You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Paul Burba <pa...@softlanding.com> on 2006/04/10 12:11:41 UTC
Re: [PATCH] was: More documentation for svn_utf_cstring_from_utf8_ex()[Show] etc.
Julian Foad <ju...@btopenworld.com> wrote on 03/26/2006 08:15:24 AM:
> Peter N. Lundblad wrote:
> > I think both of these are too messy for little gain.
>
> Well, I'd describe the first idea (requiring initialisation) as the
> "tidy" way
> in terms of its result (and I'm sure we'll want to do it sooner or
> later), but
> if you mean it's too much code churn for this particular gain then Ican
agree
> with you.
>
>
> After all that, I'm happy for you (Paul) to go with something like what
you
> had, but probably "revving" the API as Peter mentioned because I don't
really
> see a benefit in not doing so.
>
> - Julian
Hello,
Sorry for the delay in getting back to this. Let me know if this is what
you all had in mind.
Paul B.
[[[
Rev svn_utf_cstring_*_utf8_ex so xlate_handle_node_t structs are
cached based on the conversion taking place rather than a string
key argument.
* subversion/include/svn_utf.h
(svn_utf_cstring_to_utf8_ex2, svn_utf_cstring_from_utf8_ex2): New.
(svn_utf_cstring_to_utf8_ex, svn_utf_cstring_from_utf8_ex):
Deprecate.
* subversion/libsvn_subr/cmdline.c
(SVN_UTF_CONTOU_XLATE_HANDLE, SVN_UTF_UTOCON_XLATE_HANDLE,
SVN_UTF_ETOU_XLATE_HANDLE): Remove.
(svn_cmdline_cstring_from_utf8, svn_cmdline_cstring_to_utf8,
svn_cmdline__getopt_init): Change calls to svn_utf_cstring_*_utf8_ex
to use svn_utf_cstring_*_utf8_ex2.
* subversion/libsvn_subr/io.c
(SVN_UTF_ETOU_XLATE_HANDLE, SVN_UTF_UTOE_XLATE_HANDLE): Remove.
(svn_io_create_unique_link, svn_io_read_link): Change calls to
svn_utf_cstring_*_utf8_ex to use svn_utf_cstring_*_utf8_ex2.
* subversion/libsvn_subr/utf.c
(get_xlate_key): New function.
(convert_cstring): Update doc string.
(svn_utf_cstring_to_utf8_ex2): New.
(svn_utf_cstring_to_utf8_ex): Reimplement as wrapper around
svn_utf_cstring_to_utf8_ex2.
(svn_utf_cstring_from_utf8_ex2): New.
(svn_utf_cstring_from_utf8_ex): Reimplement as wrapper around
svn_utf_cstring_from_utf8_ex2.
* subversion/libsvn_client/diff.c,
subversion/libsvn_diff/diff_file.c,
subversion/libsvn_repos/hooks.c,
subversion/libsvn_subr/stream.c,
subversion/libsvn_subr/subst.c,
subversion/svn/util.c:
Change all callers of svn_utf_cstring_*_utf8_ex to use
svn_utf_cstring_*_utf8_ex2.
]]]
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
Re: [PATCH] was: More documentation for svn_utf_cstring_from_utf8_ex()[Show]
etc.
Posted by Julian Foad <ju...@btopenworld.com>.
Paul Burba wrote:
> Just an FYI: I tested this via ra_local and ra_dav on Windows XP (Apache
> 2.0.54, Neon 0.24.7) with fsfs. I also ran my OS400 versions of the test
> suite. No problems.
>
> Also, FWIW, I ran two versions of the svnserve client tests side-by-side
> against 1.3.1 on the iSeries, one with this change applied, one without,
[...]
Thanks. That's more than enough testing for me! +1 to commit it.
- Julian
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] was: More documentation for svn_utf_cstring_from_utf8_ex()[Show] etc.
Posted by Paul Burba <pa...@softlanding.com>.
Julian Foad <ju...@btopenworld.com> wrote on 04/10/2006 10:51:56 AM:
> Paul Burba wrote:
> > Sorry for the delay in getting back to this. Let me know if this is
what
> > you all had in mind.
> >
> > Paul B.
> >
> > [[[
> > Rev svn_utf_cstring_*_utf8_ex so xlate_handle_node_t structs are
> > cached based on the conversion taking place rather than a string
> > key argument.
>
> Yes, this patch looks fine to me, though I haven't tested it.
Julian,
Just an FYI: I tested this via ra_local and ra_dav on Windows XP (Apache
2.0.54, Neon 0.24.7) with fsfs. I also ran my OS400 versions of the test
suite. No problems.
Also, FWIW, I ran two versions of the svnserve client tests side-by-side
against 1.3.1 on the iSeries, one with this change applied, one without,
trying to see if there was any noticeable performance difference (i.e. Did
the string manipulation now required for every call to svn_utf_*_utf8_ex2
show up in an appreciable way?). The OS400 port makes fairly heavy use of
the *_ex2 funcs since all the command line programs use it to convert
incoming EBCDIC args to UTF-8. Not exactly scientific, but they both ran
within a few seconds of each other. These tests take several hours to
complete on the iSeries so that seems an insignificant difference.
Paul B.
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] was: More documentation for svn_utf_cstring_from_utf8_ex()[Show]
etc.
Posted by Julian Foad <ju...@btopenworld.com>.
Paul Burba wrote:
> Sorry for the delay in getting back to this. Let me know if this is what
> you all had in mind.
>
> Paul B.
>
> [[[
> Rev svn_utf_cstring_*_utf8_ex so xlate_handle_node_t structs are
> cached based on the conversion taking place rather than a string
> key argument.
Yes, this patch looks fine to me, though I haven't tested it.
- Julian
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] was: More documentation for svn_utf_cstring_from_utf8_ex()[Show] etc.
Posted by "Peter N. Lundblad" <pe...@famlundblad.se>.
Paul Burba writes:
> Julian Foad <ju...@btopenworld.com> wrote on 03/26/2006 08:15:24 AM:
>
> Sorry for the delay in getting back to this. Let me know if this is what
> you all had in mind.
>
Looks good. +1.
Regards,
//Peter
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org