--===============2084144591==
Content-Type: multipart/signed;
  boundary="nextPart1568574.VYVfWoVxMf";
  protocol="application/pgp-signature";
  micalg=pgp-sha1
Content-Transfer-Encoding: 7bit

--nextPart1568574.VYVfWoVxMf
Content-Type: text/plain;
  charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

hi,
here my changes I would like to do in a while:

a) convert use of unsigned int to int, as QT has changed to use int in near=
ly all places, it atm is a hassle to
    convert back to int in all loops, length compares and lookups, should m=
ake code more readable I guess

b) add an interface for some kind of "editor factory" or how ever to call it
    this will just a plain QObject which lets you create KTextEditor::Docum=
ents, it will avoid the current hassle
    in kate to have around one global object for ever and speed up document=
 creation for apps using it,
    if the app devs still use the old way,  nothing will break, but this wi=
ll give a faster way ;)

c) remove the DCOP interfaces from the lib, really, this should be done per=
 part, the current overhead of
   10000ths interfaces isn't worth the hassle, and it will save us from the=
 need for dpointers in the interfaces and
   the duplicated ints all over, only the document and view itself should p=
rovide some number to the outside to make
   handling easier ;) btw., perhaps should the int method just be part of i=
nterface and the implementation part should
   take the effort to make it unique, in the mini implementation it would l=
ook like the current in ktexteditor, but I think
   the interfaces should really be without any implementation at all inside

d) some little conflict solving and removing:
   - KTextEditor::Editor -> CU, this is the no view/doc seperation thing, i=
t's just not needed any more,
     yzis/kate supports it, and it's not that nice as the static queryMetho=
des don't cover it and it's a hack,
     away :)

   - SessionConfigInterface and ConfigInterface overlap, the readSessionCon=
fig... should be removed from ConfigInterface

   - CursorInterface: not sure, wasn't that usable in the current state, pe=
rhaps we should remove this now, and readd
     later a usable one, if we have all come up with one ;)

  - DynWordWrapInterface -> TRASH, not needed, this is part internal stuff,=
 why should host app want to enforce that,
    bypassing the users config, it has no effect on text content, not like =
the static word wrap, which might be needed to
    enforce ;)

  - PrintInterface -> CU, user action, too

  - UndoInterface -> CU, user action, too, + config stuff

  - ViewStatusMsgInterface -> CU, this contains only one signal, and is not=
 that usable at all, as nobody knows whats in
    this, if we really need this, we should note down in the KTextEditor::V=
iew which "non-interface" related signals parts
    should provide, just some comment would be sufficient, as this signal c=
onnection is nothing we do at compile time

e) merge some of the interfaces, which will avoid the current blow up of th=
em ;)
   - EditInterface + Ext -> EditInterface, the editBegin()/editEnd() transa=
ction methodes won't hurt to be in the basic
     editing interface, if the part doesn't support it, it should just do n=
othing there :)

   -  selection interface + ext + blockselection interface -> SelectionInte=
rface
      + this is FOR VIEW this time ;)

   - Clipboard Interface -> SelectionInterface, as this is for VIEW, too, a=
nd selection is per view, makes sense
     to collect this selection related 2 methodes in the SelectionInterface=
, too, no need for seperate, each
     editor supporting selection should have no probs with copy/paste/cut a=
nd co, not even sure if anybody
     uses this, but hey, if this are just 3 methodes in a interface, this w=
on't hurt, if it stays seperate, it will ;)

  - ConfigInterface + ConfigInterfaceExtension, simple merge, guess after t=
he conversion to int, the configPages()
     counter should just return -1 if this part doesn't allow to emebed ind=
ividual config pages ;)

  - MarkInterface + Extension -> MarkInterface

  - ViewCursorInterface -> CursorInterface, makes no sense to prefix, this =
is for view just, and good ;)
    if ever a other cursor interface for editing purpose comes up, this sho=
uld be EditCursor or CursorEdit or whatever

This would be all cleanup changes I have to propose ATM, this doesn't reall=
y change any function, just cleaning up the stuff,
which has grown to some lets say "not that small" amount ;)

cu
Christoph

=2D-=20
Christoph Cullmann
KDE Developer, kde.org Maintainance Team
http://www.babylon2k.de, cullmann@kde.org

--nextPart1568574.VYVfWoVxMf
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQBClu8MyPjDGePm9UIRAkalAKCkJsuBn1XHzqqBNg3W4m30hfYrQgCcCyb6
ABWZ7CYdk2G1Mbg2kg2XK5A=
=eKAp
-----END PGP SIGNATURE-----

--nextPart1568574.VYVfWoVxMf--

--===============2084144591==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Ktexteditor-devel mailing list
Ktexteditor-devel@kde.org
https://mail.kde.org/mailman/listinfo/ktexteditor-devel

--===============2084144591==--