Delivered-To: garrigue at math.nagoya-u.ac.jp
Delivered-To: lablgtk at yquem.inria.fr
Subject: Re: [Lablgtk] entry completion for hundreds of thousands of items
Mime-Version: 1.0 (Apple Message framework v1081)
Content-Type: text/plain; charset=us-ascii
From: Yoann Padioleau <padator at wanadoo.fr>
In-Reply-To: <9CA765FE-8942-45F2-8203-DE219559E1FA at wanadoo.fr>
Date: Sat, 4 Sep 2010 09:20:11 -0700
Content-Transfer-Encoding: quoted-printable
Message-Id: <AB0391A0-3CA2-4080-8BE7-75016331818F at wanadoo.fr>
References: <AANLkTimW0Eytj4GNJSj6NAsnF65zVrR_xo0o514H860T at mail.gmail.com>
	<AANLkTi=XN6YAjVTYLELVD1A2zT2r8cvJKcius8TLZQ7T@mail.gmail.com>
	<9CA765FE-8942-45F2-8203-DE219559E1FA@wanadoo.fr>
To: Yoann Padioleau <padator at wanadoo.fr>
Cc: lablgtk at yquem.inria.fr
Status: U


On Sep 4, 2010, at 8:34 AM, Yoann Padioleau wrote:

>=20
>=20
> On Sep 4, 2010, at 3:37 AM, Adrien wrote:
>=20
>>=20
>> Hi,
>>=20
>> I've only taken a very quick look at your code but I think that what
>> you need is described in the treeview tutorial[1].
>=20
> Your link explains that I should detach my model from the view but
> I use the entry/entry-completion widget, not the treeview, and in my =
code
> I build a model without any view/entry attached anyway to and it is =
already slow. Also
> I don't see any method to disable sorting on GTree.list_store.

Hmm I didn't see the advice on using a custom model at the first =
reading.

>=20
>>=20
>> Also, I've experienced such issues with pretty short lists or trees.
>> No need to have 1000 elements, something like 50 or 100 may be enough
>> to notice a slowdown.
>>=20
>> Hope this helps.

Yes, the custom model should help I think. I've tried inserting 100000 =
elements
using lablgtk/examples/custom_list_generic.ml and it is way faster. I =
just have
to adapt the example to use instead the entry completion widget.


>>=20
>> [1] =
http://plus.kaist.ac.kr/~shoh/ocaml/lablgtk2/treeview-tutorial/ch03s03.htm=
l#sec-treestore-adding-many-rows
>>=20
>> --
>> Adrien Nader
>>=20
>>=20
>> On 04/09/2010, yoann padioleau <yoann.padioleau@gmail.com> wrote:
>>> Hi,
>>>=20
>>> I want to provide some completion for text where the corpus of items =
is
>>> above 100000. I've tried to build a GTree.list_store with those =
100000 items
>>> but it takes a too long time to build the model. I've then tried to
>>> split those 100 000 iterms
>>> in chunk of 1000 based on common prefix.  Then as the user start to
>>> type a string, I build a model
>>> with only the relevant 1000 items. But even building a model with =
only
>>> 1000 items takes more than 2 seconds
>>> which makes the whole application look slow. Is there a way to
>>> accelerate the insertion of items
>>> in a list_store ?
>>>=20
>>> I've attached my modification of src/examples/entrycompletion.ml
>>> showing how slow it is.
>>>=20
>>=20
>> _______________________________________________
>> Lablgtk mailing list
>> Lablgtk@yquem.inria.fr
>> http://yquem.inria.fr/cgi-bin/mailman/listinfo/lablgtk
>>=20
>=20
>=20
>=20
> _______________________________________________
> Lablgtk mailing list
> Lablgtk@yquem.inria.fr
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/lablgtk
>=20



_______________________________________________
Lablgtk mailing list
Lablgtk@yquem.inria.fr
http://yquem.inria.fr/cgi-bin/mailman/listinfo/lablgtk