[prev in list] [next in list] [prev in thread] [next in thread] 

List:       lyx-devel
Subject:    Re: feature request
From:       Abdelrazak Younes <younes.a () free ! fr>
Date:       2006-04-09 21:55:36
Message-ID: e1bvsq$f3v$1 () sea ! gmane ! org
[Download RAW message or body]

Leuven, E. a écrit :
>  > I had to do that in text.C. Cursor inherits DocIterator so
>  > you can use that. "ParIterator it(cur)" should do the trick.
> 
> that's what i did, see attached.
> 
> it get's quite ugly though because selections can be set both forward 
> and backward.
> 
> it becomes especially nasty in the case when we are moving pars up 
> because ParIterator is a forward iterator and i need the paragraph 
> before the pars we are moving (am now moving the cursors temporarily 
> backwards).
> 
> suggestions on how to make it nicer are appreciated.
> 
> thanks, edwin
> 
> PS is there a way to undo the moving around?
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> Index: src/LyXAction.C
> ===================================================================
> --- src/LyXAction.C	(revision 13601)
> +++ src/LyXAction.C	(working copy)
> @@ -353,6 +353,8 @@
>  		{ LFUN_MOUSE_RELEASE, "", ReadOnly },
>  		{ LFUN_MOUSE_DOUBLE, "", ReadOnly },
>  		{ LFUN_MOUSE_TRIPLE, "", ReadOnly },
> +		{ LFUN_PARAGRAPH_MOVE_DOWN, "paragraph-move-down", Noop },
> +		{ LFUN_PARAGRAPH_MOVE_UP, "paragraph-move-up", Noop },
>  		{ LFUN_NOACTION, "", Noop }
>  	};
>  
> Index: src/lyxfunc.C
> ===================================================================
> --- src/lyxfunc.C	(revision 13601)
> +++ src/lyxfunc.C	(working copy)
> @@ -558,6 +558,18 @@
>  		flag = getStatus(func);
>  	}
>  
> +	case LFUN_PARAGRAPH_MOVE_UP: {
> +		if (cur)

What does (cur) mean? cur is not a pointer, is it?

> +			enable = cur.pit()>0;
> +		break;
> +	}
> +
> +	case LFUN_PARAGRAPH_MOVE_DOWN: {
> +		if (cur)
> +			enable = cur.pit()<cur.lastpit();
> +		break;
> +	}
> +
>  	case LFUN_MENUNEW:
>  	case LFUN_MENUNEWTMPLT:
>  	case LFUN_WORDFINDFORWARD:
> Index: src/text3.C
> ===================================================================
> --- src/text3.C	(revision 13601)
> +++ src/text3.C	(working copy)
> @@ -44,6 +44,7 @@
>  #include "ParagraphParameters.h"
>  #include "undo.h"
>  #include "vspace.h"
> +#include "pariterator.h"
>  
>  #include "frontends/Dialogs.h"
>  #include "frontends/LyXView.h"
> @@ -320,6 +321,92 @@
>  
>  	switch (cmd.action) {
>  
> +	case LFUN_PARAGRAPH_MOVE_DOWN: {
> +		// 1. FIXME: check wether we need to update counters

Is this comment still valid?

> +		bool counterUpdate = false;
> +		ParIterator pariter;
> +		ParIterator pariter_end;
> +		if (cur.anchor_.pit()<cur.pit()) {
> +			pariter = ParIterator(cur.anchor_);
> +			pariter_end = ParIterator(cur);
> +		} else {
> +			pariter = ParIterator(cur);
> +			pariter_end = ParIterator(cur.anchor_);
> +		}
> +		// include par after range we are moving down
> +		++pariter_end;
> +
> +		// one more to keep the unequal cond in the for loop
> +		++pariter_end;
> +		for (; pariter!=pariter_end; ++pariter) {
> +			if (needsUpdateCounters(cur.buffer(), pariter))
> +				counterUpdate = true;
> +		}

I would put these tests in step 4 as (as you know) they are not only tests.

> +
> +		// 2. move par (range) down
> +		pit_type begpit = cur.selBegin().pit()-1;
> +		for (pit_type pit = cur.selEnd().pit(); pit!=begpit; --pit)
> +			std::swap(pars_[pit], pars_[pit+1]);

Here we could use the splice method I proposed in RandomAccessList but 
Lars rejected that. I suppose that the number of selected paragraphs to 
move won't be too big so that's OK to swap consecutively the 
paragraph... but it is suboptimal. I believe you could also use CutAndPaste.

> +
> +		// 3. realign cursor (and selection)
> +		if (cur.selection())
> +			++cur.anchor_.pit();
> +		++cur.pit();
> +
> +		// 4. update counters
> +		if (counterUpdate) {
> +			lyxerr << "COUNTER UPDATE on DOWN!! " << endl;
> +			updateCounters(cur.buffer());
> +		}
> +
> +		needsUpdate = true;
> +		break;
> +	}

I guess my comments above are the same for the other LFUN so I stop here.

Abdel.

[prev in list] [next in list] [prev in thread] [next in thread]