Re: [Zorba-coders] [Merge] lp:~zorba-coders/zorba/new-jsoniq-update into lp:zorba

2013-09-13 Thread Ghislain Fourny
Review: Approve -- https://code.launchpad.net/~zorba-coders/zorba/new-jsoniq-update/+merge/185295 Your team Zorba Coders is subscribed to branch lp:zorba. -- Mailing list: https://launchpad.net/~zorba-coders Post to : zorba-coders@lists.launchpad.net Unsubscribe : https://launchpad.net/~zo

Re: [Zorba-coders] [Merge] lp:~zorba-coders/zorba/new-jsoniq-update into lp:zorba

2013-09-13 Thread Ghislain Fourny
After looking more in depth, I realized that there is an inconsistency across parsenodes. Some of them have const exprnode* pointers on their children (like JSONObjectLookup had), some others regular exprnode* pointers (like JSONObjectLookup has now). The latter seem to correspond to places that

Re: [Zorba-coders] [Merge] lp:~zorba-coders/zorba/new-jsoniq-update into lp:zorba

2013-09-13 Thread Ghislain Fourny
Hi Matthias, I am not sure I understand what you are suggesting. The reason for non-const is that temporary parse nodes (in that case a lookup expression node) use regular pointers that they own to child parse nodes, i.e., the destructor deletes the object pointed to at the end. However, these

Re: [Zorba-coders] [Merge] lp:~zorba-coders/zorba/new-jsoniq-update into lp:zorba

2013-09-12 Thread Matthias Brantner
Review: Approve Looks great. I'll approve and let you decide on one question: Is it really necessary to make the expressions in the parsenodes non-const? Couldn't you use a unique pointer or auto_ptr in the parser that is passed to convert_postfix_to_target_and_selector and only released when th