Re: [JDK-9] RFR: 8050370: Need new regressions tests for messageDigest with DigestIOStream

[zaiyao liu]

Hi Max,

Please help review the change:
http://cr.openjdk.java.net/~tyan/kevin/JDK-8050370/webrev02/

Thanks

Kevin
于 2014/8/20 10:35, zaiyao liu 写道:

 Hi Max,

Sorry for the confuse, how about following chang?

 136             boolean enDigest = false;
 139                 while ((k = dis.read()) != -1) {
 144                         dos.write(k);

145 if (enDigest == false) { // write data into baOut when dos and dis digest function on

 146                             baOut.write(k);
 147                         }
 143                         dis.on(enDigest);
 148                         dos.on(enDigest);
 149                         enDigest = !enDigest;
 152                 }

Thanks

Kevin
于 2014/8/19 22:42, Weijun Wang 写道:

I am understanding the models now, but still don't grasp the read/write codes very well. Take the READ example:

 136             boolean enDigest = false;
 139                 while ((k = dis.read()) != -1) {
 143                         dis.on(enDigest);
 144                         dos.write(k);
 145                         if (enDigest == false) {
 146                             baOut.write(k);
 147                         }
 148                         dos.on(enDigest);
 149                         enDigest = !enDigest;
 152                 }

dis.on() is called *after* dis.read() and dos.on() *after* dos.write(). Is it possible to change the order so that I can easily see if the read/write has any effect on the digest?

Thanks
Max

On 08/19/2014 10:29 PM, Weijun Wang wrote:

I actually don't understand the details in the program. For example,
what does the different ReadModels mean? In each case, what bytes are
actually write into the streams? There are too many on and off and I
don't know what the program is doing.

Thanks
Max