[PATCH] More tdb/ldb tests.

[Adrian Cochrane]

I've now moved the functions I was testing to be declared in 
ldb_module.h (in attachment 0001) and have also attached some 
additional tests (in attachment 0004) for Matthieu Patou's 
ldb_unpack_data_withlist (in attachments 0002 & 0003).

On Fri, Aug 7, 2015 at 2:04 PM, Adrian Cochrane 
<adrianc at catalyst.net.nz> wrote:
> Updated patches. I've fixed the style issues and have already 
> addressed the reasons for what I didn't correct.
> 
> On Fri, Aug 7, 2015 at 9:49 AM, Adrian Cochrane 
> <adrianc at catalyst.net.nz> wrote:
>> I looked into removing the use of ldb_private.h but the problem is 
>> it contains the very routines (ldb_unpack_data) that I want to test. 
>> To me the options on the private header seem to be either to move 
>> those routines to be declared in the public header to be tested or 
>> to keep the private header inclusion in the tests. There may be 
>> other solutions and I don't know which is nicer.
>> 
>> Furthermore I would consider fixing the SDDL translation a separate 
>> job and I'm not sure how to do it.
>> 
>> On Fri, Jul 31, 2015 at 6:27 PM, Stefan Metzmacher <metze at samba.org> 
>> wrote:
>>> Hi Adrian,
>>> 
>>>>  On Fri, 2015-07-31 at 12:52 +1200, Adrian Cochrane wrote:
>>>>>  This one took me a bit but this expands test coverage of tdb.
>>>> 
>>>>  Thanks Adrian.  I really appriciate your efforts to lock in the
>>>>  ldb_unpack_data behaviour.
>>>> 
>>>>  As you know, the next steps are to incorporate the
>>>>  ldb_unpack_data_withlist patches from Matthieu, and then test 
>>>> those.
>>>> 
>>>>  However, first I need you to fix up a few things.  Please read
>>>>  README.Coding.  In particular please fix up the comments and use 8
>>>>  -space tabs for indentation.
>>>> 
>>>>  For the ntSecurityDescriptor skip:
>>>>   - In the comment, explain that it can't round-trip via the SDDL 
>>>> string
>>>>  (that's what the line-noise like string starting with O: is)
>>>>   - Please don't just write 'Skipping' as the output, either 
>>>> remove the
>>>>  torture_comment or make it more meaningful.  I would prefer it 
>>>> removed
>>>>  as it will happen every time in normal behaviour.
>>> 
>>> Would it be possible to check if it's possible to fix the SDDL 
>>> translation
>>> to avoid losing information?
>>> 
>>> metze
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0002-ldb-introduce-ldb_unpack_data_withlist-to-unpack-par.patch
Type: text/x-patch
Size: 7124 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20150821/18f819b3/0002-ldb-introduce-ldb_unpack_data_withlist-to-unpack-par.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0003-ldif-increment-version-due-to-added-ldb_unpack_data_.patch
Type: text/x-patch
Size: 22211 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20150821/18f819b3/0003-ldif-increment-version-due-to-added-ldb_unpack_data_.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0004-ldb-torture-test-ldb_unpack_data_withlist.patch
Type: text/x-patch
Size: 8598 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20150821/18f819b3/0004-ldb-torture-test-ldb_unpack_data_withlist.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-ldb-torture-Test-ldb-unpacking-and-printing.patch
Type: text/x-patch
Size: 27096 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20150821/18f819b3/0001-ldb-torture-Test-ldb-unpacking-and-printing.bin>