Topic: Scribe v3.0 (1 of 4), Read 234 times
Conf: VEDIT User Applications
From: Scott Lambert
Date: Wednesday, August 04, 1999 10:35 AM


I am please to announce v3.0 of my scribe spell checker macro is now available for download at my site : http://www.pinc.com/~slambert

V3.0 allows you to setup and switch between sets of Vedit Dictionary Files. I call these VDF sets languages, and you can have a different language for spell checking and word lookup. Languages can also be file extension specific, this feature can be turned on or off. Please see the Scribe Feature List page for more info.

Comments, Bug Reports and Suggestions Welcomed.

Scott Lambert

 


Topic: Scribe v3.0 (2 of 4), Read 208 times
Conf: VEDIT User Applications
From: Scott Lambert
Date: Friday, October 08, 1999 10:14 AM

On 10/08/1999 1:04:06 AM, Peter Rejto wrote:
>Thanks Scott. Scribe 3.0 is
>very nice.

You are welcomed, feedback is appreciated.

>Here are my personal reaction:
>I would very much like to have
>the buffer displayed that I
>scribe is spell checking. In
>fact, I would also like to see
>the word in the buffer that
>scribe has picked out.

So ideally you would want to have the buffer that scribe is working on displayed, and the current word that is being checked highlighted in someway, correct ?

Or would something like

Checking : xxxx

where xxx is the current word being checked, and have it updated as scribe works be alright.

The first option may take a major rewrite, or it may just require another submacro. I would have to think it out better. The second should be easy to implement.

The "problem" here is a question of features versus speed, since Scribe is a macro, any additional code in the spell check process will slow things down, in particular if it must update the screen between words. It is something to ponder.

I am hoping to do a rewrite of Scribe when Vedit Plus v5.2 or v6.0 comes out to take advantage of the new commands since v5.0. Depends on what goodies Ted & Tom manage to cram into v5.2 as far as programming commands.

>
>Incidentally, I do appreciate
>the option of looking up the
>steps involved in spell
>checking by studying your
>system of
>.vdm files.
>
I am glad someone can understand my code. Thank you for your suggestions, if I can do it, the feature will be implemented.

Scott

 


Topic: Scribe v3.0 (3 of 4), Read 219 times
Conf: VEDIT User Applications
From: Pauli Lindgren
Date: Monday, October 11, 1999 04:38 AM

On 10/08/1999 10:14:57 AM, Scott Lambert wrote:
>
>So ideally you would want to have the
>buffer that scribe is working on
>displayed, and the current word that is
>being checked highlighted in someway,
>correct ?

Well, I did just that enhancement to Scribe 2.* long time ago, and some other minor enhancements, too. I have not yet had time to do the same for Scribe 3.0. It is actually quite easy. You just need to use a smaller window for Scribe, instead of full screen window. The erroneous word is automatically highlighted.

>
>The "problem" here is a question of
>features versus speed, since Scribe is a
>macro, any additional code in the spell
>check process will slow things down, in
>particular if it must update the screen
>between words. It is something to
>ponder.

It does not slow down the spell checking at all. The display does not need to be updated between every word, only when an error is found. And you do not even need much code to do that (IIRC, you need to but an update() command somewhere).

The problem with Scribe is that it is very slow. If I have a file containing 10 lines, and the first error is on line 2, it takes about 10 seconds to find the first error and about 3 minutes to suggest a correction.

BTW, I noticed that if you have more than one text file open, Scribe runs nearly two times slower compared to when you have just one file open. I think it has something to do with the amount of memory reserved for a buffer on Vedit.

-- Pauli

 


Topic: Scribe v3.0 (4 of 4), Read 233 times
Conf: VEDIT User Applications
From: Scott Lambert
Date: Monday, October 11, 1999 10:02 AM

On 10/11/1999 4:38:10 AM, Pauli Lindgren wrote:

>You just need
>to use a smaller window for Scribe,
>instead of full screen window. The
>erroneous word is automatically
>highlighted.

I am glad to hear it is easy to do. When I get some time, I will give it a try.

>
>The problem with Scribe is that it is
>very slow. If I have a file containing
>10 lines, and the first error is on line
>2, it takes about 10 seconds to find the
>first error and about 3 minutes to
>suggest a correction.

That may be because it is always calling up submacros and data files from disk, but with modern disk caching, it should not be a major factor.

Scribe has to conduct 3 searches to get a word from a buffer, then it has to search for the word in the dictionaries, assuming worst case, that is another four searches, and with upto 9 special dictionaries, that comes to worst case 16 searches, before it even brings up the correction submenu.

>
>BTW, I noticed that if you have more
>than one text file open, Scribe runs
>nearly two times slower compared to when
>you have just one file open. I think it
>has something to do with the amount of
>memory reserved for a buffer on Vedit.
>

That would apply I think only in the dos version, in the windows version, Vedit should be able to grab as many 64K chunks as it needs. And the only file that vedit would need to auto buffer is the main dictionary english.vdf.

And if you are trying to spell check a file over 64k with Scribe, you have too much time on your hands.

Scott