[fyi][icedyea-web] allow null in SecurityValueValidator
Adam Domurad
adomurad at redhat.com
Thu Feb 28 06:37:43 PST 2013
On 02/28/2013 09:35 AM, Jiri Vanek wrote:
> On 02/28/2013 09:18 AM, Jiri Vanek wrote:
>> I have overlooked that validate() is called also on fully constructed
>> map, not only during its loading from file.
>> In fully constructed map the null is of course possible. If detected,
>> then default is returned by getter.
>>
>> J.
>
> There was obvious NPE hidden in previous patch. Thanx Adam for catch.
>
> Version2 attached.
>
> J.
> diff -r 56c94adc52ea ChangeLog
> --- a/ChangeLog Wed Feb 27 17:00:48 2013 +0100
> +++ b/ChangeLog Thu Feb 28 15:33:49 2013 +0100
> @@ -1,3 +1,8 @@
> +2013-02-28 Jiri Vanek <jvanek at redhat.com>
> +
> + * netx/net/sourceforge/jnlp/config/SecurityValueValidator.java:
> modifed
> + null check - no considered as correct value as being valid value
> in runtime.
> +
> 2013-02-27 Jiri Vanek <jvanek at redhat.com>
>
> Added backend and settings for extended applet security
> diff -r 56c94adc52ea
> netx/net/sourceforge/jnlp/config/SecurityValueValidator.java
> --- a/netx/net/sourceforge/jnlp/config/SecurityValueValidator.java
> Wed Feb 27 17:00:48 2013 +0100
> +++ b/netx/net/sourceforge/jnlp/config/SecurityValueValidator.java
> Thu Feb 28 15:33:49 2013 +0100
> @@ -45,7 +45,10 @@
> @Override
> public void validate(Object value) throws IllegalArgumentException {
> if (value == null) {
> - throw new IllegalArgumentException("Value can't be null");
> + //null is correct, it means it is not user set
Really small nit: I find a space after // more readable
> + //and so defautl hsoudl be sued whatever it is
s/defautl hsoudl be sued/the default should be used/
> + //returning to prevent NPE in fromString
I think we can do without this comment. It isn't necessary to understand
the code.
> + return;
> }
> if (value instanceof AppletSecurityLevel) {
> //??
Otherwise looks good (feel free to push).
Thanks.
-Adam
More information about the distro-pkg-dev
mailing list