issue-#113 XhtmlExporter bug with StdLinks.Target

[Josef Templ]

meta tag added.

[Ivan Denisov]

Works well. I checked with new Firefox and old IE6.

In html text I found that there are many useless tags generated for Unicode words.

Code: Select all

<p><font size="2">3 </font><font size="2">Что</font><font size="2"> </font><font size="2">такое</font><font size="2"> </font><font size="2">БлэкБокс</font><font size="2">?</font></p>

Josef, is there some easy way to fix this? What do you think?

[Ivan Denisov]

I am suggesting to unite XhtmlEntitySets single procedure MapCharToEntity with XhtmlExporter module and eliminate this file Xhtml/Mod/EntitySets.odc from repository.

[Josef Templ]

I am suggesting to unite XhtmlEntitySets single procedure MapCharToEntity with XhtmlExporter module and eliminate this file Xhtml/Mod/EntitySets.odc from repository.

Actually I had already removed this module in my working copy but then I thought it would be better to keep it.
There are several reasons for that:
1. It has a very useful format conversion procedure and if anybody has to deal with HTML it would be possible
to import only this module without the complete XhtmlConverter.
2. I didn't want to change anything in the compile and link list at this late stage of the release.
3. If it is really necessary to introduce URL encoding it would give sense to add it to this module.

The output of the XhtmlExporter is ugly indeed.
I am afraid that I don't have the time now to look into the details.
I did some quick experiments with changing the paragraph handling
such that links are not put into their own paragraph but it turned out
to get more complicated than I expected because it violates some
fundamental invariants that I don't understand. I am sure it can be
done (also the redundant <font> elements) but it needs some time.

- Josef

[Zinn]

About XhtmlEntitySets:
- The comment line "- 20050811, mf (reported by ... " should be deleted, because this modification was deleted with the last change.
- In my opinion we should use Utf-8 instead of &#.... , because the added meta tags solves the IE problem and the the IE is obsolete.

About Xhtml:
- <font> tags allowed in HTML only. They don't allow in XHTML. They must be deleted or set into comment until we find another solution.
- Currently I have no idea how to improve the design.

- Helmut

[Josef Templ]

The change history must be considered immutable.

Utf-8 is already used in the latest version.
My guess is that the browsers all ignore the encoding of the xml header (when reading from a file)
but Chrome uses Utf-8 by default whereas IE uses Latin-1 or something else.
Just a wild guess.

Regarding the XHTML standard:
The more I read the more problems show up.
For example, even &apos; is not very well defined.
I found a place in https://www.w3.org/TR/xhtml1/#issues
where they suggest the code point format ' instead.
Also as Helmut pointed out the <font> element seems to have disappeared
between the Transitional and the Strict version.
This would mean a significant redesign of the exporter but then it
would probably produce better output.

- Josef

[Josef Templ]

There is a context menu for selecting the encoding in IE(9) indeed.
It even includes an 'automatic selection' of the encoding.

XhtmlExporter docu also needs to be updated.

- Josef

[Ivan Denisov]

I made some experimants with CSS. It is not hard to replace font tag to span with css style instructions. I will do this soon.

Also I found, that Helmut made one subsystem
http://www.zinnamturm.eu/downloadsDH.htm#Html

[Zinn]

The subsystem Html is overage and need a redesign. I shifted the redesing since 4 years because I worked on BlackBox update instead.
Before the redesign start we need some convention about the text formating and we need associated sample CSS file.
Sorry I didn't work out this topic yet.
- Helmut

[Josef Templ]

Is it possible to include the CSS part in the output html file itself (I think so)?
A separate CSS file would be very inconvenient.

With a CSS section the overall approach would probably consist of 2 phases:
1. scan the text and collect all different styles
2. scan the text and output the html with the related style.

The styles can consist of paragraph style attributes such as margin, line spacing etc.
and character style attributes such as font, size, etc.
There will be many detail problems e.g. with TABs.

- Josef