From kde-devel  Sun Aug 31 20:49:39 2008
From: Sebastian =?iso-8859-15?q?K=FCgler?= <sebas () kde ! org>
Date: Sun, 31 Aug 2008 20:49:39 +0000
To: kde-devel
Subject: Re: Patch: New KRunner for places.
Message-Id: <200808312249.39992.sebas () kde ! org>
X-MARC-Message: https://marc.info/?l=kde-devel&m=122021586030221
MIME-Version: 1
Content-Type: multipart/mixed; boundary="--===============0659511247=="

--===============0659511247==
Content-Type: multipart/signed;
  boundary="nextPart1737786.UhP7EFFhoW";
  protocol="application/pgp-signature";
  micalg=pgp-sha1
Content-Transfer-Encoding: 7bit

--nextPart1737786.UhP7EFFhoW
Content-Type: text/plain;
  charset="iso-8859-15"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

Hi David, welcome to Plasma development :)

(Michael is correct, please use plasma-devel for plasma development issues,=
=20
CC:)

On Sunday 31 August 2008 20:03:09 David Edmundson wrote:
> I've written a new KRunner for KDE. It looks at the bookmarks list for
> dolphin, the folders in the Places panel, and includes them in the krunner
> search. Nothing exciting or amazing, I wrote it because I went to use the
> functionality and was surprised that it didn't do it already.
>
> I'm posting here so it can be reviewed and committed, or so I can receive
> some constructive abuse about my code so it can be committed later.
>
> Code is attached.
>
> Part 2
>
> I imagine the follow up question I'm going to get asked is "does it also
> list devices like Dolphin does in the places bar?".
>
> I've tried doing this but I keep making Krunner crash, I've attached my
> code file for this. I think, though I may be wrong. That I get a crash wh=
en
> krunner runs my code to find a device matching the search term, I call
> solid, then krunner tells me to search for a new term. This means the Sol=
id
> objects are removed mid-way through it's dbus magic, which causes it to
> explode wildly. I've spent a while on this and I can't figure out how to
> fix it.

You shouldn't be doing all that in the match() method, as this is run very=
=20
often. If you look at other runners, they fill the list of possible matches=
 in=20
the constructor, have some mechanism to update this internal datastructure=
=20
(often just a QStringList), preferably by a signal so it doesn't get trigge=
red=20
when not possible. You can have a look at how other runners solve this for=
=20
examples, for example the recent document runner.

The match function is run very often, so it should be as short and efficien=
t=20
as possible. It's basically executed at every char that is being typed by t=
he=20
users (most runners restrict this to only trigger the matching when there a=
re=20
more than 3 chars typed).

I'd suggest to first rewrite it in this way, if your crashes (which, if I=20
understand correctly happen in match()) persist, look deeper into that.

Ow, and your code doesn't match Plasma's (kdelibs) coding style, please fix=
=20
that as well. ( http://techbase.kde.org/Policies/Kdelibs_Coding_Style for=20
details.)

Cheers,
=2D-=20
sebas

 http://www.kde.org | http://vizZzion.org |  GPG Key ID: 9119 0EF9=20


--nextPart1737786.UhP7EFFhoW
Content-Type: application/pgp-signature; name=signature.asc 
Content-Description: This is a digitally signed message part.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iQEVAwUASLsD42dNh9WRGQ75AQLn5wf/WDs+Z4Do4OuZzmQKI9YMwR1ERSu+cila
jcUncpesDsHCh/kFV2u7dIcQpF+WQ8wyN8cP0ZyoBmJOMLD1v0cT9WifJqtw1e0n
AF4PCXwMrIuU1jFrXZ+GhpZQ/pEKuFlHbwjBqFnJV7IVr9kgGDHeTxkNRmM6387K
TcryapPYIbIegtzXOO+GeFtZePdFKuy29fS9VxiZz5X9m33uQ0JforWgmDRHkAzG
jv4uy1SYe/v/u1EWO+9TH8FfUSKjoXGHPEPvDBr/JL92f3viIgk0wpN19Fe5H6Q4
HAcbURIBxyC455i8CwTisriBDxVexUF9a0yjPttSwTKGpxomkPzhVg==
=BUdg
-----END PGP SIGNATURE-----

--nextPart1737786.UhP7EFFhoW--

--===============0659511247==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

--===============0659511247==--