On 1/13/20 3:34 PM, Randy MacLeod wrote:
On 1/13/20 3:20 PM, Mingde (Matthew) Zeng wrote:
Have you by chance tried to add ptest coverage for poppler?
If not, Matthew may do that.
Have not done any ptest on this - just some runtime tests on target
with evince
Andreas
```
$ make test
Running tests...
Test project /home/mzeng/code/poppler/build
Start 1: check_qt5_attachments
1/21 Test #1: check_qt5_attachments ............***Failed 2.47 sec
Start 2: check_qt5_dateConversion
2/21 Test #2: check_qt5_dateConversion ......... Passed 0.58 sec
...
```
Poppler's `make test` runs tests from the `/build/qt5/tests`
directory, which apparently are relying on qt5.
```
/****************************************************************************
** Meta object code from reading C++ file 'check_attachments.cpp'
**
** Created by: The Qt Meta Object Compiler version 67 (Qt 5.14.0)
**
** WARNING! All changes made in this file will be lost!
*****************************************************************************/
#include <memory>
#include <QtCore/qbytearray.h>
#include <QtCore/qmetatype.h>
...
```
Is it worth it to write a ptest and sort out the meta-qt5 layer?
No, we can't have meta-openembedded packages depend
on meta-qt5 content.
The meta-oe layer only depends on oe-core (aka core):
http://cgit.openembedded.org/meta-openembedded/tree/meta-oe/conf/layer.conf#n30
Of course you could add the ptest code in a .bbappend in
the meta-qt5 layer but I won't ask you to do that.
Take a look at the bbappends in meta-qt5 if you are curious:
https://layers.openembedded.org/layerindex/branch/master/layer/meta-qt5/#appends
../Randy
Thanks for looking into it Matthew,
../Randy
Matthew
--
# Randy MacLeod
# Wind River Linux
--
_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.openembedded.org/mailman/listinfo/openembedded-devel
