Neil Hodgson wrote:
Robert Roessler:
We may not be communicating here... I have been using the ability to
remove markers AND indicators while handling SCN_MODIFIED for quite
some time now. IIRC, you suggested handling markers this way in
response to my problems with the marker "residue" being left after
lines were deleted - removing the markers in response to a
SC_MOD_BEFOREDELETE was the approved method of solving this.
I added indicators to this myself later with no observed problems... I
seem to recall you discussing loosening the restrictions even further,
for that matter.
Adding indicators inside a SC_MOD_BEFOREDELETE notofcation should
fail as enteredCount should be non-zero.
Ah, the ambiguities of [natural] language - "added" as in "added to
the mix" rather than enabling indicators... of course, I am still
zeroing the indicator bits part of the style byte.
So I verified what the CURRENT version of my Scintilla-using code is
doing... in my "notify" handler, I have the following cases that are
relevant to this discussion:
SCN_MODIFIED when (has_modifier SC_MOD_BEFOREDELETE) ->
(* remove ALL markers in range of lines *)
SCN_MODIFYATTEMPTRO ->
(* set readonly status to false *)
(* zero ALL indicator bits in buffer *)
(* remove ALL instances of specific marker *)
Perhaps there is no conflict in our world views... :)
Robert Roessler
[EMAIL PROTECTED]
http://www.rftp.com
_______________________________________________
Scintilla-interest mailing list
[email protected]
http://mailman.lyra.org/mailman/listinfo/scintilla-interest
