>>>>> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Jean-Marc Lasgouttes wrote:
>>>>>>> "Jean-Marc" == Jean-Marc Lasgouttes <[EMAIL PROTECTED]>
>>>>>>> writes:
>>
>>>>>>> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Hum the version I committed from JMarc used the
Abdelrazak> CursorSlice::operator<() which does not check for idx()
Abdelrazak> IIRC. Should add this test JMarc?
>>
Jean-Marc> I double checked and it does check for idx. However, I see
Jean-Marc> now that any container implements lexicographic ordering
Jean-Marc> tests. Question: would the following patch make sense?
Jean-Marc> Abdel, I know I have removed code you just added, but I am
Jean-Marc> wondering why we have to declare those extra things. Also,
Jean-Marc> it seems to me that the following could be removed:
>> I tried to remove some of the operators that should be defined by
>> <utility>, but could not get it to work. I give up (it is not time
>> for such cleanups anyway).
Abdelrazak> I told you already that there are some clash with the
Abdelrazak> void* operator.
Yes, and then I made the following 'explicit':
///
ParIterator(DocIterator const &);
///
ParConstIterator(DocIterator const &);
Having them implicit causees g++ to transform a DocIterator to
ParIterator for no good reason (when you want to compare two
dociterators, for example). This mostly work, but then you have to
cleanup the lazy code that mixes ParIterator and ParConstIterator.
All in all, not a job for now.
JMarc
