CEDET

Hi,

  There are two functions.  One is speedbar-vc-check-dir-p, and the
other speedbar-this-file-in-vc.

  It appears that patching speedbar to accept a configurable list of
possible VC options, or patching to just add SVN support would both be
pretty easy.

  As an asside, speedbar supports RCS, CVS, and SCCS.

Good Luck
Eric

>>> Arne Schmitz <arne.schmitz@...> seems to think that:
>
>A question concerning Speedbar: It seems that in the directory browsing mod=
>e,=20
>the Speedbar VC only wants to use CVS. I did not find a way to configure it=
>=20
>to use SVN. Is this possible? How can I do that?
>
>Arne
>

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

cvs version works correctly
thanks, Eric

_________________________________________________________________
간편하고 정확한 나만의 맞춤 검색, Live Search!
http://search.live.com

>>> "=?ISO-8859-1?Q?Per_Nordl=F6w?=" <per.nordlow@...> seems to think that:
>Noticed that the member completion of type plus the C member-operators . or
>-> is active also in comments which is probably not what we want for
>default. Especially not when writing comment sentences ending with the same
>dot.

Hi,

  Which completion mode is doing that?  This should not occur with the
idle completion function.

Thanks
Eric

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

Hi,

  Thanks for the patch.  I have applied it to CVS, as it still passes
  some of my tests.

Eric

>>> LSD <lsdsgster@...> seems to think that:
>- Forgive me bad English, here is google translation.
>
>semantic can properly resolve as "public virtual" such a class
>inheritance, it does not resolve "virtual public".
>
>I think you see below this patch, you will guess what I want to say.
>
>
>
>--- c.by.orig	2008-07-02 04:39:02 +0800
>+++ c.by	2008-07-14 10:46:59 +0800
>@@ -303,6 +303,8 @@
> one-class-parent
>   : opt-class-protection opt-class-declmods namespace-symbol
>     (TYPE-TAG (car $3) "class" nil nil :protection (car $1))
>+  | opt-class-declmods opt-class-protection namespace-symbol
>+    (TYPE-TAG (car $3) "class" nil nil :protection (car $2))
>   ;
>
> class-parents
  [ ... ]

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

Hi,

  Items added to the c include path do not automatically include
  subdirectories.  Thus, if you had the file:

/usr/include/foo/bar.h

and your C file had:

#include <foo/bar.h>

then just having /usr/include is all you need.  If you had in your C
file:

#include <bar.h>

then you would need to add /usr/include/foo to your include path.

Good Luck
Eric

>>> "=?ISO-8859-1?Q?Per_Nordl=F6w?=" <per.nordlow@...> seems to think that:
>Is it dumb to set semantic-c-dependency-system-include-path to /usr/include=
>?
>Is it slower than adding only those sub-directories to /usr/include that ar=
>e
>used by my project?
>
>/Nordl=F6w

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

Hi,

  For starters, you can use:

M-x semantic-decoration-all-include-summary RET

  to get a summary of your includes, and what is in or out of a
search.

  If all the includes you expect are there, you can use:

M-x semantic-analyze-debug-assist RET

  to get some ideas on what is missing for the analyze.  It may just
be failing to parse the local context.

In either case, this info will make it easier for me to debug what is
going on.

Eric

>>> "Antoine Levitt" <smeuuh@...> seems to think that:
>Hi,
>I'm trying to make semantic understand my project.
>Here's my .emacs :
>(setq-mode-local c-mode
>                 semanticdb-find-default-throttle
>                 '(project unloaded system recursive))
>
>(setq semantic-c-dependency-system-include-path '("/usr/include"
>                          "/home/smeuuh/eda/eda"
>                          "/usr/local/include/freesg"
>                          "/usr/local/include/agar"
>                          "/usr/include/SDL"
>                          "/usr/local/include"))
>
>Note that I use setq semantic-c-dependency-system-include-path, because the
>prescribed way, semantic-add-system-include, didn't work at all for me
>(includes not found at all)
>So now it works more or less : all includes can be found (ie, when I click
>on the include part of ECB, it finds the file). However, autocompletion
>doesn't always.
>For instance, in /home/smeuuh/eda/eda/generic/capacitor.c, I have a
>reference to a DC structure, which is declared in
>/home/smeuuh/eda/eda/core/dc.h
>The chain of include is :
>in /home/smeuuh/eda/eda/generic/capacitor.c, #include <eda.h>
>in /home/smeuuh/eda/eda/eda.h : #include <core/circuit.h>
>in /home/smeuuh/eda/eda/core/circuit.h : #include <core/dc.h>
>
>However, when I try to complete dc->, it tells me "cannot find type for dc"
>What did I do wrong ?
>

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

Hi, thanks for your quick reply.
semantic-decoration-all-include-summary lists the header I want. The
relevant line is :   dc.h :    4 tags, 0 are includes. It corresponds to
reality (two structs, two typedefs)
Here is the partial output of analyse-debug-assist, while prefixing with dc
:
Unable to find datatype for: "class ES_SimDC dc".
Declared type is: "class ES_SimDC {}"
Raw data type is: ("ES_SimDC" type (:type "class") nil nil)

Semantic could not find this data type in any of its global tables.

The context is :
static void
UpdateModel(ES_Capacitor *c, M_Real v, ES_SimDC *dc)
{
    dc

2008/7/8 Eric M. Ludlam <eric@...>:

> Hi,
>
>  For starters, you can use:
>
> M-x semantic-decoration-all-include-summary RET
>
>  to get a summary of your includes, and what is in or out of a
> search.
>
>  If all the includes you expect are there, you can use:
>
> M-x semantic-analyze-debug-assist RET
>
>  to get some ideas on what is missing for the analyze.  It may just
> be failing to parse the local context.
>
> In either case, this info will make it easier for me to debug what is
> going on.
>
> Eric
>
> >>> "Antoine Levitt" <smeuuh@...> seems to think that:
> >Hi,
> >I'm trying to make semantic understand my project.
> >Here's my .emacs :
> >(setq-mode-local c-mode
> >                 semanticdb-find-default-throttle
> >                 '(project unloaded system recursive))
> >
> >(setq semantic-c-dependency-system-include-path '("/usr/include"
> >                          "/home/smeuuh/eda/eda"
> >                          "/usr/local/include/freesg"
> >                          "/usr/local/include/agar"
> >                          "/usr/include/SDL"
> >                          "/usr/local/include"))
> >
> >Note that I use setq semantic-c-dependency-system-include-path, because
> the
> >prescribed way, semantic-add-system-include, didn't work at all for me
> >(includes not found at all)
> >So now it works more or less : all includes can be found (ie, when I click
> >on the include part of ECB, it finds the file). However, autocompletion
> >doesn't always.
> >For instance, in /home/smeuuh/eda/eda/generic/capacitor.c, I have a
> >reference to a DC structure, which is declared in
> >/home/smeuuh/eda/eda/core/dc.h
> >The chain of include is :
> >in /home/smeuuh/eda/eda/generic/capacitor.c, #include <eda.h>
> >in /home/smeuuh/eda/eda/eda.h : #include <core/circuit.h>
> >in /home/smeuuh/eda/eda/core/circuit.h : #include <core/dc.h>
> >
> >However, when I try to complete dc->, it tells me "cannot find type for
> dc"
> >What did I do wrong ?
> >
>
> --
>          Eric Ludlam:                       eric@...
>   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net
>

Hi,

  It sounds like the datatype "ES_SimDC" is somehow hidden in dc.h.
If you put dc.h into a buffer and do:

M-x bovinate RET

  you can look through the datastructure for ES_SimDC.  If it is
somehow hidden behind #defines, you will need to update your CPP
symbol table.  If it is some syntax semantic doesn't support yet, then
we'll need to identify it and update the parser.

Eric

>>> "Antoine Levitt" <smeuuh@...> seems to think that:
>Hi, thanks for your quick reply.
>semantic-decoration-all-include-summary lists the header I want. The
>relevant line is :   dc.h :    4 tags, 0 are includes. It corresponds to
>reality (two structs, two typedefs)
>Here is the partial output of analyse-debug-assist, while prefixing with dc
>:
>Unable to find datatype for: "class ES_SimDC dc".
>Declared type is: "class ES_SimDC {}"
>Raw data type is: ("ES_SimDC" type (:type "class") nil nil)
>
>Semantic could not find this data type in any of its global tables.
>
>The context is :
>static void
>UpdateModel(ES_Capacitor *c, M_Real v, ES_SimDC *dc)
>{
>    dc
>
>2008/7/8 Eric M. Ludlam <eric@...>:
>
>> Hi,
>>
>>  For starters, you can use:
>>
>> M-x semantic-decoration-all-include-summary RET
>>
>>  to get a summary of your includes, and what is in or out of a
>> search.
>>
>>  If all the includes you expect are there, you can use:
>>
>> M-x semantic-analyze-debug-assist RET
>>
>>  to get some ideas on what is missing for the analyze.  It may just
>> be failing to parse the local context.
>>
>> In either case, this info will make it easier for me to debug what is
>> going on.
>>
>> Eric
>>
>> >>> "Antoine Levitt" <smeuuh@...> seems to think that:
>> >Hi,
>> >I'm trying to make semantic understand my project.
>> >Here's my .emacs :
>> >(setq-mode-local c-mode
>> >                 semanticdb-find-default-throttle
>> >                 '(project unloaded system recursive))
>> >
>> >(setq semantic-c-dependency-system-include-path '("/usr/include"
>> >                          "/home/smeuuh/eda/eda"
>> >                          "/usr/local/include/freesg"
>> >                          "/usr/local/include/agar"
>> >                          "/usr/include/SDL"
>> >                          "/usr/local/include"))
>> >
>> >Note that I use setq semantic-c-dependency-system-include-path, because
>> the
>> >prescribed way, semantic-add-system-include, didn't work at all for me
>> >(includes not found at all)
>> >So now it works more or less : all includes can be found (ie, when I click
>> >on the include part of ECB, it finds the file). However, autocompletion
>> >doesn't always.
>> >For instance, in /home/smeuuh/eda/eda/generic/capacitor.c, I have a
>> >reference to a DC structure, which is declared in
>> >/home/smeuuh/eda/eda/core/dc.h
>> >The chain of include is :
>> >in /home/smeuuh/eda/eda/generic/capacitor.c, #include <eda.h>
>> >in /home/smeuuh/eda/eda/eda.h : #include <core/circuit.h>
>> >in /home/smeuuh/eda/eda/core/circuit.h : #include <core/dc.h>
>> >
>> >However, when I try to complete dc->, it tells me "cannot find type for
>> dc"
>> >What did I do wrong ?
>> >
>>
>> --
>>          Eric Ludlam:                       eric@...
>>   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net
>>
>

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net

public.avatar@... ([AvataR]) writes:

> In toplevel form:
> semantic-c.el:42:21:Error: Invalid read syntax: ")"
> Wrote /home/avatar/.emacs.d/cedet/semantic/bovine/semantic-el.elc

I've reported the same problem few days ago, still no idea how to fix
it. Note that if you call `byte-compile-file` on `semantic-c.el` from
running Emacs, it will compile just fine.
-- 
Happy Hacking.

http://sphinx.net.ru
む

>>> public.avatar@... ([AvataR]) seems to think that:
>Good time of day, Eric!
>Today's night snapshot don't want be compiled anymore. I get:
>
>In toplevel form:
>semantic-c.el:42:21:Error: Invalid read syntax: ")"
>Wrote /home/avatar/.emacs.d/cedet/semantic/bovine/semantic-el.elc
>
>
>If I make hack, and change eval-when-compile => eval
>it compiles, but don't work :)
  [ ... ]

Hi,

  This is because all the autoload files changed their format, and
eieio is now needed to get autoload files to load.  I bumped into
something similar at work, and used:

$ make bootstrap

  which deleted all the various generated files, then started from
scratch.  Things then went smoothly for me.

Good Luck
Eric

-- 
          Eric Ludlam:                       eric@...
   Siege: http://www.siege-engine.com          Emacs: http://cedet.sourceforge.net