[cpp-qt5] Fix memory leaks in Pet Store client sample

[etherealjoy]

PR checklist

• Read the contribution guidelines.
• Ran the shell script under ./bin/ to update Petstore sample so that CIs can verify the change. (For instance, only need to run ./bin/{LANG}-petstore.sh and ./bin/security/{LANG}-petstore.sh if updating the {LANG} (e.g. php, ruby, python, etc) code generator or {LANG} client's mustache templates). Windows batch files can be found in .\bin\windows\.
• Filed the PR against the correct branch: Default: master.
• Copied the technical committee to review the pull request if your PR is targeting a particular programming language.

Description of the PR

Allow generated code to do cleanup of allocated memory by preventing premature loop exit.
The generated code has the possibility to delete all memory allocations from itself but the sample client is terminating the loop before allowing the callback to return and memory de-allocation to take place.
With this change the callback from the timer and response schedule an event to the eventloop and then perform the clean up after which the exit is done.

Tests

• Running valgrind on the PetStore Client sample binary produces no warning of memory leak.

@stkrwork @MartinDelille @ravinikam @fvarose

[etherealjoy]

@MartinDelille
You had opened an issue to modernize the Qt sample, As of now no memory leaks are in the generated code.

However the setter are still not user friendly in that for pointer types the user has to care about de-allocating.
We can make it such that the setters themselves update the data structures internally instead of replacing the pointers and the user has to take care of his/her own allocated memory.
for example,
setters for string members, or list members which are using pointers to memory for data storage.

[MartinDelille]

I quickly review your PR but I don’t understand why you’re not using local variables instead of pointer: you wouldn’t have to deallocate the memory at the end of the block and would reduce the code size.

I will point an example in your code and make a proposal.

[etherealjoy]

@MartinDelille
Indeed, I did not add any new variables just refactor old ones. But if you think we should I can do this here itself.
Please add the example also would be great.

[etherealjoy]

@MartinDelille @stkrwork
Should I refactor here or just fix current ?
I would prefer to just fix current here and refactor with another issue.

[MartinDelille]

@ethereajoy ok do as you wish!

[MartinDelille]

What about:

const QString available = “available “

[etherealjoy]

We will do it in the next update.
I plan to switch to QVariant for all members and lets see. It will be a breaking change but I will make it with the server first.

[etherealjoy]

OK, I will create a different issue to track these for both Qt5 Client and Server since they share common code.

[etherealjoy]

@wing328
I think we can go ahead with this.

[wing328]

The CircleCI (1.0) error can be ignored as latest master has migrated to use CircleCI 2.0 with better and more stable build.