Looks good. I'll push the changeset for you. --Sean
On 08/15/2012 07:37 PM, Jason Uh wrote:
Thanks, Sean. New webrev updated with your suggestions: http://cr.openjdk.java.net/~juh/6500133/webrev.01/ Jason On 08/15/2012 10:38 AM, Sean Mullan wrote:This looks good to me. Couple of comments: 111: Can you add a comment, something like "Try parsing the URI again after encoding/escaping any illegal characters". 113-4: When this code was written there probably wasn't yet an IOException(String, Throwable) ctor. Now there is, so you can change this to: throw new IOException("invalid URI name:" + name, use2); There are also a couple other places in URIName where you can replace the same code using initCause with the IOExc ctor above. That's a low-risk refactoring you can include in this change. --Sean On 08/14/2012 11:51 PM, Jason Uh wrote:Hi all, This change fixes -- 6500133: CertificateParsingException for CRL Distribution Point with blank; and 6931888: Inconsistent behavior for invalid URI name in cert file CRs: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6500133 http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6931888 They are effectively duplicates, both regarding an exception thrown when parsing CRL Distribution Point URIs with invalid characters, like a space or backslash. This change uses sun.net.www.ParseUtil.encodePath(String) to re-encode bad URIs. Webrev: http://cr.openjdk.java.net/~juh/6500133/webrev.00/ Thanks, Jason
