Terra Informatica Forums: Last 35 Posts

Arun on "insert new elements dynamically using csss!"

Arun — Thu, 07 Aug 2008 03:41:03 +0000

hmm no .. involves div's and widgets alone.

andrew on "insert new elements dynamically using csss!"

andrew — Wed, 06 Aug 2008 11:35:52 +0000

Is this about changing value of <textarea>/<plaintext> elements?

Arun on "insert new elements dynamically using csss!"

Arun — Wed, 06 Aug 2008 01:09:13 +0000

like
<tag1>

<tag2></tag2> //wont change and no configuration allowed the user still has to type this in
<tag3></tag3>

</tag1>

Arun on "insert new elements dynamically using csss!"

Arun — Wed, 06 Aug 2008 01:07:27 +0000

Is it possible right to insert new elements just by using the csss!?

I have a template SDK code which has some redundant tags, so i want it to get added automatically, so the user doesnt have to type it in. Is it possible ?

Arun on "selector support in htmllayout"

Arun — Tue, 05 Aug 2008 13:23:47 +0000

wow thanx a lot andrew .. tat did the trick..

andrew on "selector support in htmllayout"

andrew — Tue, 05 Aug 2008 13:11:31 +0000

Try this:

<html>
<head>
  <style>
  div.heading
  {
    color:blue;
    behavior:check; // so it will flip-flop its :value from true to false
    value-changed!: self.next().hide-rows = self:value? "odd" # "even";
  }
  div.heading:checked
  {
    color:red;
  }
  div.content[hide-rows="odd"] > div:nth-child(odd) {display:none}
  div.content[hide-rows="even"] > div:nth-child(even) {display:none}
  </style>
<head>
<body>

<div id="parent">
  <div class="heading">heading</div>
  <div class="content">
    <div>1</div>
    <div>2</div>
    <div>3</div>
    <div>4</div>
  </div>
</div>

</body>
</html>

Arun on "selector support in htmllayout"

Arun — Tue, 05 Aug 2008 12:49:17 +0000

actually its not a descendant but a sibling for the heading element.

If heading value is 1 in the first div, i want to select all the even childs otherwise the odd childs (ofcourse except the first child).

andrew on "selector support in htmllayout"

andrew — Tue, 05 Aug 2008 12:34:36 +0000

For the first case:

div:not([browser=="my_browser"]) {display:none}

I am not sure I understand second case. My best guess is that you need:

div[heading="0"] > div:nth-child(odd) {display:none}
div[heading="1"] > div:nth-child(even) {display:none}

List of supported selectors is here:

http://www.terrainformatica.com/htmlayout/selectors.whtm

Arun on "selector support in htmllayout"

Arun — Tue, 05 Aug 2008 12:06:16 +0000

Does htmllayout support the following type of element selection

1) inverted condition check in attribute selectors like

div[browser!="my_browser"] {display:none}

2) n-th sibling of an element

Arun on "CSSS! - automatically identify scroll bar to an element"

Arun — Tue, 05 Aug 2008 12:00:26 +0000

hey hey got it using the :: operator iam able to access the style attrbutes.

Arun on "CSSS! - automatically identify scroll bar to an element"

Arun — Tue, 05 Aug 2008 11:54:14 +0000

thnx a lot for your quick response this is really usefull (yet i didnt find any dynamic firing events),

now i want to make some changes to the div element's child's style (in this case)
can you provide me some snippets on how to change styles using csss! ??

andrew on "CSSS! - automatically identify scroll bar to an element"

andrew — Tue, 05 Aug 2008 10:50:22 +0000

Yes, try this:

<html>
<head>
  <style>
    div {
      width:50%;
      height:50%;
      border:1px solid;
      overflow:auto;
    }
    div
    {
      active-on! : span = $1(span),
                   span:value = self.box-client-width() < self.box-content-width()?
                   "h-scrollbar" #
                   "no-h-scrollbar";
    }
  </style>
<head>
<body>
  <span>result</span>
  <div>longlonglonglonglonglonglonglonglonglonglonglonglonglonglong</div>
</body>
</html>

Arun on "CSSS! - automatically identify scroll bar to an element"

Arun — Tue, 05 Aug 2008 10:04:08 +0000

Is there a way for an element to check whether a scrollbar is drawn for it or not (due to overflow) using CSSS!
??

andrew on "gui_task"

andrew — Tue, 05 Aug 2008 09:48:33 +0000

You are passing dom::element* ptr but inside xxxElementCallback you are treating it as HELEMENT*.

BOOL CALLBACK xxxElementCallback( HELEMENT he, LPVOID param )
	{
		dom::element *pel = (dom::element *)param;
		*pel = he; // <<<<<<<<<<<<<<<<<<<<<<<<<
		return true;
	}

I've marked line that will produce different code from yours.

senyai on "gui_task"

senyai — Mon, 04 Aug 2008 07:18:23 +0000

Hi there!
From one thread I call 'gui_queue.push( new inject_in_id(rootElement, element ) );' and then i see 'access violation'. The code work ok is there is no HTMLayoutVisitElements. Please, set me on the right path.

struct inject_in_id: public htmlayout::gui_task
{
	htmlayout::dom::element el;
	htmlayout::dom::element root;

	static BOOL CALLBACK xxxElementCallback( HELEMENT he, LPVOID param )
	{
		HELEMENT *el = (HELEMENT*)param;
		(*el) = he;
		return true;
	}
	inject_in_id( const htmlayout::dom::element& _root, const htmlayout::dom::element& _el ): el(_el), root(_root) {}
	inline void exec() {
		htmlayout::dom::element where_ = NULL;
		//root.append(el);
		//root.update(true);

		HTMLayoutVisitElements(root, NULL, \"id\", L\"debuglogs\", xxxElementCallback, &where_, 0xff);
		where_.append(el);
		where_.update(true);
	}
};

So sorry for stupid question :)

DmitryYakimov on "Linux port: Sciter or HTMLayout?"

DmitryYakimov — Wed, 30 Jul 2008 02:54:24 +0000

I vote for MacOS, because it is more profitable for shareware/middle-size companies.

x-zero on "Linux port: Sciter or HTMLayout?"

x-zero — Tue, 29 Jul 2008 14:24:10 +0000

or just use wine

andrew on "HTMLite"

andrew — Tue, 29 Jul 2008 00:15:12 +0000

You can think about them as:

HTMLiteMeasure = WM_SIZE
HTMLiteRender = WM_PAINT
HTMLiteRenderOnBitmap = WM_PRINT

Not exactly but close enough.

andrew on "Scintilla"

andrew — Tue, 29 Jul 2008 00:11:06 +0000

It cannot.

adontz on "HTMLite"

adontz — Mon, 28 Jul 2008 08:50:34 +0000

May I have any comments on HTMLiteMeasure, HTMLiteRender, HTMLiteRenderOnBitmap parameters and call sequence?

x-zero on "Scintilla"

x-zero — Mon, 28 Jul 2008 02:59:10 +0000

it can work in htmlayout without change any code ?

zverok on "Bug with dynamic frames creation"

zverok — Mon, 28 Jul 2008 02:40:35 +0000

It is too old controvercy. In short: I'm not thinking about HTMLayout as about browser. Because I use it as UI library (to create and show application, not document).

andrew on "about table row"

andrew — Sun, 27 Jul 2008 21:38:56 +0000

I've fixed this in 3.3.0.9 but anyway I suggest to use last row as such a template.

andrew on "About the calendar style"

andrew — Sun, 27 Jul 2008 21:36:05 +0000

'cause <input type=date/calendar> uses table inside.

Define specifically what you want:

table#table-manipulate
{
}
table#table-manipulate > tr:current
{
}
table#table-manipulate td
{
}
table#table-manipulate tr.header
{
}

and everything will be all right.

hlqyq on "About the calendar style"

hlqyq — Sun, 27 Jul 2008 18:50:27 +0000

Why the style of calendar was affected by table style,even by my custom table style? example:

<html>
<head>
<meta http-equiv=Content-Type content="text/html;charset=gb2312">
<style>
table
{
behavior:grid;
overflow:auto;
width:80%%; height:50%%;
border-spacing:0;
background-color:white;
border:1px solid silver;
}
table > tr:current
{
color:white;
background-color:blue;

}
table td
{
padding:2px 3px;
border:1px solid;
border-color: transparent silver silver transparent;
width:*;
}
table tr.header
{
background-color:silver silver black black;
color:white;
}
</style>
</head>
<body>
<table fixedrows=1 fixedlayout id="table-manipulate">
<tr .header ><td>姓名</td><td>年龄</td><td>体重</td></tr>
<tr .prototype style="display:block">
<td>张三</td>
<td>18</td>
<td><input type="date" /></td>
</tr>
</table>
</body>
</html>

hlqyq on "about table row"

hlqyq — Sun, 27 Jul 2008 18:25:21 +0000

Thanks,I use the hidden row as a prototype row to write a common table manipulate,such as InsertRow、AppendRow、RemoveRow、ClearRows、MoveUpRow,MoveDownRow...

andrew on "Bug with dynamic frames creation"

andrew — Sun, 27 Jul 2008 17:41:41 +0000

DOM API in htmlayout is pretty low level thing. And I would like to keep it this way.
Means that not everything works as you would expect in browser.
And I have doubts that it is even possible to create frames dynamically in conventional browsers.

andrew on "Scintilla"

andrew — Sun, 27 Jul 2008 17:34:20 +0000

For Sciter - yes. That is sciter-scintilla.dll
It is in Sciter SDK and is a part of ScIDE:
http://www.terrainformatica.com/sciter/scide.zip

Screenshot:
http://www.terrainformatica.com/screenshots/scide.png

Sources of sciter-scintilla.dll are here:
http://www.terrainformatica.com/sciter/sciter-scintilla.zip

That is slightly modified version of original Scintilla distribution + behavior code (folder /sciter/). It could be ported to htmlayout if needed.

x-zero on "Scintilla"

x-zero — Sun, 27 Jul 2008 16:12:07 +0000

is there behavior who have Scintilla support ?

andrew on "About frame"

andrew — Sun, 27 Jul 2008 11:45:37 +0000

cellspacing=X is:

table { border-spacing:Xpx; }

cellpadding=X is:

table td { padding:Xpx; }

andrew on "What exectly the support is?"

andrew — Sun, 27 Jul 2008 11:38:41 +0000

I've sent an answer.

x-zero on "About frame"

x-zero — Sun, 27 Jul 2008 07:15:21 +0000

cellspacing="-2" fixed my problem.

zverok on "Bug with dynamic frames creation"

zverok — Sun, 27 Jul 2008 01:08:36 +0000

aha, old good "third-class-ness" of dynamically created DOM!

andrew on "Bug with dynamic frames creation"

andrew — Sat, 26 Jul 2008 16:15:47 +0000

That is not a bug strictly speaking. Attributes are used for initialization of runtime values. When your frameset appears in the DOM it has no children so processing of cols discarded.

Either use following styling:

frameset { flow:horizontal; border-spacing:3px; }
frameset > :nth-child(1) { width:2*; height:*; }
frameset > :nth-child(2) { width:8*; height:*; }

or create frameset content first and then add it to the DOM.

andrew on "about table row"

andrew — Sat, 26 Jul 2008 15:37:11 +0000

<table fixedlayout> uses height of first non-header row as a value of height of all other rows. In your case row has zero height (display:none) that is why all other rows are not visible.

Algorithm shall use visible rows for calculations in fact. I'll fix that.