Nabble - konqueror

Re: Supporting legacy

2008-10-11T15:45:06Z

On Fri, 10 Oct 2008, Germain Garand wrote:

> 'listing' is not aliased to 'pre' in firefox or opera, but it still gets
> treated exactly as a 'pre' in those browsers... so I intend to commit the
> attached to fix the two changes this caused in baseline[1].

Oh, okay. Please commit it then. The test I'll commit for my <image> patch
will make sure that the situation is different for this element.

Harri.

From forum: kfm-devel

Re: programatically import certificate into konqueror

2008-10-10T12:05:14Z

Am Dienstag 07 Oktober 2008 22:38:52 schrieb David Faure:

> On Tuesday 07 October 2008, Helge Deller wrote:
> > David Faure wrote:
> > > On Monday 29 September 2008, Yifan Wang wrote:
> > >> hi
> > >>
> > >> In KDE3, there is dcop that we can use to import user's
> > >> certificate(p12 format) automatically into konqueror.
> > >> But in KDE4, dcop is replaced by dbus and does not work anymore.
> > >> I have search the method to do this but so far have no result.
> > >
> > > Can you remind us which DCOP call this is about exactly?
> >
> > Running "dcop kded kssld" in a shell on KDE3 returns quite some
> > functions which are not yet available with KDE4.
> >
> > The important ones which Yifan refers to, are:
> > - bool addHomeCertificateFile(QString filename,QString password,
> > bool storePass)
> > - bool addHomeCertificatePKCS12(QString base64cert,QString passToStore)
> > - bool deleteHomeCertificateByFile(QString filename,QString password)
> > - bool deleteHomeCertificateByPKCS12(QString base64cert,
> > QString password)
> > - bool deleteHomeCertificateByName(QString name)
> >
> > It's probably a good idea to add those functions to KDE4.
>
> I see - those methods have been removed when Andreas merged in the "newssl"
> branch.
>
> However I see the implementation is still available, in
> kio/kssl/ksslcertificatehome.{cpp,h} Andreas: should I re-add the wrappers
> for those methods in kssld, or did you remove them because the switch to
> the new ssl stuff prevents those methods from working?

ksslcertificatehome is pretty much dead code, and KIO::TCPSlaveBase does not
support client certificates for now. Sorry, I didn't have enough time at the
time to do it.

From forum: kfm-devel

Re: Supporting legacy

2008-10-10T08:52:56Z

Le dimanche 5 octobre 2008, Harri Porten a écrit :

> Hi!
>
> Some reworking of our tag name representation (r867535) broke support for
> legacy HTML elements. The question is do which degree we want to fix or
> drop that. The misc/maketags made the following elements equivalent
>
> image -> img
> anchor -> a
> listing -> pre
>
> for pre HTML 4 compat. I've put up a test on
> http://www.froglogic.com/~porten/compat.html that tests both HTML parsing
> and createElement() DOM calls for those three. Results:
>
> - neither Firefox, IE nor Opera special case "anchor" and "listing". I
> was not able to find any sign of <anchor> btw.
>

'listing' is not aliased to 'pre' in firefox or opera, but it still gets
treated exactly as a 'pre' in those browsers... so I intend to commit the
attached to fix the two changes this caused in baseline[1].

Germain

[1] webcore/fast/html/listing.html
basics/listing.html

[0001-reinstate-handling-of-legacy-listing-tag.patch]

From 729a931f55b15a47d652523ef2e26727a35ac3b2 Mon Sep 17 00:00:00 2001 From: Germain Garand <germain@...> Date: Fri, 10 Oct 2008 17:44:26 +0200 Subject: [PATCH] reinstate handling of legacy 'listing' tag, basically duplicating 'pre' handling code path. diff --git a/khtml/css/html4.css b/khtml/css/html4.css index cffe228..1a916d4 100644 --- a/khtml/css/html4.css +++ b/khtml/css/html4.css @@ -459,6 +459,7 @@ samp { } pre, +listing, xmp, plaintext { display: block; @@ -528,7 +529,7 @@ bdo[dir="rtl"] { address, blockquote, body, dd, div, dl, dt, fieldset, form, frame, frameset, h1, h2, h3, h4, h5, h6, iframe, noscript, noframes, object, ol, p, ul, applet, center, - dir, hr, menu, pre, li, table, tr, thead, tbody, tfoot, + dir, hr, menu, pre, listing, li, table, tr, thead, tbody, tfoot, col, colgroup, td, th, caption { unicode-bidi: embed } */ diff --git a/khtml/css/presentational.css b/khtml/css/presentational.css index 8d6ca77..586bf5c 100644 --- a/khtml/css/presentational.css +++ b/khtml/css/presentational.css @@ -24,7 +24,7 @@ * */ -pre[wrap] { +pre[wrap], listing[wrap] { white-space: pre-wrap } diff --git a/khtml/html/dtd.cpp b/khtml/html/dtd.cpp index 38da3b2..7afbebf 100644 --- a/khtml/html/dtd.cpp +++ b/khtml/html/dtd.cpp @@ -105,6 +105,7 @@ const unsigned short KDE_NO_EXPORT DOM::tagPriorityArray[] = { 1, // ID_LAYER 1, // ID_LEGEND 5, // ID_LI + 5, // ID_LISTING 0, // ID_LINK 1, // ID_MAP 5, // ID_MARQUEE @@ -217,6 +218,7 @@ const tagStatus DOM::endTagArray[] = { REQUIRED, // ID_LEGEND OPTIONAL, // ID_LI FORBIDDEN, // ID_LINK + REQUIRED, // ID_LISTING REQUIRED, // ID_MAP REQUIRED, // ID_MARQUEE REQUIRED, // ID_MENU @@ -349,6 +351,7 @@ static const ushort tag_list_block[] = { ID_DIR, // legacy ID_MENU, // legacy // %blocktext + ID_LISTING, ID_PRE, ID_HR, ID_BLOCKQUOTE, @@ -554,6 +557,7 @@ bool DOM::checkChild(ushort tagID, ushort childID, bool strict) case ID_PRE: case ID_XMP: case ID_PLAINTEXT: + case ID_LISTING: // PRE: %flow * - _5 return check_flow(childID, true); case ID_DL: @@ -665,8 +669,10 @@ void DOM::addForbidden(int tagId, ushort *forbiddenTags) break; case ID_NOBR: forbiddenTags[ID_PRE]++; + forbiddenTags[ID_LISTING]++; // fall through case ID_PRE: + case ID_LISTING: case ID_PLAINTEXT: case ID_XMP: //forbiddenTags[ID_IMG]++; @@ -709,8 +715,10 @@ void DOM::removeForbidden(int tagId, ushort *forbiddenTags) break; case ID_NOBR: forbiddenTags[ID_PRE]--; + forbiddenTags[ID_LISTING]--; // fall through case ID_PRE: + case ID_LISTING: case ID_XMP: case ID_PLAINTEXT: //forbiddenTags[ID_IMG]--; diff --git a/khtml/html/htmlparser.cpp b/khtml/html/htmlparser.cpp index 94bfd5d..a2c5d92 100644 --- a/khtml/html/htmlparser.cpp +++ b/khtml/html/htmlparser.cpp @@ -1064,6 +1064,7 @@ NodeImpl *KHTMLParser::getElement(Token* t) case ID_PRE: case ID_XMP: case ID_PLAINTEXT: + case ID_LISTING: n = new HTMLPreElementImpl(document, t->tid); break; @@ -1366,6 +1367,7 @@ bool KHTMLParser::isAffectedByResidualStyle(int _id) case ID_DT: case ID_DD: case ID_PRE: + case ID_LISTING: return true; default: return false; diff --git a/khtml/html/htmltokenizer.cpp b/khtml/html/htmltokenizer.cpp index 011a194..0b7abd6 100644 --- a/khtml/html/htmltokenizer.cpp +++ b/khtml/html/htmltokenizer.cpp @@ -1530,6 +1530,7 @@ void HTMLTokenizer::parseTag(TokenizerString &src) discard = AllDiscard; switch( tagID ) { + case ID_LISTING: case ID_PRE: pre = beginTag; if (beginTag) diff --git a/khtml/misc/htmlnames.cpp b/khtml/misc/htmlnames.cpp index fbab9ec..12a45db 100644 --- a/khtml/misc/htmlnames.cpp +++ b/khtml/misc/htmlnames.cpp @@ -80,6 +80,7 @@ IDTable<LocalNameFactory>* LocalNameFactory::initIdTable() s_idTable->addStaticMapping(localNamePart(ID_LEGEND), "legend"); s_idTable->addStaticMapping(localNamePart(ID_LI), "li"); s_idTable->addStaticMapping(localNamePart(ID_LINK), "link"); + s_idTable->addStaticMapping(localNamePart(ID_LISTING), "listing"); s_idTable->addStaticMapping(localNamePart(ID_MAP), "map"); s_idTable->addStaticMapping(localNamePart(ID_MARQUEE), "marquee"); s_idTable->addStaticMapping(localNamePart(ID_MENU), "menu"); diff --git a/khtml/misc/htmlnames.h b/khtml/misc/htmlnames.h index 9f8866d..4b64c1f 100644 --- a/khtml/misc/htmlnames.h +++ b/khtml/misc/htmlnames.h @@ -77,209 +77,210 @@ #define ID_LI 59 #define ID_LINK 60 #define ATTR_LINK 65596 -#define ID_MAP 61 -#define ID_MARQUEE 62 -#define ID_MENU 63 -#define ID_META 64 -#define ID_NOBR 65 -#define ID_NOEMBED 66 -#define ID_NOFRAMES 67 -#define ID_NOSCRIPT 68 -#define ID_NOLAYER 69 -#define ID_OBJECT 70 -#define ATTR_OBJECT 65606 -#define ID_OL 71 -#define ID_OPTGROUP 72 -#define ID_OPTION 73 -#define ID_P 74 -#define ID_PARAM 75 -#define ID_PLAINTEXT 76 -#define ID_PRE 77 -#define ID_Q 78 -#define ID_S 79 -#define ID_SAMP 80 -#define ID_SCRIPT 81 -#define ID_SELECT 82 -#define ID_SMALL 83 -#define ID_SOURCE 84 -#define ID_SPAN 85 -#define ATTR_SPAN 65621 -#define ID_STRIKE 86 -#define ID_STRONG 87 -#define ID_STYLE 88 -#define ATTR_STYLE 65624 -#define ID_SUB 89 -#define ID_SUP 90 -#define ID_TABLE 91 -#define ID_TBODY 92 -#define ID_TD 93 -#define ID_TEXTAREA 94 -#define ID_TFOOT 95 -#define ID_TH 96 -#define ID_THEAD 97 -#define ID_TITLE 98 -#define ATTR_TITLE 65634 -#define ID_TR 99 -#define ID_TT 100 -#define ID_U 101 -#define ID_UL 102 -#define ID_VAR 103 -#define ID_VIDEO 104 -#define ID_WBR 105 -#define ID_XMP 106 -#define ID_TEXT 107 -#define ID_COMMENT 108 -#define ATTR_ACCEPT_CHARSET 65645 -#define ATTR_ACCEPT 65646 -#define ATTR_ACCESSKEY 65647 -#define ATTR_ALIGN 65648 -#define ATTR_ALINK 65649 -#define ATTR_AUTOCOMPLETE 65650 -#define ATTR_AUTOPLAY 65651 -#define ATTR_AXIS 65652 -#define ATTR_BEHAVIOR 65653 -#define ATTR_BGCOLOR 65654 -#define ATTR_BGPROPERTIES 65655 -#define ATTR_BORDER 65656 -#define ATTR_BORDERCOLOR 65657 -#define ATTR_CELLPADDING 65658 -#define ATTR_CELLSPACING 65659 -#define ATTR_CHAR 65660 -#define ATTR_CHALLENGE 65661 -#define ATTR_CHAROFF 65662 -#define ATTR_CHARSET 65663 -#define ATTR_CHECKED 65664 -#define ATTR_CLEAR 65665 -#define ATTR_CODETYPE 65666 -#define ATTR_COLOR 65667 -#define ATTR_COLS 65668 -#define ATTR_COLSPAN 65669 -#define ATTR_COMPACT 65670 -#define ATTR_CONTENTEDITABLE 65671 -#define ATTR_CONTROLS 65672 -#define ATTR_COORDS 65673 -#define ATTR_DECLARE 65674 -#define ATTR_DEFER 65675 -#define ATTR_DIRECTION 65676 -#define ATTR_DISABLED 65677 -#define ATTR_ENCTYPE 65678 -#define ATTR_END 65679 -#define ATTR_FACE 65680 -#define ATTR_FRAMEBORDER 65681 -#define ATTR_HEIGHT 65682 -#define ATTR_HIDDEN 65683 -#define ATTR_HREFLANG 65684 -#define ATTR_HSPACE 65685 -#define ATTR_HTTP_EQUIV 65686 -#define ATTR_ISMAP 65687 -#define ATTR_LANG 65688 -#define ATTR_LANGUAGE 65689 -#define ATTR_LEFT 65690 -#define ATTR_LEFTMARGIN 65691 -#define ATTR_LOOP 65692 -#define ATTR_LOOPEND 65693 -#define ATTR_LOOPSTART 65694 -#define ATTR_MARGINHEIGHT 65695 -#define ATTR_MARGINWIDTH 65696 -#define ATTR_MAXLENGTH 65697 -#define ATTR_MEDIA 65698 -#define ATTR_METHOD 65699 -#define ATTR_MULTIPLE 65700 -#define ATTR_NOHREF 65701 -#define ATTR_NORESIZE 65702 -#define ATTR_NOSAVE 65703 -#define ATTR_NOSHADE 65704 -#define ATTR_NOWRAP 65705 -#define ATTR_ONABORT 65706 -#define ATTR_ONERROR 65707 -#define ATTR_ONRESIZE 65708 -#define ATTR_OVERSRC 65709 -#define ATTR_PAGEX 65710 -#define ATTR_PAGEY 65711 -#define ATTR_PLAIN 65712 -#define ATTR_PLAYCOUNT 65713 -#define ATTR_PLUGINPAGE 65714 -#define ATTR_PLUGINSPAGE 65715 -#define ATTR_PLUGINURL 65716 -#define ATTR_POSTER 65717 -#define ATTR_READONLY 65718 -#define ATTR_REL 65719 -#define ATTR_REV 65720 -#define ATTR_ROWS 65721 -#define ATTR_ROWSPAN 65722 -#define ATTR_RULES 65723 -#define ATTR_SCOPE 65724 -#define ATTR_SCROLLAMOUNT 65725 -#define ATTR_SCROLLDELAY 65726 -#define ATTR_SCROLLING 65727 -#define ATTR_SELECTED 65728 -#define ATTR_SHAPE 65729 -#define ATTR_SIZE 65730 -#define ATTR_START 65731 -#define ATTR_TABINDEX 65732 -#define ATTR_TARGET 65733 -#define ATTR_TEXT 65734 -#define ATTR_TOP 65735 -#define ATTR_TOPMARGIN 65736 -#define ATTR_TRUESPEED 65737 -#define ATTR_TYPE 65738 -#define ATTR_UNKNOWN 65739 -#define ATTR_VALIGN 65740 -#define ATTR_VALUETYPE 65741 -#define ATTR_VERSION 65742 -#define ATTR_VISIBILITY 65743 -#define ATTR_VLINK 65744 -#define ATTR_VSPACE 65745 -#define ATTR_WIDTH 65746 -#define ATTR_WRAP 65747 -#define ATTR_Z_INDEX 65748 -#define ATTR_ACTION 65749 -#define ATTR_ALT 65750 -#define ATTR_ARCHIVE 65751 -#define ATTR_BACKGROUND 65752 -#define ATTR_CLASS 65753 -#define ATTR_CLASSID 65754 -#define ATTR_CODEBASE 65755 -#define ATTR_CONTENT 65756 -#define ATTR_DATA 65757 -#define ATTR_DATETIME 65758 -#define ATTR_FOR 65759 -#define ATTR_HEADERS 65760 -#define ATTR_HREF 65761 -#define ATTR_ID 65762 -#define ATTR_LONGDESC 65763 -#define ATTR_NAME 65764 -#define ATTR_ONBLUR 65765 -#define ATTR_ONCHANGE 65766 -#define ATTR_ONCLICK 65767 -#define ATTR_ONDBLCLICK 65768 -#define ATTR_ONFOCUS 65769 -#define ATTR_ONKEYDOWN 65770 -#define ATTR_ONKEYPRESS 65771 -#define ATTR_ONKEYUP 65772 -#define ATTR_ONLOAD 65773 -#define ATTR_ONMOUSEDOWN 65774 -#define ATTR_ONMOUSEMOVE 65775 -#define ATTR_ONMOUSEOUT 65776 -#define ATTR_ONMOUSEOVER 65777 -#define ATTR_ONMOUSEUP 65778 -#define ATTR_ONRESET 65779 -#define ATTR_ONSELECT 65780 -#define ATTR_ONSCROLL 65781 -#define ATTR_ONSUBMIT 65782 -#define ATTR_ONUNLOAD 65783 -#define ATTR_PROFILE 65784 -#define ATTR_PROMPT 65785 -#define ATTR_SCHEME 65786 -#define ATTR_SRC 65787 -#define ATTR_STANDBY 65788 -#define ATTR_SUMMARY 65789 -#define ATTR_USEMAP 65790 -#define ATTR_VALUE 65791 -#define ID_LAST_TAG 108 +#define ID_LISTING 61 +#define ID_MAP 62 +#define ID_MARQUEE 63 +#define ID_MENU 64 +#define ID_META 65 +#define ID_NOBR 66 +#define ID_NOEMBED 67 +#define ID_NOFRAMES 68 +#define ID_NOSCRIPT 69 +#define ID_NOLAYER 70 +#define ID_OBJECT 71 +#define ATTR_OBJECT 65607 +#define ID_OL 72 +#define ID_OPTGROUP 73 +#define ID_OPTION 74 +#define ID_P 75 +#define ID_PARAM 76 +#define ID_PLAINTEXT 77 +#define ID_PRE 78 +#define ID_Q 79 +#define ID_S 80 +#define ID_SAMP 81 +#define ID_SCRIPT 82 +#define ID_SELECT 83 +#define ID_SMALL 84 +#define ID_SOURCE 85 +#define ID_SPAN 86 +#define ATTR_SPAN 65622 +#define ID_STRIKE 87 +#define ID_STRONG 88 +#define ID_STYLE 89 +#define ATTR_STYLE 65625 +#define ID_SUB 90 +#define ID_SUP 91 +#define ID_TABLE 92 +#define ID_TBODY 93 +#define ID_TD 94 +#define ID_TEXTAREA 95 +#define ID_TFOOT 96 +#define ID_TH 97 +#define ID_THEAD 98 +#define ID_TITLE 99 +#define ATTR_TITLE 65635 +#define ID_TR 100 +#define ID_TT 101 +#define ID_U 102 +#define ID_UL 103 +#define ID_VAR 104 +#define ID_VIDEO 105 +#define ID_WBR 106 +#define ID_XMP 107 +#define ID_TEXT 108 +#define ID_COMMENT 109 +#define ATTR_ACCEPT_CHARSET 65646 +#define ATTR_ACCEPT 65647 +#define ATTR_ACCESSKEY 65648 +#define ATTR_ALIGN 65649 +#define ATTR_ALINK 65650 +#define ATTR_AUTOCOMPLETE 65651 +#define ATTR_AUTOPLAY 65652 +#define ATTR_AXIS 65653 +#define ATTR_BEHAVIOR 65654 +#define ATTR_BGCOLOR 65655 +#define ATTR_BGPROPERTIES 65656 +#define ATTR_BORDER 65657 +#define ATTR_BORDERCOLOR 65658 +#define ATTR_CELLPADDING 65659 +#define ATTR_CELLSPACING 65660 +#define ATTR_CHAR 65661 +#define ATTR_CHALLENGE 65662 +#define ATTR_CHAROFF 65663 +#define ATTR_CHARSET 65664 +#define ATTR_CHECKED 65665 +#define ATTR_CLEAR 65666 +#define ATTR_CODETYPE 65667 +#define ATTR_COLOR 65668 +#define ATTR_COLS 65669 +#define ATTR_COLSPAN 65670 +#define ATTR_COMPACT 65671 +#define ATTR_CONTENTEDITABLE 65672 +#define ATTR_CONTROLS 65673 +#define ATTR_COORDS 65674 +#define ATTR_DECLARE 65675 +#define ATTR_DEFER 65676 +#define ATTR_DIRECTION 65677 +#define ATTR_DISABLED 65678 +#define ATTR_ENCTYPE 65679 +#define ATTR_END 65680 +#define ATTR_FACE 65681 +#define ATTR_FRAMEBORDER 65682 +#define ATTR_HEIGHT 65683 +#define ATTR_HIDDEN 65684 +#define ATTR_HREFLANG 65685 +#define ATTR_HSPACE 65686 +#define ATTR_HTTP_EQUIV 65687 +#define ATTR_ISMAP 65688 +#define ATTR_LANG 65689 +#define ATTR_LANGUAGE 65690 +#define ATTR_LEFT 65691 +#define ATTR_LEFTMARGIN 65692 +#define ATTR_LOOP 65693 +#define ATTR_LOOPEND 65694 +#define ATTR_LOOPSTART 65695 +#define ATTR_MARGINHEIGHT 65696 +#define ATTR_MARGINWIDTH 65697 +#define ATTR_MAXLENGTH 65698 +#define ATTR_MEDIA 65699 +#define ATTR_METHOD 65700 +#define ATTR_MULTIPLE 65701 +#define ATTR_NOHREF 65702 +#define ATTR_NORESIZE 65703 +#define ATTR_NOSAVE 65704 +#define ATTR_NOSHADE 65705 +#define ATTR_NOWRAP 65706 +#define ATTR_ONABORT 65707 +#define ATTR_ONERROR 65708 +#define ATTR_ONRESIZE 65709 +#define ATTR_OVERSRC 65710 +#define ATTR_PAGEX 65711 +#define ATTR_PAGEY 65712 +#define ATTR_PLAIN 65713 +#define ATTR_PLAYCOUNT 65714 +#define ATTR_PLUGINPAGE 65715 +#define ATTR_PLUGINSPAGE 65716 +#define ATTR_PLUGINURL 65717 +#define ATTR_POSTER 65718 +#define ATTR_READONLY 65719 +#define ATTR_REL 65720 +#define ATTR_REV 65721 +#define ATTR_ROWS 65722 +#define ATTR_ROWSPAN 65723 +#define ATTR_RULES 65724 +#define ATTR_SCOPE 65725 +#define ATTR_SCROLLAMOUNT 65726 +#define ATTR_SCROLLDELAY 65727 +#define ATTR_SCROLLING 65728 +#define ATTR_SELECTED 65729 +#define ATTR_SHAPE 65730 +#define ATTR_SIZE 65731 +#define ATTR_START 65732 +#define ATTR_TABINDEX 65733 +#define ATTR_TARGET 65734 +#define ATTR_TEXT 65735 +#define ATTR_TOP 65736 +#define ATTR_TOPMARGIN 65737 +#define ATTR_TRUESPEED 65738 +#define ATTR_TYPE 65739 +#define ATTR_UNKNOWN 65740 +#define ATTR_VALIGN 65741 +#define ATTR_VALUETYPE 65742 +#define ATTR_VERSION 65743 +#define ATTR_VISIBILITY 65744 +#define ATTR_VLINK 65745 +#define ATTR_VSPACE 65746 +#define ATTR_WIDTH 65747 +#define ATTR_WRAP 65748 +#define ATTR_Z_INDEX 65749 +#define ATTR_ACTION 65750 +#define ATTR_ALT 65751 +#define ATTR_ARCHIVE 65752 +#define ATTR_BACKGROUND 65753 +#define ATTR_CLASS 65754 +#define ATTR_CLASSID 65755 +#define ATTR_CODEBASE 65756 +#define ATTR_CONTENT 65757 +#define ATTR_DATA 65758 +#define ATTR_DATETIME 65759 +#define ATTR_FOR 65760 +#define ATTR_HEADERS 65761 +#define ATTR_HREF 65762 +#define ATTR_ID 65763 +#define ATTR_LONGDESC 65764 +#define ATTR_NAME 65765 +#define ATTR_ONBLUR 65766 +#define ATTR_ONCHANGE 65767 +#define ATTR_ONCLICK 65768 +#define ATTR_ONDBLCLICK 65769 +#define ATTR_ONFOCUS 65770 +#define ATTR_ONKEYDOWN 65771 +#define ATTR_ONKEYPRESS 65772 +#define ATTR_ONKEYUP 65773 +#define ATTR_ONLOAD 65774 +#define ATTR_ONMOUSEDOWN 65775 +#define ATTR_ONMOUSEMOVE 65776 +#define ATTR_ONMOUSEOUT 65777 +#define ATTR_ONMOUSEOVER 65778 +#define ATTR_ONMOUSEUP 65779 +#define ATTR_ONRESET 65780 +#define ATTR_ONSELECT 65781 +#define ATTR_ONSCROLL 65782 +#define ATTR_ONSUBMIT 65783 +#define ATTR_ONUNLOAD 65784 +#define ATTR_PROFILE 65785 +#define ATTR_PROMPT 65786 +#define ATTR_SCHEME 65787 +#define ATTR_SRC 65788 +#define ATTR_STANDBY 65789 +#define ATTR_SUMMARY 65790 +#define ATTR_USEMAP 65791 +#define ATTR_VALUE 65792 +#define ID_LAST_TAG 109 #define ID_CLOSE_TAG 16384 -#define ATTR_LAST_ATTR 255 -#define ATTR_LAST_CI_ATTR 212 +#define ATTR_LAST_ATTR 256 +#define ATTR_LAST_CI_ATTR 213 #define caseSensitiveAttr(id) (((localNamePart(id)) > ATTR_LAST_CI_ATTR || (id) == ATTR_ABBR || (id) == ATTR_CITE || (id) == ATTR_CODE || (id) == ATTR_LABEL || (id) == ATTR_OBJECT || (id) == ATTR_TITLE)) diff --git a/khtml/misc/htmltags.in b/khtml/misc/htmltags.in index 8692720..63a8852 100644 --- a/khtml/misc/htmltags.in +++ b/khtml/misc/htmltags.in @@ -58,6 +58,7 @@ layer legend li link +listing map marquee menu diff --git a/khtml/xml/dom_docimpl.cpp b/khtml/xml/dom_docimpl.cpp index d41d2b4..6692f16 100644 --- a/khtml/xml/dom_docimpl.cpp +++ b/khtml/xml/dom_docimpl.cpp @@ -972,6 +972,7 @@ ElementImpl *DocumentImpl::createHTMLElement( const DOMString &name ) case ID_PLAINTEXT: case ID_XMP: case ID_PRE: + case ID_LISTING: n = new HTMLPreElementImpl(docPtr(), id); break; -- 1.5.3.2

From forum: kfm-devel

Re: Please comment: ComboBoxWidget change

2008-10-09T16:01:38Z

Le mardi 7 octobre 2008, David Faure a écrit :
> On Sunday 05 October 2008, Germain Garand wrote:
> > unfortunately, QComboBox doesn't seem to offer the necessary hooks to the
> > actual popup widget, so we cannot set the position we want.
>
> Which kind of hook do you need? The popup widget is accessible as view().
> Knowing when it's going to show up could be done with an event filter for
> ShowEvent on that view widget.

I don't remember exactly... the popup is actually a private container for the
view. It is certainly possible to walk to it from view(), but that isn't the
problem.
IIRC the QComboBox::showPopup virtual is actually 200 loc doing all kind of
fancy sizing calculations with some private methods, that won't work on our
offscreen widgets.

So the bottom line is, it probably is possible to do it, at the expense of
reimplementing the functionality of those 200 loc.

Germain

From forum: kfm-devel

[Patch] Smoother smooth-scrolling in khtml

2008-10-08T09:56:46Z

Hello to the list,

I have made a patch that effects khtml's smooth scrolling. Currently, if you start scrolling, it steps to maximum scrolling speed, to nicely decelerate. But the step to maximum speed isn't easy to follow by your eyes.

With this patch, it keeps the current speed (which is zero, if you start scrolling), so it accelerate first to maximum speed and then decelerate. Which is easier to follow with your eyes (especially if you scroll with your mouse wheel, which is much smoother with this patch).

Like to hear what you think about it :)

Manuel Mommertz

[khtml_smoothscolling.patch]

Index: kdelibs/khtml/khtmlview.cpp =================================================================== --- kdelibs/khtml/khtmlview.cpp (Revision 869251) +++ kdelibs/khtml/khtmlview.cpp (Arbeitskopie) @@ -93,6 +93,8 @@ #include <QtGui/QPrinter> #include <QtGui/QPrintDialog> +#include <cmath> + //#define DEBUG_FLICKER //#define DEBUG_PIXEL @@ -127,7 +129,7 @@ static const int sLayoutAttemptIncrement = 20; static const int sParsingLayoutsIncrement = 60; -static const int sSmoothScrollTime = 140; +static const int sSmoothScrollTime = 280; static const int sSmoothScrollTick = 14; class KHTMLViewPrivate { @@ -228,7 +230,9 @@ scrollingFromWheel = QPoint(-1,-1); borderX = 30; borderY = 30; - dx = dy = ddx = ddy = rdx = rdy = dddx = dddy = 0; + dx = dy = 0; + ax = ay = bx = by = 0.; + stepsX = stepsY = 0; paged = false; clickX = -1; clickY = -1; @@ -346,9 +350,7 @@ { smoothScrollTimer.stop(); dx = dy = 0; - ddx = ddy = 0; - rdx = rdy = 0; - dddx = dddy = 0; + stepsX = stepsY = 0; updateContentsXY(); smoothScrolling = false; shouldSmoothScroll = false; @@ -412,7 +414,9 @@ int zoomLevel; int borderX, borderY; - int dx, dy, ddx, ddy, rdx, rdy, dddx, dddy; + int dx, dy; + double ax, ay, bx, by; + int stepsX, stepsY; KConfig *formCompletions; int clickX, clickY, clickCount; @@ -3999,39 +4003,90 @@ void KHTMLView::setupSmoothScrolling(int dx, int dy) { - // full scroll is remaining scroll plus new scroll - d->dx = d->dx + dx; - d->dy = d->dy + dy; + int maxSteps = sSmoothScrollTime/sSmoothScrollTick; + int &stepsX = d->stepsX, &stepsY = d->stepsY; + double &ax = d->ax, &ay = d->ay; + double &bx = d->bx, &by = d->by; - if (d->dx == 0 && d->dy == 0) return; + double maxPoint, maxValue; + double maxAllowedX, maxAllowedY; - int steps = sSmoothScrollTime/sSmoothScrollTick; + // full scroll is new scroll plus remaining scroll + double x = dx + (bx - ax*stepsX) * stepsX * stepsX; + double y = dy + (by - ay*stepsY) * stepsY * stepsY; + d->dx += dx; + d->dy += dy; - // average step size (stored in 1/16 px/step) - d->ddx = (d->dx*16)/(steps+1); - d->ddy = (d->dy*16)/(steps+1); + // get cur speed + double sx = (3*ax*stepsX - 2*bx) * stepsX; + double sy = (3*ay*stepsY - 2*by) * stepsY; - if (abs(d->ddx) < 64 && abs(d->ddy) < 64) { - // Don't move slower than average 4px/step in minimum one direction - if (d->ddx > 0) d->ddx = qMax(d->ddx, 64); - if (d->ddy > 0) d->ddy = qMax(d->ddy, 64); - if (d->ddx < 0) d->ddx = qMin(d->ddx, -64); - if (d->ddy < 0) d->ddy = qMin(d->ddy, -64); - // This means fewer than normal steps - steps = qMax(d->ddx ? (d->dx*16)/d->ddx : 0, d->ddy ? (d->dy*16)/d->ddy : 0); - if (steps < 1) steps = 1; - d->ddx = (d->dx*16)/(steps+1); - d->ddy = (d->dy*16)/(steps+1); + if (sx) { + // save maximum for X + if (ax*bx > 0) { + maxPoint = 2*bx / (3*ax); + if(maxPoint > stepsX) maxPoint = stepsX; + } else { + maxPoint = stepsX; + } + maxAllowedX = (bx - ax*maxPoint) * maxPoint * maxPoint + dx; + + if (sx > 0) { + maxAllowedX = qMax(maxAllowedX, (double)qMax(dx, 0)); + } else { + maxAllowedX = qMin(maxAllowedX, (double)qMin(dx, 0)); + } } - // step size starts at double average speed and ends at 0 - d->ddx *= 2; - d->ddy *= 2; + if (sy) { + // save maximum for Y + if (ay*by > 0) { + maxPoint = 2*by / (3*ay); + if(maxPoint > stepsY) maxPoint = stepsY; + } else { + maxPoint = stepsY; + } + maxAllowedY = (by - ay*maxPoint) * maxPoint * maxPoint + dy; - // deacceleration speed - d->dddx = (d->ddx+1)/steps; - d->dddy = (d->ddy+1)/steps; + if (sy > 0) { + maxAllowedY = qMax(maxAllowedY, (double)qMax(dy, 0)); + } else { + maxAllowedY = qMin(maxAllowedY, (double)qMin(dy, 0)); + } + } + stepsX = stepsY = maxSteps; + + ax = (2*x + sx*stepsX) / (stepsX*stepsX*stepsX); + ay = (2*y + sy*stepsY) / (stepsY*stepsY*stepsY); + bx = x/(stepsX*stepsX) + ax*stepsX; + by = y/(stepsY*stepsY) + ay*stepsY; + + if (ax*bx > 0 && sx) { + maxPoint = 2*bx / (3*ax); + if (maxPoint < stepsX) { + maxValue = (bx - ax*maxPoint) * maxPoint * maxPoint; + if (abs(maxValue) > abs(maxAllowedX)) { + stepsX = qMax(int(sqrt(maxAllowedX/maxValue) * stepsX), + qMax(int(-2*x/sx), 1)); + ax = (2*x + sx*stepsX) / (stepsX*stepsX*stepsX); + bx = x/(stepsX*stepsX) + ax*stepsX; + } + } + } + if (ay*by > 0 && sy) { + maxPoint = 2*by / (3*ay); + if (maxPoint < stepsY) { + maxValue = (by - ay*maxPoint) * maxPoint * maxPoint; + if (abs(maxValue) > abs(maxAllowedY)) { + stepsY = qMax(int(sqrt(maxAllowedY/maxValue) * stepsY), + qMax(int(-2*y/sy), 1)); + ay = (2*y + sy*stepsY) / (stepsY*stepsY*stepsY); + by = y/(stepsY*stepsY) + ay*stepsY; + } + } + } + if (!d->smoothScrolling) { d->startScrolling(); scrollTick(); @@ -4039,52 +4094,25 @@ } void KHTMLView::scrollTick() { - if (d->dx == 0 && d->dy == 0) { + int &stepsX = d->stepsX, &stepsY = d->stepsY; + if (!stepsX && !stepsY ) { d->stopScrolling(); return; } - // step size + remaining partial step - int tddx = d->ddx + d->rdx; - int tddy = d->ddy + d->rdy; + if(stepsX) --stepsX; + if(stepsY) --stepsY; - // don't go under 1px/step - if (tddx > 0 && tddx < 16) tddx = 16; - if (tddy > 0 && tddy < 16) tddy = 16; - if (tddx < 0 && tddx > -16) tddx = -16; - if (tddy < 0 && tddy > -16) tddy = -16; + // pixel steps to scroll in this step + int ddx = d->dx - int((d->bx - d->ax*stepsX)*stepsX*stepsX); + int ddy = d->dy - int((d->by - d->ay*stepsY)*stepsY*stepsY); - // full pixel steps to scroll in this step - int ddx = tddx / 16; - int ddy = tddy / 16; - // remaining partial step (this is especially needed for 1.x sized steps) - d->rdx = tddx % 16; - d->rdy = tddy % 16; - - // limit step to requested scrolling distance - if (abs(ddx) > abs(d->dx)) ddx = d->dx; - if (abs(ddy) > abs(d->dy)) ddy = d->dy; - - // Don't stop if deaccelerated too fast - if (!ddx) ddx = d->dx; - if (!ddy) ddy = d->dy; - // update remaining scroll d->dx -= ddx; d->dy -= ddy; d->shouldSmoothScroll = false; scrollContentsBy(ddx, ddy); - - // update scrolling speed - int dddx = d->dddx; - int dddy = d->dddy; - // don't change direction - if (abs(dddx) > abs(d->ddx)) dddx = d->ddx; - if (abs(dddy) > abs(d->ddy)) dddy = d->ddy; - - d->ddx -= dddx; - d->ddy -= dddy; }

From forum: kfm-devel

Re: programatically import certificate into konqueror

2008-10-07T13:38:52Z

On Tuesday 07 October 2008, Helge Deller wrote:

> David Faure wrote:
> > On Monday 29 September 2008, Yifan Wang wrote:
> >> hi
> >>
> >> In KDE3, there is dcop that we can use to import user's certificate(p12
> >> format) automatically into konqueror.
> >> But in KDE4, dcop is replaced by dbus and does not work anymore.
> >> I have search the method to do this but so far have no result.
> >
> > Can you remind us which DCOP call this is about exactly?
>
> Running "dcop kded kssld" in a shell on KDE3 returns quite some
> functions which are not yet available with KDE4.
>
> The important ones which Yifan refers to, are:
> - bool addHomeCertificateFile(QString filename,QString password,
> bool storePass)
> - bool addHomeCertificatePKCS12(QString base64cert,QString passToStore)
> - bool deleteHomeCertificateByFile(QString filename,QString password)
> - bool deleteHomeCertificateByPKCS12(QString base64cert,
> QString password)
> - bool deleteHomeCertificateByName(QString name)
>
> It's probably a good idea to add those functions to KDE4.

I see - those methods have been removed when Andreas merged in the "newssl" branch.

However I see the implementation is still available, in kio/kssl/ksslcertificatehome.{cpp,h}
Andreas: should I re-add the wrappers for those methods in kssld, or did you remove
them because the switch to the new ssl stuff prevents those methods from working?

--
David Faure, faure@..., sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).

From forum: kfm-devel

Re: programatically import certificate into konqueror

2008-10-07T11:14:51Z

David Faure wrote:
> On Monday 29 September 2008, Yifan Wang wrote:
>> hi
>>
>> In KDE3, there is dcop that we can use to import user's certificate(p12
>> format) automatically into konqueror.
>> But in KDE4, dcop is replaced by dbus and does not work anymore.
>> I have search the method to do this but so far have no result.
>
> Can you remind us which DCOP call this is about exactly?

Running "dcop kded kssld" in a shell on KDE3 returns quite some
functions which are not yet available with KDE4.

The important ones which Yifan refers to, are:
- bool addHomeCertificateFile(QString filename,QString password,
bool storePass)
- bool addHomeCertificatePKCS12(QString base64cert,QString passToStore)
- bool deleteHomeCertificateByFile(QString filename,QString password)
- bool deleteHomeCertificateByPKCS12(QString base64cert,
QString password)
- bool deleteHomeCertificateByName(QString name)

It's probably a good idea to add those functions to KDE4.

Helge

From forum: kfm-devel

Re: programatically import certificate into konqueror

2008-10-07T07:11:10Z

On Monday 29 September 2008, Yifan Wang wrote:
> hi
>
> In KDE3, there is dcop that we can use to import user's certificate(p12
> format) automatically into konqueror.
> But in KDE4, dcop is replaced by dbus and does not work anymore.
> I have search the method to do this but so far have no result.

Can you remind us which DCOP call this is about exactly?

--
David Faure, faure@..., sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).

From forum: kfm-devel

Re: Please comment: ComboBoxWidget change

2008-10-07T07:10:31Z

On Sunday 05 October 2008, Germain Garand wrote:
> unfortunately, QComboBox doesn't seem to offer the necessary hooks to the
> actual popup widget, so we cannot set the position we want.

Which kind of hook do you need? The popup widget is accessible as view().
Knowing when it's going to show up could be done with an event filter for ShowEvent on that view widget.

--
David Faure, faure@..., sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).

From forum: kfm-devel

new webarchiver: please review

2008-10-06T13:34:00Z

or I will commit :-)

I have finally come around to port my webarchiver
(http://www.kde-apps.org/content/show.php?content=32549)
to KDE4 head.

I needed to add two methods to KHTML to get the base URL and charset.
Please take a look at the attached kdelibs patch.

Comments welcome,

Matthias Grimrath

webarchiver-extragear-base.diff.bz2 (27K) Download Attachment
webarchiver-kdelibs.diff.bz2 (744 bytes) Download Attachment

From forum: kfm-devel

Re: KHTML and jQuery

2008-10-06T04:40:43Z

On the bright side, it's an improvement over 3.5:

Tests completed in 72240 milliseconds.
61 tests of 1271 failed.

for 4.1.2 versus:

Tests completed in 163399 milliseconds.
101 tests of 1262 failed.

for 3.5.10

--
Luciano Montanaro

Anyone who is capable of getting themselves made President should on
no account be allowed to do the job. -- Douglas Adams

From forum: kfm-devel

KHTML and jQuery

2008-10-06T03:22:03Z

Hi,

jQuery is a famous JavaScript library. A lot of companies and open source projects use it . It support many browser but Konqueror not one of them. I run the jQuery Test Suite ( http://jquery.com/test/ ) and I get :
Tests completed in 66518 milliseconds.
61 tests of 1271 failed.
I'm using konqueror 4.1.2 . Do you think there are some bugs in KHTML part or it is fault of jQuery ? I want fill some bugs report but I do not know how I can start.

Best regards.
Zayed Al-Saidi

From forum: kfm-devel

Re: Please comment: ComboBoxWidget change

2008-10-05T11:53:36Z

Le dimanche 5 octobre 2008, Harri Porten a écrit :
> On Sun, 5 Oct 2008, Rafael Fernández López wrote:
> > First of all, thanks for your time and investigations. The reason of this
> > setParent() calls is to make the dropdown list be correctly positioned
> > when you click on the combobox and it is into an iframe. We need to
> > reparent it, so we are talking correctly about its coordinates.
>
> But coordinates can be calculated with the appropriate mapTo() and
> mapFrom() calls, too, no?
>

that's what we do with KCompetionBox's popup menu.

unfortunately, QComboBox doesn't seem to offer the necessary hooks to the
actual popup widget, so we cannot set the position we want.

Greetings,
Germain

From forum: kfm-devel

Supporting legacy

2008-10-05T06:56:45Z

Hi!

Some reworking of our tag name representation (r867535) broke support for
legacy HTML elements. The question is do which degree we want to fix or
drop that. The misc/maketags made the following elements equivalent

image -> img
anchor -> a
listing -> pre

for pre HTML 4 compat. I've put up a test on
http://www.froglogic.com/~porten/compat.html that tests both HTML parsing
and createElement() DOM calls for those three. Results:

- neither Firefox, IE nor Opera special case "anchor" and "listing". I
was not able to find any sign of <anchor> btw.

- Opera and IE map "image" onto "img" both in the parse and
createElement()
- Firefox does special-case "image" on createElement() only.

My suggestion would be to follow IE and Opera. Firefox is inconsistent. A
proposed patch is attached but I'm not fully sure it's correct and is
applied at the correct level. Maksim or Vyacheslav may be able to suggest
a simpler if (id == ID_IMAGE) id = ID_IMG; hack at 1-2 central places
only.

Harri.

Index: misc/htmltags.in

=========================
=========================
=================

--- misc/htmltags.in (revision 868005)

+++ misc/htmltags.in (working copy)

@@ -47,6 +47,7 @@

i

iframe

ilayer

+image

img

input

ins

Index: html/htmlprospectivetokenizer.cpp

=========================
=========================
=================

--- html/htmlprospectivetokenizer.cpp (revision 868005)

+++ html/htmlprospectivetokenizer.cpp (working copy)

@@ -698,6 +698,7 @@

const DOMString value = DOMString(m_attributeValue.data(),
m_attributeValue.size()); // ####

switch (tag) {

case ID_SCRIPT:

+ case ID_IMAGE:

case ID_IMG:

if (attribute == ATTR_SRC && m_urlToLoad.isEmpty())

m_urlToLoad = parseURL(value);

@@ -845,7 +846,7 @@

CachedObject* o = 0;

if (tag == ID_SCRIPT)

o = m_document->docLoader()->requestScript( m_urlToLoad,
m_document->view()->part()->encoding() );

- else if (tag == ID_IMG)

+ else if (tag == ID_IMAGE || tag == ID_IMG)

o = m_document->docLoader()->requestImage( m_urlToLoad );

else if (tag == ID_LINK && m_linkIsStyleSheet)

o = m_document->docLoader()->requestStyleSheet( m_urlToLoad,
m_document->view()->part()->encoding() );

Index: html/dtd.cpp

=========================
=========================
=================

--- html/dtd.cpp (revision 868005)

+++ html/dtd.cpp (working copy)

@@ -95,6 +95,7 @@

1, // ID_I

1, // ID_IFRAME

1, // ID_ILAYER

+ 0, // ID_IMAGE

0, // ID_IMG

0, // ID_INPUT

1, // ID_INS

@@ -206,6 +207,7 @@

REQUIRED, // ID_I

REQUIRED, // ID_IFRAME

REQUIRED, // ID_ILAYER

+ FORBIDDEN, // ID_IMAGE

FORBIDDEN, // ID_IMG

FORBIDDEN, // ID_INPUT

REQUIRED, // ID_INS

@@ -304,6 +306,7 @@

// %special

ID_A,

ID_OBJECT,

+ ID_IMAGE,

ID_IMG,

ID_APPLET,

ID_IFRAME,

@@ -496,6 +499,7 @@

case ID_BR:

case ID_AREA:

case ID_LINK:

+ case ID_IMAGE:

case ID_IMG:

case ID_PARAM:

case ID_HR:

Index: html/htmlparser.cpp

=========================
=========================
=================

--- html/htmlparser.cpp (revision 868005)

+++ html/htmlparser.cpp (working copy)

@@ -1089,6 +1089,7 @@

break;

// images

+ case ID_IMAGE:

case ID_IMG:

if (t->attrs&&

KHTMLGlobal::defaultHTMLSettings()->isAdFilterEnabled()&&

Index: xml/dom_docimpl.cpp

=========================
=========================
=================

--- xml/dom_docimpl.cpp (revision 868005)

+++ xml/dom_docimpl.cpp (working copy)

@@ -996,6 +996,7 @@

// images

case ID_IMG:

+ case ID_IMAGE: // legacy name

n = new HTMLImageElementImpl(docPtr());

break;

case ID_CANVAS:

From forum: kfm-devel

Re: [PATCH] Bugs #158932 and #161385 : remounting does not reload the view

2008-10-05T06:01:17Z

Peter Penz schreef:

Hi Laurens,

I've added Kevin to CC (= maintainer of Solid), as I'm quite sure a simpler 
approach must be possible.

Am Thursday, 11. September 2008 11:33:36 schrieb Laurens Vanhove:

Hi all !

I've managed to find a way to fix bugs 158932 and 161385.
To reproduce the bug : mount some volume in dolphin, unmount it (the
view gets emptied), remount it (the view stays empty).

So the problem was that the view didn't get updated when a volume was
(re)mounted at or under the current path. My patch adds a class (used by
DolphinView) that keeps track of solid accessibility changes. When a
volume accessibility changes to true (=> the volume is mounted) and the
mountpoint is a substring of the current path a signal is emitted.
DolphinView gets this signal and reloads.

I suppose the could be a simpler way to get to this result, as the
sidebartreeview automatically updates, which means that the event
probably gets eaten somewhere on the way to the view. But to find out
where that happens would require some serious debugging ... and the fix
might not be trivial either.
Fortunately because the monitor class is simple and gets created by
DolphinView the performance impact should be absolutely minimal :-)


First thanks for taking care of those 2 bugs!

I'd like to prevent adding a new class representing a workaround (even if the 
class is quite straight forward). When unmounting a device the view gets 
already notified that the items should disappear. I did not analyze the code 
yet (I was two weeks on vacation and got 187 Dolphin related bugs.kde.org 
mails ;-)), but I guess that KDirLister gets informed that the directory 
should get empty when the device gets unmounted (-> KDirModel and hence the 
view gets updated). In Dolphin's information panel (infosidebarpage.cpp) I 
manually connected to signals from KDirNotify (see patch 
http://websvn.kde.org/trunk/KDE/kdebase/apps/dolphin/src/infosidebarpage.cpp?r1=818005&r2=818004&pathrev=818005) 
and I guess that maybe KDirLister only misses to emit a corresponding signal 
if the device gets mounted again.

It would be great if you'd have the time to check whether this can be solved 
in a general way in KDirLister, but as KDirLister is a quite complex class I 
fully understand if you see this issue out of your scope ;-) Just let me know 
how you thank we should proceed here.

Thanks!
Peter

I'd be gratefull if anyone would want to review (and if possible commit)
the patch !
And tell me if there is anything wrong ;-)


Regards,
Laurens

Hi Peter !

Sorry for the long delay ... I've been quite busy the last few weeks.
I'm not surprised that you'd rather not use a hack like I made ;-) Even more because it should be unneeded, the sidebar treeview somehow manages to do without ...

I'll see if I can find some time to find out some other solution (should probably suffice to connect some signals, like you said). But if someone else wants to try, go for it !

Regards,
Laurens

From forum: kfm-devel

Re: Please comment: ComboBoxWidget change

2008-10-04T20:31:35Z

On Sun, 5 Oct 2008, Rafael Fernández López wrote:

> First of all, thanks for your time and investigations. The reason of this
> setParent() calls is to make the dropdown list be correctly positioned when
> you click on the combobox and it is into an iframe. We need to reparent it, so
> we are talking correctly about its coordinates.

But coordinates can be calculated with the appropriate mapTo() and
mapFrom() calls, too, no?

> I actually think your patch looks good, but I wonder if an update() wouldn't
> work.

Didn't do the job. QWidget::update() is a noop if isVisible() returns
false.

I noticed the other problems you described, too.

Harri.

From forum: kfm-devel

Re: Please comment: ComboBoxWidget change

2008-10-04T19:17:13Z

Hi Harri !

First of all, thanks for your time and investigations. The reason of this
setParent() calls is to make the dropdown list be correctly positioned when
you click on the combobox and it is into an iframe. We need to reparent it, so
we are talking correctly about its coordinates.

I actually think your patch looks good, but I wonder if an update() wouldn't
work.

I also have to look at the problem of the combobox making the rest of the form
not being hovered after having clicked on it. Better explained: click on the
combobox so the dropdown list is shown, try to hover a button: it won't. Now
click on any place of the page, and now hover the button: it does...

There is also a small issue because of the "setParent" thing. Try to click on
a combobox so the dropdown list is shown, while the dropdown list is shown,
drag the Konqueror window and move it: the dropdown list is not hidden, and it
keeps shown at the same position while you move Konqueror window.

To sum up: this is pretty hackish, and we need to improve some things here =)

Thanks !

Regards,
Rafael Fernández López.

signature.asc (196 bytes) Download Attachment

From forum: kfm-devel

Please comment: ComboBoxWidget change

2008-10-04T18:22:46Z

Hello (Germain and Rafael),

can you comment on the correctness of this proposed change - or rather the
background of why it seems necessary:

I noticed that combo box elements like

<select>
<option>A</option>
<option>B</option>
<option>C</option>
</select>

have some rendering issues. They can be used once just fine but on the
second use the new selection is not updated automatically. The internal
state is fine, just the rendering isn't.

After a long debugging session (really long, as I first had to dive into
KHTML's redirection mechanism and Qt's backing store) I discovered that
the update is not done because the KComboBoxWidget is explicitly hidden!
None of the places were we call QWidget::hide() seems guilty. But the
setParent() calls in ComboBoxWidget::showPopup() have this (documented)
side effect. I saw that Rafael had reinstated those calls upon Germain's
recommendation:

http://websvn.kde.org/trunk/KDE/kdelibs/khtml/rendering/render_form.cpp?r1=862207&r2=862897

I am just curious about their purpose. Since I assume they are needed I
suggested the following patch to undo their side effect:

--- rendering/render_form.cpp (revision 867904)
+++ rendering/render_form.cpp (working copy)
@@ -1352,6 +1352,8 @@
blocked = blockSignals(true);
if (v != parent) {
setParent(parent);
+ // undo side effect of setParent()
+ show();
}
move( p );
blockSignals(blocked);

Harri.

From forum: kfm-devel

Re: Regression on page

2008-10-03T07:38:08Z

> Dear all,
>
> since a few month now, I'm trying out KDE 4.1.1. on Kubuntu hardy.
> There are still many many bugs and it is much less usable than 3.5.9.
> My hope was to get all the advances in konqueror. But to my surprise,
> konqueror does NOT work as good in 4.1.1 and has really many
> regressions and crashes.

Thanks for the feedback.

>
> Take
> http://ipyorkshire.blogspot.com/2008/10/glossary-of-intellectual-property-terms.html
> in 3.5.9 it works like charm
> in 4.1.1 I get a blank canvas

I can't confirm the problem with ~4.1.2; the page shows up fine (though
the JS on the page suggests we should support [[Get]] of attribute name on
NamedNodeMap).

> Also the W3C slidemaker does not work in konqueror 4.1.1 but they
> don't work in 3.5.9 either. So I will ask back and see if the
> javascript stuff in there is responsible. Try
> http://www.w3.org/2008/Talks/0908-Matt-Geo-Widgets/

I can confirm that. The problem seem to be that we return .hash as
%-encoded, which confuses the slide number parser in the tool.
David, does the below look sane?

--- ecma/kjs_window.cpp (revision 865420)
+++ ecma/kjs_window.cpp (working copy)
@@ -2529,7 +2532,7 @@
KUrl url = m_frame->m_part->url();
switch(token) {
case Hash:
- return jsString( UString(url.ref().isNull() ? QString("") : '#' +
url.ref()) );
+ return jsString( UString(url.htmlRef().isNull() ? QString("") : '#'
+ url.htmlRef()) );
case Host: {
UString str = url.host();
if (url.port() > 0)

From forum: kfm-devel

Re: Regression on page

2008-10-03T02:06:40Z

On Friday 03 October 2008, Rigo Wenning wrote:
> I use the full power of konqueror, but in kubuntu, I get the feeling
> that I can't access features (like nepomuk) that are integrated with
> dolphin. But this may be due to the typical kubuntu strip-down of
> menus that I hate so much. I will have to change the kubuntu profile
> to the default profile. I made that for 3.5, but it was a major
> operation (One has to remember which files to change)

Kubuntu is not to blame for this. Blame me rather ;)
Features like nepomuk are only available in dolphin at the moment.
Konqueror uses the dolphin part for filemanagement, however this integration
was done with "works as good as kde3" as a priority, so the new-in-kde4
stuff that needs additional GUI elements, like nepomuk, isn't available
in konqueror right now.

--
David Faure, faure@..., sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).

From forum: kfm-devel

Regression on page

2008-10-02T23:36:15Z

Dear all,

since a few month now, I'm trying out KDE 4.1.1. on Kubuntu hardy.
There are still many many bugs and it is much less usable than 3.5.9.
My hope was to get all the advances in konqueror. But to my surprise,
konqueror does NOT work as good in 4.1.1 and has really many
regressions and crashes.

Take
http://ipyorkshire.blogspot.com/2008/10/glossary-of-intellectual-property-terms.html
in 3.5.9 it works like charm
in 4.1.1 I get a blank canvas

Also the W3C slidemaker does not work in konqueror 4.1.1 but they
don't work in 3.5.9 either. So I will ask back and see if the
javascript stuff in there is responsible. Try
http://www.w3.org/2008/Talks/0908-Matt-Geo-Widgets/

I use the full power of konqueror, but in kubuntu, I get the feeling
that I can't access features (like nepomuk) that are integrated with
dolphin. But this may be due to the typical kubuntu strip-down of
menus that I hate so much. I will have to change the kubuntu profile
to the default profile. I made that for 3.5, but it was a major
operation (One has to remember which files to change)

Best,

Rigo

signature.asc (196 bytes) Download Attachment

From forum: kfm-devel

kio_bookmarks intregration in konqueror

2008-09-29T14:13:12Z

Hello

As kio_bookmarks in now in kdebase trunk, I'm trying to integrate it well in
konqueror. I see four ways :
- add it to the start page choices in konqueror options (done, r866049)
- add it to the bookmarks menu, not easy as it's defined in kdelibs. Can
someone help me do this ?
- getting kcm_bookmarks displayed in konqueror configuration, need your
approval for this.
- add it to the konqueror about: page, maybe replacing the "applications"
menu, need your approval too.

Please consider points 3 and 4 and tell me what I'm allowed to do :)
--
Xavier Vello

From forum: kfm-devel

Re: Fwd: New Issue from qt-bugs Re: [Issue N228549] Qt glyph (greek polytonic) substitution bug (Greek polytonic (ancient greek) font encoding problem)

2008-09-29T11:26:36Z

Hi again

In case anyone is interested... This is the bug report (public task) created
in qt-bugs

--
Regards
Spiros

--
gpg key URL: http://members.hellug.gr/sng/gpg-hellug.html

Hi Spiros

> This is a test case to demonstrate the problem

We're still not sure that this is a bug in Qt. I've nevertheless created
a public task on it:
http://www.trolltech.com/developer/task-tracker/index_html?id=229099&method=entry

Thank you

--

Helder Correia
Support Engineer
Qt Software, Nokia
======================================================================
Qt Developer Days 2008
REGISTRATION NOW OPEN!!!
MUNICH, Germany: October 14th-15th
REDWOOD CITY, California: October 29th-30th
Learn more: www.trolltech.com/qtdevdays2008
======================================================================

signature.asc (204 bytes) Download Attachment

From forum: kfm-devel

programatically import certificate into konqueror

2008-09-29T08:06:15Z

hi

In KDE3, there is dcop that we can use to import user's certificate(p12 format) automatically into konqueror.
But in KDE4, dcop is replaced by dbus and does not work anymore.
I have search the method to do this but so far have no result.
Do anyone know how to do this in KDE4?
Any hints or ideas will be great.

Regards
Yifan

From forum: kfm-devel

Re: "Suspicious code" in khtml

2008-09-29T00:16:32Z

Le dimanche 28 septembre 2008, Harri Porten a écrit :

> Hi!
>
> I've fixed 3 of the "suspicious code" issues reported by Christoph
> Bartoscheck on kde-core-devel last week. Here are the remaining ones:
>
> Subject: Suspicious code in kdelibs/kdebase revision 864329
>
> - kdelibs/khtml/xml/xml_tokenizer.cpp:505
>
> Line 502 indicates that body can be NULL here. A crash follows.
>
> - kdelibs/khtml/xml/dom_xmlimpl.cpp:483
>
> Line 481 indicates that m_sheet can be NULL here. A crash follows.
>
> - kdelibs/khtml/xml/dom_selection.cpp:827
>
> Line 824 indicates that renderNode can be NULL here. A crash follows.
>
> - kdelibs/khtml/html/html_baseimpl.cpp:735
> kdelibs/khtml/html/html_imageimpl.cpp:160
> kdelibs/khtml/rendering/bidi.cpp:1103,1154
>
> A break might be missing. A fall through comment would be nice.

I fixed all those except the bidi algorithm. It's not clear to me whether the
missing breaks are intended.

Germain

From forum: kfm-devel

Re: "Suspicious code" in khtml

2008-09-29T00:15:20Z

On Mon, 29 Sep 2008, Germain Garand wrote:

> I fixed all those except the bidi algorithm. It's not clear to me whether the
> missing breaks are intended.

This is the one I absolutely did not understand either.

Thanks,

Harri.

From forum: kfm-devel

"Suspicious code" in khtml

2008-09-28T14:51:41Z

Hi!

I've fixed 3 of the "suspicious code" issues reported by Christoph
Bartoscheck on kde-core-devel last week. Here are the remaining ones:

Subject: Suspicious code in kdelibs/kdebase revision 864329

- kdelibs/khtml/xml/xml_tokenizer.cpp:505

Line 502 indicates that body can be NULL here. A crash follows.

- kdelibs/khtml/xml/dom_xmlimpl.cpp:483

Line 481 indicates that m_sheet can be NULL here. A crash follows.

- kdelibs/khtml/xml/dom_selection.cpp:827

Line 824 indicates that renderNode can be NULL here. A crash follows.

- kdelibs/khtml/html/html_baseimpl.cpp:735
kdelibs/khtml/html/html_imageimpl.cpp:160
kdelibs/khtml/rendering/bidi.cpp:1103,1154

A break might be missing. A fall through comment would be nice.

Harri.

From forum: kfm-devel

Re: konqueror crash

2008-09-28T02:35:12Z

El Viernes 26 Septiembre 2008, Nick Shaforostoff escribió:

> sometimes i get the following crash starting konqueror from trunk.

> Thread 1 (Thread 0xb5ff3720 (LWP 5344)):

> [KCrash Handler]

> #5 0xb6a3f277 in QAction::isChecked () from /usr/lib/libQtGui.so.4

> #6 0xb7e77ea1 in KonqViewManager::saveViewProfileToGroup () from