[
https://issues.apache.org/jira/browse/HTTPCLIENT-1487?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13939832#comment-13939832
]
Chris Oliver commented on HTTPCLIENT-1487:
------------------------------------------
The issue is that the existing cache implementations although they store files
on disk, don't recover the cache when the process is restarted. Instead they
repopulate the cache by downloading the files from the server unnecessarily. In
my use cases some files are quite large and this is unacceptable. I've
implemented my own HttpCacheStorage that does this but had to copy many package
private classes from the BasicHttpStorage implementation to do that. In
addition, I had to guess implement serializing and deserializing the
HttpCacheEntry information to/ from the disk. It would be nice if this use case
(which is like that of standard web browser) was supported in a future version
at least as an option. Thanks for your work, btw.
> CachingHttpClient doesn't persist cache
> ---------------------------------------
>
> Key: HTTPCLIENT-1487
> URL: https://issues.apache.org/jira/browse/HTTPCLIENT-1487
> Project: HttpComponents HttpClient
> Issue Type: Bug
> Reporter: Chris Oliver
> Assignee: Jon Moore
> Priority: Blocker
>
> Unless I'm mistaken, the current implementation doesn't maintain the cache
> between executions of the java process. In addition, it appears the current
> api for HttpCacheStorage makes it difficult / error prone to do this myself,
> since my implementation would have to somehow produce an HttpCacheEntry. Any
> suggestions?
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
