Your latest changes look fine to me.
On 21 Jul 2014, at 13:00, Gross, Lukas <lukas.gr...@sap.com> wrote: > Hi Gavin, > > I implemented the suggested changes (new syntax and NSDateFormatter > caching). I also discussed 1) with Florian and he said that they had a > similar discussion when they implemented escaping for OpenCMIS and they > decided to go for the ‘?’ approach because it is easier to implement and > well known from JDBC implementations. > > Best Regards, > Lukas > > PS: I will start the release this afternoon if there are no more comments > on my changes > > On 21/07/14 01:04 pm, "Gavin Cornwell" <gavin.cornw...@alfresco.com> wrote: > >> Hi Lukas, >> >> I discussed 1) with Mike this morning and we don’t think it needs doing >> for the 0.4 release. In fact, as you point out, this is the approach used >> for Prepared statements and the OpenCMIS implementation is always our >> reference so if that’s the way it’s been implemented there too, let’s >> leave it as it is. >> >> It would be good to do the others though, I’m still used to the old way >> of doing things too, but I do agree that new code should be added using >> the new/modern syntax. >> >> Regards, >> >> Gavin >> >> >> >> On 18 Jul 2014, at 14:20, Gross, Lukas <lukas.gr...@sap.com> wrote: >> >>> Hi, >>> >>> Mike has some valid points regarding the escaping implementation. Please >>> find my comments below. >>> >>> 1- I used Œ?¹ as parameter token because I basically followed the >>> OpenCMIS >>> implementation. Furthermore this is a common pattern also in other >>> Prepared Statement implementations. >>> If I got you right you suggest to use {1} Š {n} parameters in the string >>> and then use rangeOf to find and replace them instead of parsing the >>> string character by character. >>> >>> 2 - I will change this one >>> >>> 3 - The new syntax is definitively nicer and I use it especially when >>> initializing large arrays or dictionaries. However I¹m still kind of >>> used >>> to writing old-fashioned code... but I can change this, if it makes you >>> sleep better :) >>> >>> 4- I will check this >>> >>> Is 1) something we should consider changing before releasing 0.4? Is >>> everyone the same opinion as Mike? The current approach follows the >>> OpenCMIS implementation. Do we want to have something different here? >>> >>> Regards, >>> Lukas >>> >>> On 18/07/14 02:39 pm, "Gavin Cornwell" <gavin.cornw...@alfresco.com> >>> wrote: >>> >>>> Hi, >>>> >>>> I raised and fixed https://issues.apache.org/jira/browse/CMIS-823 to >>>> cover the category issue I mentioned in the status call earlier this >>>> week. >>>> >>>> I¹ve also committed an update to release.sh to use the version number >>>> defined in ObjectiveCMIS.xcconfig so we only have to change it in one >>>> place. >>>> >>>> I had a quick look at the query escaping changes, I don¹t have anything >>>> else to add over and above what Mike said. >>>> >>>> That completes all the changes we wanted to make for 0.4. >>>> >>>> Regards, >>>> >>>> Gavin >>>> >>>> >>> >> >