FileCacheLoader create() Method Implementation

[jiwils]

From FileCacheLoader.java:

public void create() throws Exception {
 if(!root.exists())
 throw new FileNotFoundException(root.toString());
}

Is there a reason that the FileCacheLoader implementation of create throws an exception if the "root" cache loader location is not found rather than just create it? Checking File.canRead() and File.canWrite() might make sense as a sanity check here as well.

If such a reason exists, would an additional configuration property named createLocation that controlled creation of the "root" cache loader location make sense? I would be happy to provide a patch via JIRA that provides this behavior.

[belaban]

"jiwils" wrote:
From FileCacheLoader.java:

public void create() throws Exception {
 if(!root.exists())
 throw new FileNotFoundException(root.toString());
}

Is there a reason that the FileCacheLoader implementation of create throws an exception if the "root" cache loader location is not found rather than just create it?

No, you can change this

Checking File.canRead() and File.canWrite() might make sense as a sanity check here as well.

Go ahead

If such a reason exists, would an additional configuration property named createLocation that controlled creation of the "root" cache loader location make sense? .

Okay, under the CacheLoaderConfig attribute would be okay. Go ahead and make those changes, thanks,

[jiwils]

Based on your response, the removal of the exception throwing behavior is probably the route to take as the addition of a property would not buy us much. Unless I misread what you wrote (please correct me if so), I will head down this path and post a feature request with patch to JIRA shortly...

[jiwils]

"jiwils" wrote:
...I will head down this path and post a feature request with patch to JIRA shortly...

http://jira.jboss.com/jira/browse/JBCACHE-158

[belaban]

Don't you have CVS write access. Go ahead and commit those changes once the FileCacheLoaderTest passes...

[jiwils]

Don't you have CVS write access. Go ahead and commit those changes once the FileCacheLoaderTest passes...

I have committer access for JGroups, but not for the JBoss stuff, so I just created a JIRA issue...access would be good...