------------------+-------------------------------------
            Reporter:  jinmei        |                        Owner:
                Type:  task          |  UnAssigned
            Priority:  medium        |                       Status:
           Component:  data source   |  reviewing
            Keywords:                |                    Milestone:
           Sensitive:  0             |  Sprint-20130611
         Sub-Project:  DNS           |                   Resolution:
Estimated Difficulty:  5             |                 CVSS Scoring:
         Total Hours:  0             |              Defect Severity:  N/A
                                     |  Feature Depending on Ticket:
                                     |  shared memory data source
                                     |          Add Hours to Ticket:  0
                                     |                    Internal?:  0
-------------------------------------+-------------------------------------

Comment (by jinmei):

 You should at least need b3befd1.

 In my case the original code indeed segfaulted, and b3befd1 fixed it.
 Even if not for you, this fix itself is necessary, and, it seems to
 reveal a couple of meta issues:

 - I suspect you should have seen something like this:
 {{{
 ERROR: test_find_mapped (__main__.ClientListTest)
 ----------------------------------------------------------------------
 Traceback (most recent call last):
   File
 "/Users/jinmei/src/isc/git/bind10-2853/src/lib/python/isc/datasrc/tests/clientlist_test0.py",
 line 61, in test_find_mapped
     result[1].load()
 SystemError: error return without exception set
 }}}
   which should have suggested the source cause of the issue.
 - This type of programming errors should normally be caught by
   compilers with our default warning settings, and I then noticed
   its Makefile is missing these settings.

 For the second issue, the attached patch should recover the expected
 settings.  But there seem to be other warnings in the wrapper library,
 so I'm not committing it right now.  They are probably easy to fix,
 in which case we should probably fix them in this task.  We should
 also check other Makefiles (at least for Python wrappers) to see if
 there's any such misconfiguration.

 And one other thing I've noticed: (IMO) pydoc should generally be as
 detailed as the original C++ API documentation.  From a quick look, it
 largely misses information in the original doc.  You seem to copy some
 part of the C++ doc by hand, cutting much of it.  It has some obvious
 drawbacks: it can easily be incomplete; still it's time consuming;
 it's more easy to cause discrepancy between the original and wrapper
 doc over time.

 I suggest you use the doxygen2pydoc.py tool available under
 src/lib/util/python/ of the "trac933preview" branch.

 p.s. please do not return the ticket to me (yet).  I've just looked
 into it for the urgent issue of crash for now.

-- 
Ticket URL: <http://bind10.isc.org/ticket/2853#comment:7>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development