#615: added TYPEFLAGS constants for TYPEATTR in ITypeInfo.GetTypeAttr() #621
Conversation
| @@ -43,6 +43,7 @@ Features | |||
| * [#608](https://github.com/java-native-access/jna/pull/608): Mavenize the build process - change parent and native pom artifactId/name to differentiate in IDE and build tools. [@bhamail](https://github.com/bhamail) | |||
| * [#613](https://github.com/java-native-access/jna/pull/613): Make Win32Exception extend LastErrorException [@lgoldstein](https://github.com/lgoldstein). | |||
| * [#613](https://github.com/java-native-access/jna/pull/614): Added standard 'Kernel32Util#closeHandle' method that throws a Win32Exception if failed to close the handle [@lgoldstein](https://github.com/lgoldstein). | |||
| * [#621](https://github.com/java-native-access/jna/pull/621): COM: Added TYPEFLAGS constants for TYPEATTR in ITypeInfo.GetTypeAttr() [@SevenOf9Sleeper](https://github.com/SevenOf9Sleeper). | |||
There was a problem hiding this comment.
Can you please try to make this look more like the other CHANGELOG entries please (and fix anything above that doesn't look the same)?
A correct one is:
* [#569](https://github.com/java-native-access/jna/pull/569): Added `com.sun.jna.platform.win32.Winspool.PRINTER_INFO_2` support. Added GetPrinter and ClosePrinter functions in `com.sun.jna.platform.win32.Winspool` - [@IvanRF](https://github.com/IvanRF).
There's an issue number, a dash before the Github name, all code is back-quoted and we mention namespaces, particularly important since JNA is not just Win32.
|
Squash your commits, too, please. |
SevenOf9Sleeper
force-pushed
the
com_add_typeflags
branch
2 times, most recently
from
563cc85 to
10b96be
Compare
cosmetic corrections done
SevenOf9Sleeper
force-pushed
the
com_add_typeflags
branch
from
10b96be to
a6a81f2
Compare
|
I have done the demanded changes. But I hav no clue why the build has failed?! Seems so, that it has nothing to do with my changes... |
|
I just reissued the testrun and it came back green. The failing test relies on unloading and (I asume) garbage collection - the JVM has not guarantees for this, just observed behaviour. I have one last nit-pick - the other constants in this interface are decared on the interface level, not the structure level (see FADF_*). The values are verified, so then it could be merged. BTW: Nice work in CHANGES.md - much appretiated! |
|
referring to your "nit-pick"...:-) I have thought about where to place the constants. And I have decided to place them in the structure TYPEATTR where they are needed. |
|
Merged this. Constants should be declared where they are declared in the Windows SDK. Is that what you're referring to? |
Motivation: Creating a new Mac instance each time is wasteful. As the method is expected to be called from the EventLoop we should better cache these in a FastThreadLocal Modifications: Cache instances in FastThreadLocal Result: Less overhead
No description provided.