[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-finance-apps
Subject:    Re: GSoc GDB pretty printers: Automatic testing and printing class pointers
From:       Thomas Baumgart <thb () net-bembel ! de>
Date:       2021-07-02 9:55:41
Message-ID: 2060065.OBFZWjSADL () thb-nb
[Download RAW message or body]


On Mittwoch, 30. Juni 2021 20:09:27 CEST Da Viper wrote:

[...]

> The current problem i am facing now is how to automate the test cases,
> i am not sure how to go about that, it is to make sure that older classes
> are not affected when new printer classes are created

I investigated this a bit and here's what I came up with: you can use
the gdb -batch mode feature to automate tests, collect the results in a
first run and store them as ground truth. When run again, compare the
current output against the ground truth and make sure it does not differ.

Here's what I did in the tests subdir:

a) I compiled your test program with

   g++ -g -I /usr/include/qt5/QtCore/ -I /usr/include/qt5  -lstdc++ -lQt5Core -o test qstring.cpp

b) Next I wrote a file and called it 'autotest'. Its content is

---8<---
file test
break main
run
n
p str
n
p str
p flist
n
n
n
p flist
---8<---

c) Then I ran

   gdb -batch -x autotest > ground.truth

   and made sure that the content of ground.truth is what I expected. The file here
   contains

---8<---
Breakpoint 1 at 0x400ec8: file qstring.cpp, line 7.
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".

Breakpoint 1, main (argc=1, argv=0x7fffffffd908) at qstring.cpp:7
7               QString str("firstString");
8               QList<QString> flist;
$1 = "firstString"
9               flist.append(str);
$2 = "firstString"
$3 = <empty>
10              flist.append("second");
11              flist.append("third");
13              QList<QString> alist{ "thsit", "ssfldsf", "sfsfs"};
$4 = {"firstString", "second", "third"}
---8<---


d) At this point, you usually want to make changes to your code. Once done
   you continue with

e) and run

   gdb -batch -x autotest > verification
   diff ground.truth verification

and make sure that there is no difference.

Hope that gives you some ideas.


-- 

Regards

Thomas Baumgart

https://www.signal.org/       Signal, the better WhatsApp
-------------------------------------------------------------
'The compiler doesn't read comments and neither do I'
-------------------------------------------------------------

["signature.asc" (application/pgp-signature)]

[prev in list] [next in list] [prev in thread] [next in thread]