[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [cdt-dev] Fix for bug 307311
- From: Marc Khouzam <marc.khouzam@xxxxxxxxxxxx>
- Date: Thu, 20 Oct 2011 13:51:13 -0400
- Accept-language: en-US
- Acceptlanguage: en-US
- Delivered-to: firstname.lastname@example.org
- Thread-index: AcyKl/fhOq3zyWenTFGmYLPI3bAWtgEuBvmw
- Thread-topic: [cdt-dev] Fix for bug 307311
> -----Original Message-----
> From: cdt-dev-bounces@xxxxxxxxxxx
> [mailto:cdt-dev-bounces@xxxxxxxxxxx] On Behalf Of Mathias Kunter
> Sent: Friday, October 14, 2011 1:37 PM
> To: cdt-dev@xxxxxxxxxxx
> Subject: [cdt-dev] Fix for bug 307311
apologies for not being able to answer earlier.
It is nice to see your effort for this contribution
and I hope a committer will have time to review it soon.
I personally have no time right now, but I'll keep this
in the back of my mind if time does become available and no
one else has jumped in.
> I've already posted to this mailing list some weeks ago, but didn't
> receive any reply so far - so I'm trying again now: I've fixed bug
> 307311 and attached a corresponding patch. Please see
> The fix brings UTF-8 Unicode support to the Eclipse CDT
> debugging views.
> The patch works well for me, but a short review might cover the
> following things:
> 1) Changes have been made to the MIConst.java file contained within
> the package org.eclipse.cdt.dsf.mi.service.command.output, but what
> about the package org.eclipse.cdt.debug.mi.core.output? Does it have
> to be changed, too?
org.eclipse.cdt.dsf.mi.service.command.output is used for the DSF integration of GDB while
org.eclipse.cdt.debug.mi.core.output is used for the CDI integration of GDB.
CDT uses the DSF integration by default, but the CDI one is still available if the
user selects it explicitly.
It is up to you if you want to add this feature to both integration of GDB
or only one of them. Doing it in the DSF one (as you have done) will make
it available by default, which is nice.
> 2) MIVarEvaluateExpressionInfo and MIDataEvaluateExpressionInfo should
> probably call MIConst.getString instead of MIConst.getCString in order
> to support UTF-8 for evaluated expressions too, if this has no other
> side effects (I didn't notice any, at least).
> 3) This patch is related to the following gdb options:
> --> gdb's "host-charset" must be set to "UTF-8" (if it's not, the
> patch possibly can't recognize all UTF-8 strings correctly). As far as
> I know, this is gdb's default setting on Linux (and Mac?), but not on
> Windows. CDT users on Windows can set this via the .gdbinit file, but
> a separate Eclipse preference option would probably be handier.
> --> gdb's "print sevenbit-strings" must be set to "on" (again, if it's
> not, the patch possibly can't recognize all UTF-8 strings correctly).
> This is the default setting for the gdb/mi interface on all platforms
> as far as I know.
> I'd be happy about any feedback / comments. Thank you very much for
> your valued time and work.
> Best regards,
> Mathias Kunter
> cdt-dev mailing list