Re[1]: AW: AW: [Ecb-list] ecb and semantic

[Eric M. L. <er...@si...>]

This is interesting.  Perhaps a little patch like this fixes the
problem so that ECB doesn't have to worry about such things?

Eric

*** semantic.el.~1.163.~	Thu Sep 12 23:24:34 2002
--- semantic.el	Wed Oct 23 11:25:12 2002
***************
*** 582,587 ****
--- 582,589 ----
  The optional argument CHECKCACHE is ignored.  It is maintained for
  compatibility with previous versions of Semantic."
    (and
+    ;; Is this a semantic enabled buffer?
+    (semantic-active-p)
     ;; Application hooks say the buffer is safe for parsing
     (run-hook-with-args-until-failure
      'semantic-before-toplevel-bovination-hook)


>>> "Berndl, Klaus" <kla...@sd...> seems to think that:
  [ ... ]
>
>Ok, please excuse my making so many noise....
>
>i have just checked the ECB code deeper. The ECB-function added to
>the hook 'semantic-after-toplevel-cache-change-hook' is not only
>called via this hook but also internal on demand by ECB itself. It
>seems that semantic is working correctly i.e. the hooks are NOT
>called for not-semantic-buffers like text- or tex-buffers.
>
>But nevertheless: IMHO the behavior of ECB (clearing the
>method-buffer for not-semantic-buffers) is correct and senseful
>because otherwise we would have tokens displayed for a not-tokenized
>buffer (which would be very bad because then clicking on a token in
>the methopd-buffer would jump to somewhere in the current text- tex-
>or etc...buffer)! Semantic does not parse these buffers because
>(semantic-bovinate-toplevel t) returns fast and immediatelly nil for
>not-semantic-buffers (right, Eric?) so the only thing ECB does is to
>rebuild its method-buffer with a nil-tokencache which results in
>clearing the method-buffer. What is the problem?
  [ ... ]

-- 
          Eric Ludlam:                 za...@gn..., er...@si...
   Home: http://www.ludlam.net            Siege: www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: www.gnu.org