On Thu, 17 Jun 2021 12:05:28 GMT, Dongbo He <dongb...@openjdk.org> wrote:
>> src/java.base/share/classes/sun/security/util/AbstractAlgorithmConstraints.java
>> line 72:
>>
>>> 70: algorithmsInProperty = property.split(",");
>>> 71: for (int i = 0; i < algorithmsInProperty.length; i++) {
>>> 72: algorithmsInProperty[i] =
>>> algorithmsInProperty[i].trim().toLowerCase(Locale.ENGLISH);
>>
>> Is it possible to keep the current behavior that the property could be
>> sensitive? It may be not desired to allow "keysize" for "keySize" spec in
>> the property.
>
> If we keep property sensitive, we may need to use TreeSet. I have updated the
> PR with TreeSet. Fortunately, the performance hasn't changed much.
I did not get the point to use TreeSet. Is it sufficient if the toLowerCase()
is not added (and don't compare keywords like "keySize" by ignoring cases)?
- algorithmsInProperty[i] = algorithmsInProperty[i].trim().toLowerCase(...);
+ algorithmsInProperty[i] = algorithmsInProperty[i].trim();
-------------
PR: https://git.openjdk.java.net/jdk/pull/4424
