Strange default for StringBody default charset

2010-12-15 Thread sebb
The default charset for StringBody is currently set to Charset.defaultCharset(). This seems odd, as it varies between hosts. The equivalent class in HC 3.1 appears to be StringPart, which uses the default of "US-ASCII" for the charset name. That seems to be a more sensible default. It's probably

Re: HttpParams read-only access as Set

2010-12-15 Thread Oleg Kalnichevski
On Wed, 2010-12-15 at 19:12 +, sebb wrote: > On 15 December 2010 15:26, Oleg Kalnichevski wrote: > > ... > > > >> In the case of DefaultedHttpParams, I think the local names should > >> also be made available, e.g. through a method such as getLocalNames(). > >> But the default getNames() shoul

Re: HttpParams read-only access as Set

2010-12-15 Thread sebb
On 15 December 2010 15:26, Oleg Kalnichevski wrote: > ... > >> In the case of DefaultedHttpParams, I think the local names should >> also be made available, e.g. through a method such as getLocalNames(). >> But the default getNames() should include all names. >> It's easy create a new Set with th

[jira] Updated: (HTTPCLIENT-975) add support to caching module for RFC 5861 (stale-on-error and stale-while-revalidate)

2010-12-15 Thread Michajlo Matijkiw (JIRA)
[ https://issues.apache.org/jira/browse/HTTPCLIENT-975?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Michajlo Matijkiw updated HTTPCLIENT-975: - Attachment: stale-while-revalidate.updated.patch Here is the patch bought up

Re: HttpParams read-only access as Set

2010-12-15 Thread Oleg Kalnichevski
... > In the case of DefaultedHttpParams, I think the local names should > also be made available, e.g. through a method such as getLocalNames(). > But the default getNames() should include all names. > > The default names can be easily extracted using getDefaults().getNames(). > > Though I do w

Re: HttpParams read-only access as Set

2010-12-15 Thread sebb
On 15 December 2010 09:34, Oleg Kalnichevski wrote: > On Tue, 2010-12-14 at 22:25 +, sebb wrote: >> I've now committed a version of the HttpParams hierarchy using the new >> HttpParamsSet interface. >> I chose to stay with entrySet(), rather than iterator(), as that gives >> a bit more flexibi

[jira] Updated: (HTTPCLIENT-1035) cache should invalidate obsoleted entries mentioned in Content-Location

2010-12-15 Thread Jon Moore (JIRA)
[ https://issues.apache.org/jira/browse/HTTPCLIENT-1035?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jon Moore updated HTTPCLIENT-1035: -- Attachment: content-location-response-invalidation-test.patch > cache should invalidate ob

[jira] Created: (HTTPCLIENT-1035) cache should invalidate obsoleted entries mentioned in Content-Location

2010-12-15 Thread Jon Moore (JIRA)
cache should invalidate obsoleted entries mentioned in Content-Location --- Key: HTTPCLIENT-1035 URL: https://issues.apache.org/jira/browse/HTTPCLIENT-1035 Project: HttpComponents Htt

Re: HttpParams read-only access as Set

2010-12-15 Thread Oleg Kalnichevski
On Tue, 2010-12-14 at 22:25 +, sebb wrote: > I've now committed a version of the HttpParams hierarchy using the new > HttpParamsSet interface. > I chose to stay with entrySet(), rather than iterator(), as that gives > a bit more flexibility. > > The key question is: do we actually want to keep