Eric M. Ludlam wrote:
>=20
>=20
> Oi, I think I built the autoload files with CVS Emacs again.
> Perhaps this would make it work:
>=20
> make clean-all
> make
This solved the problem, thanks.
>>- Next I tried semantic-ia-complete-symbol and now I'm getting "No=20
>>method definition: semanticdb-needs-refresh-p (nil)". Since I had no=20
>>better ideas, I tried nuking my semantic caches but it didn't help.
>>
>>I haven't had the time to dig any deeper yet, nor to build a newer=20
>>emacs, but I thought I'd send off these few issues now and dig more lat=
er.
>=20
> [ ... ]
>=20
> Hmmm, curious. I'd be curious to see a stack trace.
> M-x toggle-debug-on-error RET
>=20
> My investigations showed a different error I need to look into when
> semantic-ia-complete-symbol is already on a complete symbol.
>=20
Here's the stack trace. I think I've found a starting point but I=20
haven't investigated further: it works just fine when my header file is=20
a ".h" file in C mode. But unfortunately, it is not. When it's a=20
".pan" in fundamental mode I get this. Opening up the .pan file and=20
manually switching to c mode causes it to work, but there is some=20
interesting behaviour here.
What I did to test this was put a struct definition in the .pan file=20
and then I have a c file which includes this .pan file and declares a=20
variable of that structure type. I then use semantic-ia-complete-symbol=20
to try to complete a field in that struct.
- If I load just the .c file and attempt a completion I get the "no=20
method defintion" error.
- If I then switch to the .pan file (which was loaded by the attempt to=20
complete the symbol) and change it to C mode, then switch back to the c=20
file, I no longer get the error when I try to complete -- but I don't=20
get the correct completion either, it gives me argc, argv, and other=20
things like that.
- But, if I switch the .pan file to c mode before attempting any=20
completions, everything is just fine. I just don't know what bits of=20
state I need to clear out in order to reset things when I get stuck in=20
the 2nd mode above.
Anyway, here's the trace. It appears to my untrained eye that the=20
culprit is semanticdb-file-table-object.
Debugger entered--Lisp error: (no-method-definition=20
semanticdb-needs-refresh-p (nil))
signal(no-method-definition (semanticdb-needs-refresh-p (nil)))
eieio-generic-call(semanticdb-needs-refresh-p (nil))
semanticdb-needs-refresh-p(nil)
semanticdb-file-table-object("test.pan")
semanticdb-find-table-for-include-default(("test.pan" include nil nil=20
#<overlay from 1 to 20 in test.c>) nil)
semanticdb-find-table-for-include(("test.pan" include nil nil=20
#<overlay from 1 to 20 in test.c>))
semanticdb-find-translate-path-includes-default(nil)
semanticdb-find-translate-path-default(nil nil)
semanticdb-find-translate-path(nil nil)
semanticdb-find-tags-collector(#[(table tags) "=C3=08 \n#=1B,G"=1B(B [=
table=20
prefix tags semanticdb-find-tags-for-completion-method] 4] nil nil)
semanticdb-find-tags-for-completion("")
semantic-analyze-find-tags-by-prefix("")
semantic-analyze-possible-completions-default([object=20
semantic-analyze-context "context" (102 . 102) ("foo" "") (function=20
variable) nil nil nil (("i" variable ... ... [60 66]) ("f" variable ...=20
... [75 88])) #<buffer test.c>])
semantic-analyze-possible-completions([object=20
semantic-analyze-context "context" (102 . 102) ("foo" "") (function=20
variable) nil nil nil (("i" variable ... ... [60 66]) ("f" variable ...=20
... [75 88])) #<buffer test.c>])
semantic-ia-get-completions([object semantic-analyze-context=20
"context" (102 . 102) ("foo" "") (function variable) nil nil nil (("i"=20
variable ... ... [60 66]) ("f" variable ... ... [75 88])) #<buffer=20
test.c>] 102)
semantic-ia-complete-symbol(102)
* call-interactively(semantic-ia-complete-symbol)
--=20
Jonathan Emmett
jonathan.emmett@...
|