CEDET

Hi,

  I have a couple guesses why this might happen, but I cannot
replicate the problem.

  The position vector [number number] occurs when a tag is unlinked
from a buffer because it was reparsed, for example.  This could occur
if a tag was cached somewhere, and a full reparse was needed.  I'm not
sure where that would happen though.  Any ideas how to reliably
replicate it?

  I don't remember any changes in this area since the beta.

Eric

>>> "Jans, Hauke" <hauke.jans@...> seems to think that:
>Hi all.
>=20
>Sometimes i get an error when i want to jump to a symbol
>definition with semantic-complete-jump.
>=20
>The first time it works ok but the second time i get:
>=20
>=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>Signaling: (wrong-type-argument extentp [15492 15692])
>  semantic-overlay-get([15492 15692] face)
>  (cons (semantic-overlay-get o (quote face)) (semantic-overlay-get o =
>(quote old-face)))
>  (semantic-overlay-put o (quote old-face) (cons (semantic-overlay-get o =
>...) (semantic-overlay-get o ...)))
>  (let ((o ...)) (semantic-overlay-put o (quote old-face) (cons ... =
>...)) (semantic-overlay-put o (quote face) (or face ...)))
>  semantic-highlight-tag(("SM_Set_Reset_Reason_IF" variable =
>(:constant-flag t :default-value (nil)) nil [15492 15692]) nil)
>  semantic-momentary-highlight-tag(("SM_Set_Reset_Reason_IF" variable =
>(:constant-flag t :default-value (nil)) nil [15492 15692]))
>  #<compiled-function nil "...(72)" [semanticdb-search-system-databases =
>tag nil semantic-complete-read-tag-project "Symbol: " 2 3 push-mark =
>semantic-go-to-tag switch-to-buffer semantic-momentary-highlight-tag =
>working-message "%S: %s "] 5 =
>("h:\\.xemacs\\cedet-1.0beta3b\\semantic\\semantic-complete.elc" . =
>39125) nil>()
>  call-interactively(semantic-complete-jump)
>=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>=20
>To me it looks like the extent is already cached via semanticdb
>and so is expressed as [number number] but the calling function=20
>(in this case semantic-highlight-tag) expects an living extent.
>=20
>Maybe smenatic-overlay-get must be made aware of this?
>=20
>Or is it already fixed in the CVS?
>=20
>This is with cedet-1.0b3 and XEmacs
  [ ... ]