287 Commits

Author	SHA1	Message	Date
rbs%maths.uq.edu.au	74698af0b2	since bad end of lines cause troubles on some platforms, extend the little perl magic for dos2unix to dos2{unix or mac} so that the conditional addition of mathml.css to the ua.css works reliably everywhere	2002-02-08 03:34:57 +00:00
rbs%maths.uq.edu.au	bdad95f005	Updated comments	2002-02-08 00:58:15 +00:00
rbs%maths.uq.edu.au	8ad7c6897f	Opted to move a functionality to the specialized class where it belongs	2002-02-08 00:57:39 +00:00
rbs%maths.uq.edu.au	9cd3cfff6d	By definition <mi> renders in italic type when its textual content consists of a single char, and in normal type when it consists of mutiple chars. This is a fix to re-resolve the style data properly when the text content (and its length) get changed back and forth dynamically	2002-02-08 00:53:19 +00:00
rbs%maths.uq.edu.au	ba931a911b	s/gInvariantChar/gInvariantCharArray/g and avoid an unnecessary string copy in LookupInvariantChar()	2002-02-08 00:42:32 +00:00
rbs%maths.uq.edu.au	6c7c1056cf	[MathML only - candidate version for 0.9x] * Hooked robust support for advanced dynamic updates via JavaScript and the DOM across the entire MathML module (with particular attention to the numerous presentational attributes that MathML has). * Eliminated some functions (or folded their functionality into recently added APIs). * Reviewed the module as a whole and simplified several chunks of code owing to the benefit of hindsight now. * Added/reviewed the documentation in anticipation of the freezing of Mozilla 1.0, to improve the understanding of people that may read the archived code and learn from it under the appropriate context of this time. * Reviewed the XXX comments and addressed those that could be addressable in this timeframe. * Some renamings (e.g., ReLayout -> ReLayoutChildren to avoid ambiguities as to who does what in terms of the terminology used when referring to the frame hierarchy).	2002-02-07 04:38:08 +00:00
rbs%maths.uq.edu.au	9063863332	Add a flag to distinguish a real infix operator from its 'isolated' form, e.g., as in x^{+}	2002-02-07 02:12:46 +00:00
rbs%maths.uq.edu.au	19be27a26f	Add a helper function to lookup all the forms under which an operator is listed in the Operator Dictionary	2002-02-06 20:48:35 +00:00
rbs%maths.uq.edu.au	bc20957975	Using the recently added APIs to respond to dynamic changes: a _much_ simpler (re)fix of bug 114909 - dynamically changing displaystyle attribute in <mstyle> tag doesn't change display	2002-02-03 21:15:29 +00:00
rbs%maths.uq.edu.au	ee9e501c96	Add ReLayout() - a driver that hides the details needed to re-sync a frame subtree in case of dynamic changes: rebuild of automatic data, re-resolve script styles, fire a reflow	2002-02-03 21:06:51 +00:00
rbs%maths.uq.edu.au	f5893aef35	slight simplification - the removal will recurse down to the wrapped child	2002-02-03 20:11:37 +00:00
rbs%maths.uq.edu.au	0a35caab09	Continuing to factor handlers to rebuild automatic data during dynamic changes	2002-02-02 05:43:03 +00:00
heikki%netscape.com	3a6fb922c1	Bug 122710, dtd folder in dist/bin should be moved into res. r=rbs@maths.uq.edu.au&harishd, sr=sfraser.	2002-02-01 22:28:55 +00:00
rbs%maths.uq.edu.au	c33e0f4dbe	Added two APIs, InheritAutomaticData() & TransmitAutomaticData(), to respond to dynamic changes that affect automatic data (i.e., data inferred from the actual definition of each MathML tag). The computation of the data was so far mostly locked in Init() and SetInitialChildList() which are meant to be called only once in the life-time of a frame. Factored this computation into the two APIs to be used to rebuild the data, and which can be called repeatedly as appropriate, when dynamic changes arise in the content model. Shuffled code to stay in sync and consolidated the hooks used dynamic changes with the APIs	2002-02-01 15:10:50 +00:00
rbs%maths.uq.edu.au	580f85ab56	activate the rhelper to wrap foreign frames	2002-02-01 14:31:21 +00:00
rbs%maths.uq.edu.au	f91be4f749	Helper class use for wrapping non-MathML elements (e.g., html:img) so that they can mix better with other surrounding MathML markups. Factored previous code here and added hooks that were needed for child list operations	2002-02-01 14:23:07 +00:00
mkaply%us.ibm.com	4cd0454803	Add PR_CALLBACK to get MathML building on OS/2	2002-01-28 23:24:17 +00:00
rbs%maths.uq.edu.au	8a12dec1b6	revert an unintended change that crept in	2002-01-27 22:23:07 +00:00
rbs%maths.uq.edu.au	e803323d31	Another renaming for ease of readability: s/core/coreFrame/g and s/next/nextFrame/g	2002-01-27 22:10:45 +00:00
rbs%maths.uq.edu.au	7499dddea5	When looking up the real sibling of a cell, use colspan/rowspan (instead of just 1) as the increment to cater for cases of spanning cells in the cellmap	2002-01-27 19:26:42 +00:00
rbs%maths.uq.edu.au	8c3fe11744	typos	2002-01-25 21:57:47 +00:00
rbs%maths.uq.edu.au	16fb7a5ab0	Remove leftover from debugging	2002-01-25 07:34:49 +00:00
rbs%maths.uq.edu.au	a1a725e6af	Improvements to <mtable>	2002-01-25 07:03:43 +00:00
rbs%maths.uq.edu.au	c0c66c4dbc	In case of markup errors, provide a more discoverable feedback to users. Drop in the other functions from the nsIFrame APIs that steve.swanson@mackichan.com needs to consolidate his Dynamic JavaScripted MathML editor	2002-01-25 06:48:30 +00:00
rbs%maths.uq.edu.au	808a02672c	Fix MathML bug 121037. Background-color is not showing up in stretchy chars that use additional leaf style contexts. More reliable recovery from the background revamps in bug 116161	2002-01-25 06:08:03 +00:00
rbs%maths.uq.edu.au	2d36f0f82f	Updated comments to add a reference to bug 69325 which details the 8 possible cases that may arise depending on whether displaystyle, largeop, stretchy are true or false	2002-01-25 06:02:52 +00:00
rbs%maths.uq.edu.au	7f70f2cd28	Fix regression from the fix for bug 116161. The picky treatment of stretchy MathML characters involves additional style contexts whereas the default search of the background stops at the primary style context	2002-01-15 07:48:10 +00:00
rbs%maths.uq.edu.au	cef48900e6	Add two default declarations to support display=block|inline introduced in MathML 2.0	2002-01-14 08:59:01 +00:00
rbs%maths.uq.edu.au	98190fd75f	Additional reduction of code size by using a helper to propagate the update of presentation data in a range of child frames	2002-01-14 00:55:53 +00:00
rbs%maths.uq.edu.au	cf9d24490b	Continuation of tyding	2002-01-12 08:06:08 +00:00
rbs%maths.uq.edu.au	a83e90c318	Compact code - use a helper function for a re-usable chunk of code	2002-01-12 03:04:14 +00:00
rbs%maths.uq.edu.au	05c1973f14	Some tweaks	2002-01-11 05:08:11 +00:00
rbs%maths.uq.edu.au	14edaf7b93	Alert! MissingMathML fonts. Here goes the prompt that the user will see when it is detected that crucial stretchy MathML fonts are not installed. The list of those fonts that are missing is shown so that the user can install them. The message is hardcoded and would need to be localized	2002-01-11 02:57:03 +00:00
rbs%maths.uq.edu.au	0185ec37f1	[MathML only] Rename the pointer on the embellished operator from |firstChild| to |next| because it is more clearer with the latter that it is an hierarchy. Also made the inheritance of the displaystyle more reliable by walking up to the root <math> tag if necessary, e.g., if a MathML frame happens to be surrounded by non-MathML helper frames needed for its support	2002-01-10 23:40:40 +00:00
rbs%maths.uq.edu.au	9e0fa92a99	Finishing touch to MathML bug 114909	2002-01-10 21:30:41 +00:00
rbs%maths.uq.edu.au	07dff3093b	Fix MathML bug 114909 - dynamically changing the displaystyle attribute in <mstyle> tag doesn't change display	2002-01-10 04:42:55 +00:00
rbs%maths.uq.edu.au	06c0b341cd	[MathML only] shuffle shared functions between files to balance the weights of the files	2002-01-09 18:51:30 +00:00
rbs%maths.uq.edu.au	74dabf2bd2	MathML bug 118475 - Allow stretchy MathML fonts to be customized via the pref system	2002-01-08 02:15:23 +00:00
dbaron%fas.harvard.edu	c9c0a97163	Replace the BodyFixupRule with code in nsCSSRendering to fix all the problems with dynamic style changes and backgrounds on the root element and the BODY element. Change the way the HTMLDocumentColorRule (for the table color quirk) is initialized since it used to depend on the BodyFixupRule. b=116161 (also fixes other bugs) r=attinasi sr=hyatt	2002-01-07 23:46:07 +00:00
rbs%maths.uq.edu.au	8552445f56	Move a getter to the safer place where the font was just set in the rendering context (its state could change, who knows)	2002-01-06 19:50:51 +00:00
rbs%maths.uq.edu.au	9b899cf2fb	Add an additional clause when picking the desired stretchy size in the self-stretch situation, and give more descriptive comments	2002-01-06 19:47:24 +00:00
rbs%maths.uq.edu.au	aceedc85d5	Simple cleanup	2002-01-06 19:42:08 +00:00
rbs%maths.uq.edu.au	c2330df4a5	Fix leaky style context	2002-01-06 19:40:04 +00:00
rbs%maths.uq.edu.au	5475678622	MathML only - Further tweaks to getting the preferred stretchy size. Decide that the core <mo> is the provider of the 'em' unit for the lspace & rspace of embellished containers. Add support of the 'bevelled' attribute in <mfrac> by leveraging on the code of <mfenced>	2002-01-05 01:15:04 +00:00
rbs%maths.uq.edu.au	7d8032905d	MathML only - remove outdated APIs	2002-01-05 01:08:09 +00:00
rbs%maths.uq.edu.au	fe33b4e8ce	MathML only - Reset the stretchy flags also when presentation data get changed	2002-01-05 01:06:05 +00:00
rbs%maths.uq.edu.au	0f015abd9f	Reset the stretchy flags of underscript/overscript frames when their geometry is expected to be like subscript/superscirpt frames	2002-01-03 23:18:01 +00:00
rbs%maths.uq.edu.au	a81d5f1b96	Follow-up to MathML bug 117652 - additional tweak to the preferred stretchy size for the case where the direction of stretching varies during the descend	2002-01-02 22:42:52 +00:00
rbs%maths.uq.edu.au	a938c08678	MathML bug 117652 - Tune the preferred stretchy size of stretchy containers by controlling the inclusion/exclusion of embellishments in the calculations, and defer the stretching of siblings until after finishing treating the embellished child	2002-01-02 05:32:33 +00:00
rjesup%wgate.com	2e680f1b17	Bug 96108: Inline ElementAt(), add SafeElementAt(), voidarray usage fixes, ... assertions for bad voidarray indexes. r=jst, sr=waterson	2001-12-21 01:10:07 +00:00