I need to integrate a WYSIWYG Rich Text Editor into an existing D3 system
whereby I can pass some text to the editor, manipulate the text using the
WYSIWYG editor and then pass the resultant string back into D3.

I've had a look around and come across this utility
http://www.kevinroth.com/rte/demo.htm which seems to be able to accept a
string, manipulate it and then output the string for the modified text. (To
me, it seems to do most of the "hard" stuff?)

However, I now need to integrate this into D3.

In case you haven't guessed already, this isn't my forte so is there anybody
here that can help with this please?

I'm willing to pay for any such help, so an idea of how much said help is
going to cost would be much appreciated.

Once this is done, I intend to offer to finished code back here so that, if
somebody has a similar request in the future, the code is all done for them.

Thanks

With Kindest Regards

Mike Wooding

One is tempted to ask what you are going to do with the 'rich text'
once you are finished ..? You also need to know/realize that the
'output' appears to be HTML (as I would expect from a 'plain' browser),
and there also seems to be a problem with formatting (try highlighting
some text & changing style.

Within Visage we generate "real" Rich Text via an ActiveX control - if
this is what you need to do (RTF vs HTML), then you will probably need
to follow a similar path I believe. That being the case, I'm assuming
you are using some kind of terminal emulator, so the question then
becomes "which one" (curiosity - doesn't impact the implementation,
just the mecahnics of data transfer).

I imagine someone like Kevin Powick, TonyG & others will be able to
help you (probably even WITHOUT mv.NET :-) - if you get stuck, we
certainly can, though depending on the solution you finally elect, you
may not be able to post code back here

Ross Ferris
Stamina Software Pty Ltd
Visage > Better by Design

"Ross Ferris" <rossf (AT) stamina (DOT) com.au> wrote


It will be uploaded to a web-site.


I need HTML. I put it in the subject of the original post, but not the
body. Mea culpa.


Changing the "Style" affects to whole paragraph. The only way to mix styles
within a paragraph is to manually edit the font, size, etc., parameters. I
am aware of this limitation and in this instance, it doesn't cause a
problem.


I need HTML. A typical input from D3 into the Editor (or vice-versa) might
look like this:

<p>This is bold.&nbsp; This is italics.&nbsp; This is underlined.</p>
<p>Left Justified.</p>
<p align="center">Centred</p>
<p align="right">Right Justified</p>
<p align="justify">Fully Justified.</p>
<p></p>
<hr />
<p></p>
<p>(Horizontal Rule above)</p>
<ol>
<li>Numbered List 1</li>
<li>Numbered List 2</li>
<li>Numbered List 3</li></ol>
<ul>
<li>Bullet List 1</li>
<li>Bullet List 2</li>
<li>Bullet List 3</li></ul><blockquote dir="ltr" style="MARGIN-RIGHT: 0px">
<p>Indent</p><blockquote dir="ltr" style="MARGIN-RIGHT: 0px">
<p>Indent More</p><blockquote dir="ltr" style="MARGIN-RIGHT: 0px">
<p>Indent Even More</p></blockquote></blockquote></blockquote>
<p>Not indented.</p>
<p>This is black, <font color="#ff0000">red</font> and<font color="#6633ff">
blue</font>.&nbsp; (Apologies if that's wrong.&nbsp; Colour blindness can be
a pain sometimes.)</p>
<p>Likewise with the backgrounds, black, <font style="BACKGROUND-COLOR:
#ff0000">red</font> and <font style="BACKGROUND-COLOR:
#3366ff">blue</font>.</p>
<p>I think you get the idea?</p>

The editor will render it WYSIWYG. Users can edit it and then pass the
changed details back to D3 which will upload them to a web-site.


Accuterm.


Never underestimate the power of curiosity. :-)


I'm officially "stuck" now. Transferring the data in and out of D3 /
Windows via Accuterm isn't a problem. Manipulating the text using the
editor isn't a problem. The two bits I'm missing are getting the code from
D3/Accuterm into the editor, and then getting the modified code out of the
editor back into D3/Accuterm.

The editor comes with full source and I was hoping somebody could tweak that
source for me to achieve the necessary. I don't think it is too hard, but
then I hate it when clients tell me how easy my job should be, so I'll shut
up now.

Thanks for the help Ross.

Regards

Mike Wooding

Here are some ideas on how you could implement it. I haven't looked at the
actual product you mentioned, but this should work. First, transfer data
using Accuterm. Accuterm writes the data out to a file on the local
machines hard drive. Invoke the editor with the file name to edit like
"edit xyz.html". The main trick is recognizing when the person is finished
editing the file and then using accuterm to send the data back to D3. I
would have a continue prompt in the accuterm window which would signal that
they finished editing the file and then transfer the data back.

Hope this will help you get the project finished.

Matt Harting

Mike Wooding <mikewooding (AT) email (DOT) com> wrote

The "trick" is muich more than this! Mike's "editor" is as actually
jScript in a web page manipulating the DOM, so tghe problem isn't just
"how do I get the text out", but also populating the puppy in the first
place.

If you are using FC, life is a little easier - you could just have a
SUBMIT (changes) on the page, and have a process pick things up - but I
have a feeling the required solution needs to be cheaper than this? For
a few hundred bucks he could also get a copy of Visage (heck, I'll even
drop a copy of the editor on to a blank form for you), that could be
iniated by Accuterm , but I have a feeling this would also be
unacceptable.

So, in terms of initially loadig the document, assuming all clients
have access to a common network drive [letter], using OSFI (and samba
if on NIX) you can simply write out a blank page INCLUDING any initial
text, then access this by getting accuterm to do a (client DOS) execute
by using something like :

PRINT ESC :STX :">iexplore x:\"field(oconv("u50bb"," ",1):".htm": CR


assuming X: is the common mapped drive letter - thisis the form that
waits until the execute program finishes

OK, so now have your "editor" displayed onscreen ... now go &
google/find the jScript.chm from Microsoft & ook up the "File System
Object" .... you see, IE allows your browser to write information to
the file system - in this example, the "X" drive.

Write the information out to a know name (you could substitute this
into the page when you inially write it, and now your D3 program can
read it .... oh, and if there is no file with the expeted name, it
means that must have closed the browser down

Hopefully this will give you enough to complete the project under your
own steam

As Matt suggests, it is very easy to push a pick document to your PC,
edit it, and then pull it back into pick. Here is a some code that
allows .rtf documents to be stored on a pick host and
created/edited/printed/viewed using wordpad.exe.

/Scott Ballinger
Pareto Corporation
Edmonds WA USA
206 713 6006

-----------begin sub snippet-----------

7000 * edit the document using the application associated with .rtf
files

IF ACCUTERM ELSE
MSG = "Cannot edit document unless you are using a PC running
Accuterm."
GOSUB 9000
RETURN
END

READ DOCUMENT.OLD FROM DOCUMENTS,DOC.ID THEN GO 7100

DOCUMENT.OLD = ""

* this is a new document, so load the default stuff...

READ INIT FROM DOCUMENTS,"INIT.rtf" ELSE INIT = ""

MAX = DCOUNT(INIT,@AM)
INIT = DELETE(INIT,MAX) ;* remove end of document marker ("\par }}")

* fill in letterhead name, addr, phones

READ COMPANY.REC FROM COMPANY.FILE,COMPANY.ID ELSE COMPANY.REC = ""
FNAME = COMPANYREC<COMPANY$NAME>
FADDR = COMPANYREC<COMPANY$ADDR>
PH1 = COMPANYREC<COMPANY$PHONE> ;* main
PH2 = COMPANYREC<COMPANY$PH800> ;* toll free
PH3 = COMPANYREC<COMPANY$FAX> ;* fax

IF FNAME EQ "" THEN FNAME = "Pareto Corporation"
IF FADDR EQ "" THEN FADDR = "11201 Whitcomb Pl, Edmonds WA 98020"
IF PH1 EQ "" THEN PH1 = "206-713-6006"
IF PH2 EQ "" THEN PH2 = "425-670-0831"
IF PH3 EQ "" THEN PH3 = "425-672-4630"

INIT = CHANGE(INIT,"@COMPNAME",FNAME)
INIT = CHANGE(INIT,"@COMPADDR",FADDR)
INIT = CHANGE(INIT,"@PH1",PH1)
INIT = CHANGE(INIT,"@PH2",PH2)
INIT = CHANGE(INIT,"@PH3",PH3)

P = @AM:"\par " ;* rtf paragraph marker

DATE = OCONV(DATE(),"D4-")
MO = DATE[1,2]
DA = DATE[4,2] + 0
YR = DATE[7,4]
CALL PC.GET.MONTH(MO,MONTH) ;* return pretty month, e.g. "January"
IF MONTH NE "" THEN DATE = MONTH:" "A:", ":YR

****
**** here is where you get the addressee's name, address, and
city-st-zip, then...
****

ADR = P:ADDRESSEE:X:P
MAX = DCOUNT(ADDR,@VM)
FOR N = 1 TO MAX
IF ADDR<1,N> NE "" THEN ADR := ADDR<1,N>:P
NEXT N
ADR := CSZ:P

INIT := P:P:P:PATE:P:P:P:ADR

* get formletter boilerplate if indicated

IF FORM.ID NE "" THEN
READ FREC FROM FORMLETTERS,FORM.ID ELSE FREC = ""
FREC = CHANGE(FREC,@AM:@AM,@VM) ;* mark paragraphs with vm
FREC = CHANGE(FREC,@AM," ") ;* make remove am at end of every
line
FREC = CHANGE(FREC,@VM,P) ;* convert vms to rtf paragraph
marks
INIT := P:FREC
END

INIT := P:P:P:P:P:AUTHORS.NAME:P:AUTHORS.TITLE:P:P

* allow for CC: info at bottom of letter, including cc: address if
provided

IF CCADR NE "" OR CCNAME NE "" THEN
CCADR1 = CCADR<1,1>
CCADR2 = CCADR<1,2>
INIT := P:P:"CC: ":CCNAME:P
IF CCADR1 NE "" THEN INIT := SPACE(7):CCADR1:P
IF CCADR2 NE "" THEN INIT := SPACE(7):CCADR2:P
INIT := SPACE(7):CCCSZ:P
END

INIT := P:P:POC.ID:P:"}}" ;* end of document marker

WRITE INIT ON DOCUMENTS,DOC.ID

7100 * download doc to pc, edit it, then upload it back here

DATA "O","1=T","","S","K","DOCUMENTS",DOC.ID,"C:\","DOC .rtf"
EXECUTE "FT"

* at2k2 doesn't need the 'start'
* accuterm variable is 1 if old accuterm, 2 if accuterm 2k2

IF ACCUTERM EQ 1 THEN X = "start" else X = ""
PRINT ESC:STX:">":X:" c:\DOC.rtf":CR:
IF ACCUTERM EQ 2 THEN
PRINT ESC:STX:"I":CR:
END ELSE
PRINT
PRINT "Press RETURN to continue...":
END
INPUT XXX

DATA "O","1=T","","R","K","C:\","DOC.rtf","DOCUMENTS",D OC.ID
EXECUTE "FT"

READ DOCUMENT.NEW FROM DOCUMENTS,DOC.ID ELSE
DOCUMENT = ""
RETURN
END

* since we are using wordpad to view the document, they might have
changed it.
* test for "sent" indicator and disallow changes to document.

IF DOCUMENT.OLD NE "" THEN
IF DOCUMENT.OLD NE DOCUMENT THEN
IF SENT NE "" THEN
MSG = "Sorry, can't change a document after its been sent."
GOSUB 9000
WRITE DOCUMENT.OLD ON DOCUMENTS,DOC.ID
RETURN
END
CHANGED = DATE()
END
END

RETURN

"Mike Wooding" wrote:
Hi Mike, good to see you again. Not to take anything from this
developer, but if you're going to use a web-based text editor, I'd
suggest you use one that's extremely popular, supported on all popular
browsers, and with a variety of other technologies:
http://www.fckeditor.net/
It was Project of the Month on SourceForge last month - quite an honor
among tens (hundreds) of thousands of projects.

I haven't tested it, but I believe Ross provided the code which will
cause AccuTerm to launch the browser to edit text using whatever
editor you have in there. Seems to me like you have everything you
need to accomplish this.

Other options with AccuTerm include ObjectBridge and DOSSVC - this
last one is a callable sub that does the write/execute/read pattern
that many of us use - I just found this recently but haven't tried it
yet.

HTH,
Tony
TG@ removethisNebula-RnD.com

"Ross Ferris" <rossf (AT) stamina (DOT) com.au> wrote

Bringing all this up-to-date, as recommended by Tony, I have now switched to
using "FCKEditor", so populating the puppy in the first place is no longer
an issue. (See post elsewhere to see the HTML code I'm creating from within
D3 and how it gets populated.)


Sadly, the site in question isn't. Hopefully, in the not too distant
future, it will have mv.NET and, although that is for a related project, it
would appear from the replies in this thread, that it won't be needed for
this.


If it needs FlashConnect, then it needs it. However, I don't think it does,
so I'd prefer the client didn't have to spend the money if it isn't needed.

As it will only make like "a little easier", it seems hard to justify?


American, Canadian, or Austrialian? :-)


At the moment, I need a solution. If that solution is going to cost a few
hundred bucks, so be it.

Whether that few hundred bucks needs to go on a product like Visage, or to
somebody more knowledgeable than me in these matters for them to write me a
solution, doesn't really matter. I need a solution for the client.

If the solution is going to cost significantly more than a few hundred
bucks, the client may decide they no longer *need* the solution.



Using FCKEditor (as recommended by Tony elsewhere in this thread) I have
this bit working fine.

Thanks for the help with that.


From searching, it seems Microsoft have renamed "jScript" to "Windows
Script" which seems to be at version 5.6 at present? Hence a search for
jScript.chm didn't reveal much, whereas a search for Windows Script 5.6 chm
was much more enlightening. (Does that seem right, or am I way off base,
again!)

Having said that, it wasn't as enlightening as I think you thought it might
be, IYSWIM?


Sadly, your hopes for me are misplaced. :-(

I know, (OK, strongly suspect), that I should be using the FileSystemObject
to cause the amended data to be written back to the filesystem rather than
posted somewhere using an asp call.

But I don't know how? Any further pointers you can offer would be
gratefully received.

Thanks for the assistance to date.

With Kindest Regards

Mike Wooding

"Tony Gravagno" <g6q3x9lu53001 (AT) sneakemail (DOT) com.invalid> wrote

Good to be seen again. :-)

Thanks. As noted elsewhere, this is what I'm now using.


Again, like Ross, I'm sory to report that your confidence in me is
misplaced. :-(

I'm sure I do have everything I need to accomplish this. But I don't know
what to do with all the bits. :-((


Likewise on DOSSVC. (I found it whilst I was searching for documentation of
FTS which I was going to mention elsewhere in this thread.)

I see a lot of basic code which stacks all the inputs and then performs an
EXECUTE "FT". I've yet to see a routine that does this which includes error
checking (user pressed "Abort", file exists and can't be overwritten, no
such Windows path, etc.)

I find the FTS subroutine superior to FT for a number of reasons:

(1) instead of stacking the data prior to an execute, the options are passed
as parameters.
(2) it includes error checking as standard.
(3) I can pass whatever I want transferring as a variable, rather than
giving it a Pick file and item name (and let's face it - 99% of the time, I
either have the data I want sending to Windows in a variable within my prog
already (in which case I don't need to write it to Pick - I can just send it
using FTS) or I can read it into a variable, which allows me to handle duff
file / item names before they get anywhere near AccuTerm.

Anyway, I fear I'm preaching to the choir here. So, on that note, I'll
finish as it is late and I have a plane to catch in three hours, so I'm
calling it a night.

With Kindest Regards

Mike Wooding

"Mike Wooding" wrote:
That is a very important observation. People here very frequently
look for solutions to problems. When I suggest some sort of for-fee
component, they counter with the DIY freeware option. The reason I
often lead with for-fee tools is that I know those tools are complete
solutions which include error handling and other things that are
necessary for real world applications. Most often when people
implement the DIY freeware, they will not add any kind of error
handling. The put in the code, it does the job, they ship it to the
end-user. If/when an error does occur, the end-user is faced with
some weird message that no one has ever seen - it's not coming from
the BASIC code, the VAR doesn't recogize it, it's probably coming from
the freeware but what does it really mean? In my mind, such quick
fixes are not true "solutions". The tools I recommend and those I
write are complete with error handling - they tell you exactly what
went wrong, they don't blow up the application or mess up the screen,
and they don't require research on the part of the VAR. The process
of adding error handling and masking cryptic responses takes a lot of
time. I put in this time to save the VAR from having to do so
themselves, and that's why I charge for my products and services. The
question for each individual is whether the cost of for-fee components
and services is truly more than the cost of dealing with support
calls, end-user frustration, research time, and doing their own coding
to re-invent what's already been done by others.

T