--- In peditors@y..., Paul Nevai <nevai@m...> wrote:
> Dear Japanese peditor guys [if any]:
>
> If you use regular expressions or TeX and such, then what do you
use instead
> of "\" [backSlash]? Do you simply use "yen" which is the same as
backSlash or
> some other symbol? Thanks! Best regards, Paul

I don't use Unix, but quick search on Tex on-line tutorial in
Japanese site shows it is indeed "yen" symbol that is used.
for example.

\begin{verbatim}
#$%^&*()
(^^;)
\end{verbatim}

Of course, you are now completely confused.

In my browser or any japanese text enabled viewer, symbols
before "begin" and "end" are in fact 'yen' symbol.  I suspect that
you are actually seeing 'backslash' because in your original quote, a
symbol in "[this symbol]" [backslash] is indeed "yen" symbol on my
screen. Sorry for poor explanation. I believe the point is that ascii
character x5c (or whatever that was, I don't have ascii table handy)
that is assigned to backslash is assigned to "yen" symbol in Japanese
ascii set. I don't know the history behind it, (so, let's blame
Microsoft) it was always a case in DOS-PCs. In Japanese DOS prompt,
path delimiter is "yen" symbol, not backslash symbol, etc.,
Therefore, to me, it was refreshining to see both yen and backslash
in Palm devices indeed.

Masaki Adachi

# #DO: John Harms photojhh@... 30-Sep-2001
# LapTopHack version 5.76 (12:12:49 2001/09/24)
# Need way to directly access a pScript button page

Hi JohnH:

Done. Will send you for testing soon. Maye today.

USAGE: "ESC B x B" where x is any number between 0 and 999 [no need for
leading zeros].

Best regards, Paul

Dear Adachi: [this is your "first" name, right?]

# I don't use Unix, but quick search on Tex on-line tutorial in
# Japanese site shows it is indeed "yen" symbol that is used.
# for example.
#
# \begin{verbatim}
# #$%^&*()
# (^^;)
# \end{verbatim}
#
# Of course, you are now completely confused.

Not at all. I understood you 100%.

# In my browser or any japanese text enabled viewer, symbols
# before "begin" and "end" are in fact 'yen' symbol.  I suspect that
# you are actually seeing 'backslash' because in your original quote, a
# symbol in "[this symbol]" [backslash] is indeed "yen" symbol on my
# screen. Sorry for poor explanation. I believe the point is that ascii

No, no, no, your explanation is 100% perfect.  Many many thanks.  It will
help me a lot when I do regular expressions. Best regards, Paul

P.S. As in Japanese, in Hungarian too: last_name first_name.

> # Need way to directly access a pScript button page
> Done. Will send you for testing soon. Maye today.
> USAGE: "ESC B x B" where x is any number between 0
> and 999 [no need for
> leading zeros].

Thanks!

-- John


__________________________________________________
Do You Yahoo!?
Listen to your Yahoo! Mail messages from any phone.
http://phone.yahoo.com

> # 2. Aho, Alfred V. et al. "Compilers Principles, Techniques,
> and Tools"
> # 1986 Addison-Wesley
> http://www.aw.com/product/0,2627,0201100886,00.html (If
>
> This was written by Jeffrey D. Ullman as co-author who, guess
> what, is the brother of one of my close friends [Joe died about
> 6 years ago].

And I used to study from this book for my compiler building course in
college. I still vividly remember it by both its cover and its
authors.

Talking about regex and compilers, Alfred Aho also lends his name and
ingenuity to AWK (a killer programmable text filter under Unix),
together with Jim Weinberger and Brian W. Kernigan, while the latter
co-developed the C programming language with Dennis M. Ritchie (I'm
recalling these names from memory so forgive me for any typos or name
swapping :)). Isn't it a small world?

Roy.

i tried to find a way to understand how pScripts work. it took quite a
time to get the helping input from the manual and pScriptors right. i
want to tell and show you my findings to hear if i am right.

- hypothesis a: chains are the pScript-way for repeating tasks
- hypothesis b: chains are the pScript-way to get things in the right
order
- hypothesis c: recursive loops don't work (very well) in pScripts

definitions:
- chain - a series of pScripts. 1 calls 2 at its end, 2 calls 3 at its
end and so on. the last may call the first to repeat a task.
- recursive loop - one basic pScript calling itself (at its end),
calling a series of pScripts 1, 2 and 3 ... (this is the programming way
with a main routine and subroutines)

explanation:
one pScript completely finishes and empties the queues when it is closed
(it reaches "}"). it then starts the next called pScript. a loop partly
processes the script and partly goes on filling the queues leading to
unpredictable results and the "runtime calls over the limit" message.

i have a visual imagination on pScripts going through chains
(http://groups.yahoo.com/group/peditors/files/pedit%20Scripting/Tips%20a
nd%20Techniques/testlab%20-%20loops%20and%20chains/chain.gif) and loops
(http://groups.yahoo.com/group/peditors/files/pedit%20Scripting/Tips%20a
nd%20Techniques/testlab%20-%20loops%20and%20chains/loop.gif) which i
want to show in 2 animated gifs to illustrate the context (saved to the
files area, see urls)

conditions:
+ KeyQSize = 55 (to immediately see the effect)
+ have the /&runtimeInit [] pFunktion handy for experimenting (i have it
on a customButton and call it every time i get the "runtime calls over
the limit" message)
+ build and tested only under LTH and pedit 5.76

why is the test as it is?
i constructed the test with only /&mess [] pFunctions to show the way
the script goes and to avoid additional influences. both units do the
same in different ways. they are build the same way, the differences
only depend on the different realizations.

how loop-lab works:
first initialize var $0 and call the main loop
main creates output and calls pScript 2 (output) pScript 3 (output),
pScript 4 (breaking decision) and calls itself
----+----+

custBtn.004.1
{loop-lab::
     /&varSet [$0, 0]
     /&script [@@loop-lab-main@@]
}

{loop-lab-main::
     /D0
     /&mess [@@1.1@@,,30]
     /&mess [@@1.2@@,,30]
     /&mess [@@1.3@@,,30]
     /&script [@@loop-lab-01@@]
     /&script [@@loop-lab-02@@]
     /&script [@@loop-lab-03@@]
     /&script [@@loop-lab-main@@]
}

{loop-lab-01::
     /D0
     /&mess [@@2.1@@,,30]
     /&mess [@@2.2@@,,30]
     /&mess [@@2.3@@,,30]
}

{loop-lab-02::
     /D0
     /&mess [@@3.1@@,,30]
     /&mess [@@3.2@@,,30]
     /&mess [@@3.3@@,,30]
}

{loop-lab-03::
     /D0
     /&varAdd@ [$0, 1]
     /&ifAbort@ f [$0 == 5]
}
----+----+


how chain-lab works:
first initialize var $0 and call link 1
chain-lab-01 creates output and calls link 2
chain-lab-02 creates output and calls link 3
chain-lab-03 creates output and calls link 4
chain-lab-04 contains the breaking decision, calls link 1 when false,
link 5 when true
chain-lab-05 output "done"
----+----+

custBtn.004.2
°0  {chain-lab::
     /&varSet [$0, 0]
     /&script [@@chain-lab-01@@]
}

°1  {chain-lab-01::
     /D0
     /&mess [@@1.1@@,,30]
     /&mess [@@1.2@@,,30]
     /&mess [@@1.3@@,,30]
     /&script [@@chain-lab-02@@]
}

°2  {chain-lab-02::
     /D0
     /&mess [@@2.1@@,,30]
     /&mess [@@2.2@@,,30]
     /&mess [@@2.3@@,,30]
     /&script [@@chain-lab-03@@]
}

°3  {chain-lab-03::
     /D0
     /&mess [@@3.1@@,,30]
     /&mess [@@3.2@@,,30]
     /&mess [@@3.3@@,,30]
     /&script [@@chain-lab-04@@]
}

°4  {chain-lab-04::
     /D0
     /&varAdd@ [$0, 1]
     /&ifScript@ [$0 != 5, @@chain-lab-01@@]
     /&script [@@chain-lab-05@@]
}

°5  {chain-lab-05::
     /D0
     /&mess [@@done@@,,]
}
----+----+


the results:
loop-lab produces "1.1" "1.2" "1.3" "runtime calls over the limit"
"runtime calls over the limit" "2.1" "2.2" "2.3" "3.1" "3.2" "1.1" "1.2"
and then ends - "3.3" doesn't show up

chain-lab produces 5 times "1.1" "1.2" "1.3" "2.1" "2.2" "2.3" "3.1"
"3.2" "3.3" and then "done"

i think hypothesis a, b and c are verified :-)
by the way i tested chain-lab with 100 loops and it worked fine.


question:
- are recursive loops possibly somehow similar to a long pScript?

i was very happy with this cognition and hope it will help some of you
to get their pScripts working.

regards
urs

ps. paul's recurrent suggestion is to chain pScripts. in the beginning i
read it but i didn't understand it because i thought on loops and tried
loops with no result - but it settled in my brain. after struggling with
a repeating task which i could by no means get to work i tried to make a
chain of the pScripts - and it worked. this was the trigger for the
"testlab".

pps.
norbert preining wrote in his posting "Iterating in pedit (funny
scripting ;-)" Do 2001-09-27 14:01
> I have to call runtimeinit because very fast I get to
> "Runtime calls over the limit"
>  What can I do against this? Really, this IS a problem. See the
paragraph
>  on further development.
norbert, your problem relating to this point will imho be minimized by
chaining.

--
urs hunkler                                     u.hunkler@...
wir design gmbh                                 fon +49. 5 31. 8 88 10
d-38118 braunschweig                            fax +49. 5 31. 8 88 12
34
sophienstrasse 40                               germany

# - hypothesis a: chains are the pScript-way for repeating tasks

Yes. Preferably combined with /&ifScript[@] [...].

# - hypothesis b: chains are the pScript-way to get things in the right
# order

Correct.

# - hypothesis c: recursive loops don't work (very well) in pScripts

Probably. I need to study your examples. Maybe some of you guys will and
then I don't need to.

BTW, recursive loops can get real nasty. Keep in mind that the Palm has 3
different queues: event, pen, and key.

AMB=Paul

On Mit, 03 Okt 2001, Paul Nevai wrote:
> # - hypothesis c: recursive loops don't work (very well) in pScripts
>
> Probably. I need to study your examples. Maybe some of you guys will and
> then I don't need to.

Well I have some recursive loops in my iteration pack and it works
well. You must of course call with ifScript...


Best wishes

Norbert

-----------------------------------------------------------------------
Norbert Preining <preining@...>
University of Technology Vienna, Austria            gpg DSA: 0x09C5B094
-----------------------------------------------------------------------
WRITTLE (vb.)

Of a steel ball, to settle into a hole.

			 --- Douglas Adams, The Meaning of Liff

> Very interesting thing. I have to digest this. But I know most
> of these things from problems I tried to circumevent, but
> you described all the problems very clearly.

it took me hours to get everything together and my english dictionary to
write it down ;-)

>
...

>
> Well I tried to chain the scripts, but still got the message. THe
problem
> looks to me that I have around 10 indirect calls, partly with @ and
> partly without, and all this seems to screw up. I erased all I did
> wrt selection and will do a clean new start.

i become used to make the pScripts smaller and smaller until they work.
at the beginning it was strange to me to have small scraps of code sewed
together with /&script [@] calls. but since I recognised the benefit i
start enjoying this way of reduction.

paul wrote Mi 2001-10-03 17:39 to this topic
> BTW, recursive loops can get real nasty. Keep in mind that the Palm
has 3 different queues: event, pen, and key.

this may be the reason, why some of them have to become very small to
just keep the information for one queue.

>
> Best wishes
>
> Norbert
>

regards
urs

--
urs hunkler                                     u.hunkler@...
wir design gmbh                                 fon +49. 5 31. 8 88 10
d-38118 braunschweig                            fax +49. 5 31. 8 88 12
34
sophienstrasse 40                               germany

# > Well I tried to chain the scripts, but still got the message. THe problem
# > looks to me that I have around 10 indirect calls, partly with @ and
# > partly without, and all this seems to screw up. I erased all I did wrt
# > selection and will do a clean new start.

pScripting has a limit of 8 "runtime" calls of each function. Best regards,
Paul

On Wed, 3 Oct 2001 16:42:08 +0200, Urs <u.hunkler@...> wrote:

> i tried to find a way to understand how pScripts work. it took quite a
> time to get the helping input from the manual and pScriptors right. i
> want to tell and show you my findings to hear if i am right.

I do not currently have the level of pScripting proficiency required to
tell whether you are right. But I have saved your message for further
reference because, even if your model is not accurate, your thoughts
provide useful insights for further investigation and experimentation on
how pScripts work.

Thanks,


Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://web.mclink.it/amoroso/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]

On Tue, 02 Oct 2001 21:26:16 -0000, Roy van der Woning <rvdw@...>
wrote:

> > # 2. Aho, Alfred V. et al. "Compilers Principles, Techniques,
> > and Tools"
[...]
> And I used to study from this book for my compiler building course in
> college. I still vividly remember it by both its cover and its

Speaking of useful resources on formal languages and parsing, I recalled
that there was this link buried among my bookmarks:

   Parsing Techniques: A Practical Guide
   http://www.cs.vu.nl/~dick/PTAPG.html

The full text of the book may be downloaded freely. It includes an
extensive bibliography.


Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://web.mclink.it/amoroso/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]

BUG: Mark A. Mandel 2001-10-03

description: None of the modifiers listed in the manual under "The FIND
SilkScreen button" work for me, with or without FindHack running. I am
running pedit 5.76 on a Visor Deluxe with no Springboard modules installed.
Hacks, in top-to-bottom order from HackMaster (v1) display; parenthesized
version numbers are as reported by SoftTrack, others are in the name as
shown by HackMaster:

  Switcheroo (1.3)
  LTH 5.76
  SwitchHack (1.62)
  Swipe! 1.0
  PopUp Calculator (0.97)
  PopMe! (3)

(the following two hacks are disabled)
  ClipHack (1.0)
  FindHack (1)

Trying the example given in the manual, in the paragraph beginning

      ".^" gives you the titles of all your memos
      in the current category. Try it in ListView.

without FindHack, in pedit's ListView, I get

      Matches for ".^"
      --- memos via pedit --- 1
      --- pedit32 memos ----- 2
      --- Datebk3 ----------- 3
           [...]
      --- Memos ------------- 7
      FIND SilkScreen button in pedit   8
      The FIND SilkScreen button        9

The last two are versions of the text of that section from the manual. Note
that the button finds the string ".^" (without quotes) itself... but only
in MemoPad, not in pedit!!

With FindHack and specifying "Current App", I get

      No matches for ".^"
      --- memos via pedit --- 1
      --- pedit32 memos ----- 2

Similarly, in the category that contains Paul's list of Web locations for
the latest versions, whose first line is "Dear peditors, LapTopHackers, and
peditHacker:", a search for
      .dear
finds no matches. And likewise for the other modifiers.

                   Mark A. Mandel : Senior Linguist
  Dragon Systems, a Lernout & Hauspie company : speech recognition
  320 Nevada St., Newton, MA 02460, USA : http://www.dragonsys.com

BUG: Mark A. Mandel 2001-10-03

description: I am running on a Visor Deluxe with no Springboard modules
installed. In pedit 5.76, when I select Recent from ListView and choose a
memo from the list, the cursor is where I left it in that memo. That's
fine, that's OK. HOWEVER, if the first Graffiti stroke I enter is CAPITAL
(i.e., straight upward), it often is not recognized, and if I then stroke a
letter it comes out lowercase. If I don't enter the letter, I have to
re-enter the CAPITAL stroke to get the solid up-arrow display in the lower
right corner that means the next Graffiti character will be capital.

This doesn't happen all the time, but it is reliably reproducible in a memo
that I use to log my exercise. The relevant part looks like this:

OCT 01
Sun  Mon  Tue  Wed  Thu  Fri  Sat
      1TS  T    *                   6
                                    13
                                    20
                                    27
                     31

The asterisk is not there in the memo; it marks the current cursor
location. Note that there are no SPACE characters on the lines of the
calendar grid, just TABs.


By the way, I've made a set of these as blank forms (seven: for months
beginning on each day of the week); I adjust the last line(s) to the actual
month length when copying a form into place. I use a different letter for
each kind of exercise, with capital vs. small to mark whether I've done a
complete set. Of course the forms could be used to track anything day by
day, as long as an entry wasn't wider than a single TAB. I'm still
experimenting with them. I could delete the "1" that marks the (indented)
first day of the month after putting an entry there, maybe with a special
symbol like "." or "-" to explicitly show "no entry for this day". Is there
enough interest in these for me to upload the memo?

                   Mark A. Mandel : Senior Linguist
  Dragon Systems, a Lernout & Hauspie company : speech recognition
  320 Nevada St., Newton, MA 02460, USA : http://www.dragonsys.com

--- Mark_Mandel@... wrote:
> BUG: Mark A. Mandel 2001-10-03
>
> description: None of the modifiers listed in the manual under "The FIND
> SilkScreen button" work for me, with or without FindHack running. ...
>

Check if the preference "silkscreen Find is fancy" is checked on in
Preferences.

-- Yaakov.



__________________________________________________
Do You Yahoo!?
NEW from Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1

Hi Everyone,

A bit ago there was some talk on this group regarding where to get
refurbished palms.  I just received a link to the 'reopened'
Refurbished Palm Store (click on 'get the full story' for more
details):

http://www.palmblvd.com/articles/2001/10/2001-10-1-Palm-Re-Opens.html

Take care,

Christine K.

Our mailer is dead so I announce it here:

http://www.math.ohio-state.edu/~nevai/x/pc_beta.zip

Best regards, Paul

P.S. You will probably get another announcement when our mailer
comes back.

At 1:49 +0000 4/10/01, nevai@... wrote:

>Our mailer is dead so I announce it here:
>
>http://www.math.ohio-state.edu/~nevai/x/pc_beta.zip
>
>Best regards, Paul

REGEX is here! Awsome!

Can I get in with the first bug reports...

1. I have a document which contains "[a variety of text]", and also
one or two "[]". I did a REGEX search with the expression "[*]",
which should find all of the instances of pairs of square brackets
either full or empty. The search works, but the selection of the
found item is wrong - in the case of "[]" it works fine (both chars
selected), but when it hits "[a variety of text]" only the closing
square bracket is selected. This isn't a problem now, but when
replace is enabled...

2. The same document contains "*a variety of text*". I should be able
to find these with the REGEX "\**\*", but pedit tells me this is
illegal. Presumably it is spotting the "**", but not registering that
the first "*" is backslash escaped.

Steven
--

--- In peditors@y..., nevai@m... wrote:
> Our mailer is dead so I announce it here:
>
> http://www.math.ohio-state.edu/~nevai/x/pc_beta.zip
>
> Best regards, Paul
>
> P.S. You will probably get another announcement when our mailer
> comes back.

Welcome REGEX!

Bug?
REGEX        ^(*)$
finds a line consisting only of "()" but not "(xxx)".

Best regards
Joachim

Hi Joachim:

Our mailer is still dead....

#####################################
REGEX        ^(*)$
finds a line consisting only of "()" but not "(xxx)".
#####################################

This is not a bug! This is how REGEX works. You need

REGEX       ^(.*)$
REGEX       ^(x*)$

to find "(xxx)". Please remember that the REGEX * means "zero or more
of the previous character".

Right? Paul

Hi Steven:

###
I have a document which contains "[a variety of text]", and also
one or two "[]". I did a REGEX search with the expression "[*]",
which should find all of the instances of pairs of square brackets
either full or empty. The search works, but the selection of the
found item is wrong - in the case of "[]" it works fine (both chars
selected), but when it hits "[a variety of text]" only the closing
square bracket is selected. This isn't a problem now, but when
replace is enabled...
####

Just like Joachim, you also made a little mistake. "[*]" is not supposed
to find "[Steven]". You need "[.*]". Right? This is REGEX!

Paul

Hi Steven:

###
The same document contains "*a variety of text*". I should be able
to find these with the REGEX "\**\*", but pedit tells me this is
illegal. Presumably it is spotting the "**", but not registering that
the first "*" is backslash escaped.
###

Yep, this is a bug. It will cause me a headache to fix it.
I need a run time error checking instead of the current
one which checks the errors prior to doing the search job.

Dear peditors, LapTopHackers, and peditHackers:

NEW in the latest LapTopHack? Please see the very end of LapTopHack_man.html.
NEW in the latest pedit? Please see the very end of pedit_man.html.

NOTE. If you use the peditorial and scriptorial features of LapTopHack, then
you need to get the latest pedit and LapTopHack from pc_beta.zip [pc stands
for PaulComputing]

	      http://www.math.ohio-state.edu/~nevai/x/pc_beta.zip

			  PaulComputing VERSION INFO

	     http://www.math.ohio-state.edu/~nevai/x/versions.html

		  AND [you probably don't need any of these]

	    http://www.math.ohio-state.edu/~nevai/x/pedit_beta.zip
	  http://www.math.ohio-state.edu/~nevai/x/LapTopHack_beta.zip
	      http://www.math.ohio-state.edu/~nevai/x/pScript.zip
	  http://www.math.ohio-state.edu/~nevai/x/LapTopHack_man.html
	   http://www.math.ohio-state.edu/~nevai/x/pedit_man.zip
	  http://www.math.ohio-state.edu/~nevai/x/LapTopHack_man.zip
	   http://www.math.ohio-state.edu/~nevai/palm/pedit_man.html
        http://www.math.ohio-state.edu/~nevai/x/pedit_man_annotated.zip
	  http://www.math.ohio-state.edu/~nevai/x/LapTopHack_beta.hqx
	 update info posted at http://groups.yahoo.com/group/peditors

Best regards, Paul

Name:    Paul Nevai
E-mail:  2me@...
Home:    http://www.PaulComputing.com

Archive:  http://www.math.ohio-state.edu/~nevai/x/pc_beta.zip
-rw-------  2.2 unx    1325 tx defN  3-Oct-01 17:33 pc_beta.txt
-rw-------  2.2 unx  244383 bx defN  4-Oct-97 12:45 LapTopHack.prc
-rw-------  2.2 unx  386884 bx defN  4-Oct-97 12:40 peditPro.prc
-rw-r--r--  2.2 unx  250358 bx stor  3-Oct-01 17:33 LapTopHack_man.zip
-rw-r--r--  2.2 unx  297885 bx stor  3-Oct-01 17:33 pedit_man.zip
5 files, 1180835 bytes uncompressed, 800744 bytes compressed:  32.2%

Mark_Mandel@... wrote to Paul Nevai:
#
# BUG: Mark A. Mandel 2001-10-03
#
# description: None of the modifiers listed in the manual under "The FIND
# SilkScreen button" work for me, with or without FindHack running. I am

Did you activate it via the prefs? Best regards, Paul

--- In peditors@y..., nevai@m... wrote:
> Hi Joachim:
>
> Our mailer is still dead....
>
> #####################################
> REGEX        ^(*)$
> finds a line consisting only of "()" but not "(xxx)".
> #####################################
>
> This is not a bug! This is how REGEX works. You need
>
> REGEX       ^(.*)$
> REGEX       ^(x*)$
>
> to find "(xxx)". Please remember that the REGEX * means "zero or
more
> of the previous character".
>
> Right? Paul

Hi Paul,
got it.
Thanks, Joachim

BUG: Yaakov Shlafman 2001-10-04
REGEX ending in .* highlights one character less than it should


for example, given the regexp "pe.*" in the following memo

--- Start of Memo ---
EXAMPLE
Paul Nevai is the author of pedit
--- End of Memo ---

find only highlights the string "pedi"; should be "pedit".

On Don, 04 Okt 2001, shlaf@... wrote:
> Paul Nevai is the author of pedit
> --- End of Memo ---
>
> find only highlights the string "pedi"; should be "pedit".

Well, depends on the implementation (match shortest or longest first),
but you are right, this is neither shortes nor longest ;-)

Best wishes

Norbert

-----------------------------------------------------------------------
Norbert Preining <preining@...>
University of Technology Vienna, Austria            gpg DSA: 0x09C5B094
-----------------------------------------------------------------------
FARDUCKMANTON (n. archaic)

An ancient edict, mysteriously omitted from the Domesday Book,
requiring that the feeding of fowl on village ponds should be carried
out equitably.

			 --- Douglas Adams, The Meaning of Liff

# BUG: Yaakov Shlafman 2001-10-04
# REGEX ending in .* highlights one character less than it should

Will check it. This was so far the hardest for me: to make REGEX greedy,
i.e., find the longest match. Obviously, I missed some stuff. AMB=Paul

Hi Steven:

I fixed the "\**" bug. AMB=Paul

--- Norbert Preining <preining@...> wrote:
> On Don, 04 Okt 2001, shlaf@... wrote:
>
> Well, depends on the implementation (match shortest or longest first),
> but you are right, this is neither shortes nor longest ;-)
>
> Best wishes
>
> Norbert
>

All the regex implementations I know about do "greedy" match.  In case of
"something.*" there's not very much sense to find shortest much :)

-- Yaakov


__________________________________________________
Do You Yahoo!?
NEW from Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
http://geocities.yahoo.com/ps/info1