This is an editing object for HTML text.
It is based on the standard ATK text object.
It is alpha release, and it is available
from ftp.cs.city.ac.uk:/htmltext/
To use:
1) Compile it up:
genmake
make Install
2) Modify your .ezinit or your .atkinit or your global.ezinit, etc....
to have the following:
include $(ANDREWDIR)/lib/html.init
where ANDREWDIR is the location of your installed andrew tree.
3) edit a file with the .html extension!
Send comments to njw@cs.city.ac.uk.
Notes on this version:
* It has the point size of the fonts explicitly stated in the template file.
This is an attempt to make the fontsize similar to the size used in
NCSA Mosaic. But not quite. HoHum.
* Many features would be nice (for example, following a link, loading in
a file via http, seeing the document title in a nice label widget, etc) but
aren't there. This is because this is purely an HTML text widget. A real
HTML editor would contain an htmltext object and be wrapped up with other
things to do the features. e.g. http retrieval should be placed into
an http version of the frame object. I have attempted to use ADEW to
develop a more complete editor, but I can't quite grok ADEW yet. Perhaps
someone else could do this...?
* Known drawbacks/bugs:
- the title of a document isn't visible enough, and the variables associated
with the are not editable in a nice or consistent way compared with
the rest of the system. I'll be fixing this when I work out some nicer
way of placing the htmltext into a real editor frame (using ADEW).
- the menus aren't particularly user-friendly. If some comments on HCIness
of the system come back to me, I'll try fixing this up.
- what is written to file is not what is seen on screen. Especially
whitespace and newline type stuff.
- cutting and pasting is horrible. There are several solutions better than
what happens now (anything is better than a core dump :-!), but I haven't
had the time to look at this. Currently, the cutbuffer is assumed to
contain HTML. Badness can result from this.
- description lists are not enabled. Again, I haven't had time to implement
a "nice" way of doing these.
- the code is particularly gross. I would recommend anyone who wants to
better this object to understand this one first, and then start again.
especially with the WriteSubString() routine, which I made the mistake
of basing on the plain text object. Because they write such completely
different streams, taking the original code and modifying it has caused
a lot of grief. When time allows, this particular rewrite is first on
my list of things to do.