Hi William, Sorry for the delay, but I had to take a look in the current codebase first - to be honest, I'm not familiar with a lot of internal Cactus stuff (most code was created by Vincent).
So, here are some comments... On 4/24/06, William Ferguson <[EMAIL PROTECTED]> wrote: > I have had to provide an exploded implementation of the EarArchive, > WarArchive and JarArchive to allow the EarParser to handle an exploded Makes sense, cool! > deployment. Currently they implement the Cactus interfaces as this is > what is required to allow integration into EarParser. How exactly would EarParser use the new classes? I mean, currently the parse() method instantiate a DefaultEarArchive; would you be creating a new method that takes the EarArchive as parameter and refactor the current method to call the new one? > 1) Submit as is, in which case when the Cargo integration is completed, > they must be enhanced and converted such that they implement the cargo > interfaces. This should be the easiest option, but... > 2) Change them to implement the Cargo interfaces now, but this means > that EarParser (WarParser should be changed too) would have to rely on > the Cargo implementations. If you are concerning with Cactus depending on Cargo jars, that's fine - that dependency already exists. > I am not sure that this wouldn't break something else. Me neither. But if it doesn't break the test cases, then in theory it should be fine :-) > It would also be better to have done this in another changeset than the one > I have for 238. You mean implementing them as option 1 first (and closing 238) and later doing the refactor to use Cargo? > Which path would you prefer I follow? I prefer path 2 because it would be the best option in the long term. But as the Cargo integration is not ready yet, path 1 might be a wiser choice... > Also, I have the doco and a testcase ready to go. Great, these are the most important pieces of the puzzle ;-) > So whenever you give me feedback on above I should be able to submit a > patch containing the lot. Ok, sorry again for the delay... -- Felipe --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
