Drat, I had a function that did this very thing, but I replaced it with the
dbschema_getTable() function in sqlYoga which returns everything you would want
to know about a table schema. I was not aware that the CREATE TABLE command
worked with sqLite.
Bob
On Dec 22, 2011, at 9:34 AM, Pete w
Yes, I know about that but the project I'm working on requires me to parse
out all the individual elements of the CREATE statement, not actually
create the table. The PRAGMA commands provide some information about the
table but you have to parse the CREATE command to get the rest.
On Thu, Dec 22,
On Dec 22, 2011, at 7:17 AM, Jim Hurley wrote:
> Dave,
>
> Ah! Those mighty regular expressions. Sm day I'll have to study up on
> them.
DO YOU VALUE YOUR SANITY??? DON'T DO IT
___
use-livecode mailing list
use-livecode@lists.runrev.com
Pleas
Open the sqLite_master table. In it you will find a column called SQL with
entries for how to create each table.
Bob
On Dec 21, 2011, at 2:08 PM, Pete wrote:
> I think I tried that but the arguments aren't always comma delimited
> unfortunately. Somewhat related,I wish SQLite provided more f
Dave,
Ah! Those mighty regular expressions. Sm day I'll have to study up on them.
Yes that works very nicely.
The other method (one that I understand ;) ) now boils down to this:
on mouseUp
put field 1 into tText
put containsWord("time", tText)
end mouseUp
function containsWord tWor
Very good Hugh. This adds another level of versatility, first by the addition
of a "rule" and then by returning a word count.
You are right about what constitutes a word. That is what is nice about having
a "rule".
BTW: The previous "token" method can be changed into a count by replacing this
Late to the thread.
If the purpose is just to see whether the word exists in a string, would this
work (using regular expressions)?
on mouseUp
put "Some timely text with time, and more." into tString
put "time" into tTarget
put containsWord(tTarget, tString)
end mouseUp
function conta
Aha. So we are now only testing 'exist', and not the word number? Since I've
already written this, I'll post it anyway...
The problem (as always when this topic is raised) is the definition of a
'word', as indicated by the additional inclusion of 'token' in the language,
and thus the definition of
Strike most of my last message. It appears that most of the function can be
replaced with an examination of the entire text (dah) as in:
put tWord is among the tokens of tList into tTest
return tTest
This tests the whole text; it is not necessary to test each string containing
the wor
Thanks to all for their help with this. I learned a new key word in "token".
So far the function below handles everything reasonable I have thrown at it,
including finding "time" in the less than reasonable text in field 1:
"Now is timely the timeless time.-for, all good."
on mouseUp
pu
True enough, except that a comma *might* be there under normal circumstances,
whereas I cannot think of an example where two words can be joined by a comma
in normal English grammar. But again, it might not be english grammar we are
trying to work with, so any function devised could not be so ge
I think I tried that but the arguments aren't always comma delimited
unfortunately. Somewhat related,I wish SQLite provided more formatted
access to the table structures. Some stuff is available via the PRAGMA
statements but for other information, there's no alternative but to parse
the CREATE TA
Bob-
Wednesday, December 21, 2011, 12:54:46 PM, you wrote:
> I suppose it could be argued that everything would break down if
> the text were "now,is,the,time,for,all,goo,men..." but at some point
> we have to presume the absence of the absurd? to have 2 words split
> by a comma is a problem with
OIC good point. But wouldn't it be better simply isolate the comma delimited
arguments (usually in parens) in an SQL statement and then deal with them as
items? But I suppose there are all kinds of special circumstances that would
make a general purpose word function very unwieldy.
Bob
On Dec
Not really - try parsing out an SQL SELECT statement. Granted that's not
really English text but if you want to make this a general purpose
function, you have to deal with that circumstance. I needed to do that and
ended up going through the text and inserting a space after each comma if
there wa
I suppose it could be argued that everything would break down if the text were
"now,is,the,time,for,all,goo,men..." but at some point we have to presume the
absence of the absurd? to have 2 words split by a comma is a problem with the
data, not with the software.
Bob
On Dec 21, 2011, at 12:0
On Dec 21, 2011, at 12:28 PM, Jim Hurley wrote:
> where I just replace the quote with a carrot or whatever
You must mean caret? ;-) Otherwise I think you just found an algorithm that
could solve the problem of world hunger!
Bob
___
use-livecode m
Plus it doesn't work for the case where the sentence ends with a period and
you search for "men" - that returns zero.
On Wed, Dec 21, 2011 at 12:09 PM, Mark Wieder wrote:
> FlexibleLearning-
>
> Wednesday, December 21, 2011, 12:00:29 PM, you wrote:
>
> > An alternative approach:
>
> > Assuming t
Thanks, Hugh.
Unfortunately there are a couple of problems. It won't work for this line:
Now is the time.
because of the period.
Nor for the line
"Now is the time for all"
because of the quotes.
Stripping the lead and trailing characters is cumbersome, but fairly
reliable--except for
FlexibleLearning-
Wednesday, December 21, 2011, 12:00:29 PM, you wrote:
> An alternative approach:
> Assuming that the phrase "now is the time, for all good men" is in fld 1,
> then this button script
> on mouseUp
> put findWord(fld 1,"time")
> end mouseUp
> function findWord pContent,pS
An alternative approach:
Assuming that the phrase "now is the time, for all good men" is in fld 1,
then this button script
on mouseUp
put findWord(fld 1,"time")
end mouseUp
function findWord pContent,pStr
if pStr is among the tokens of pContent then
return num of words of char 1 to o
Bob et. al.
>
> Message: 14
> Date: Wed, 21 Dec 2011 08:24:59 -0800
> From: Bob Sneidar
> To: How to use LiveCode
> Subject: Re: is among the words AND find words
> Message-ID:
> Content-Type: text/plain; charset=us-ascii
>
> I think this underscores the ne
Before rushing for a tokenOffset feature request, perhaps it might be
worth considering some other options. Perhaps there could be a way of
signalling a string of items which would mark either what should be
considered a non-word, or a non-token. That way such a search feature
could be used with
Feature request! Sign me up!
Bob
On Dec 21, 2011, at 10:12 AM, Pete wrote:
> The token keyword works with "time," and "time-bomb" (assuming you want to
> find just "time"). I've found it to be really useful in a number of
> circumstances in parsing out chunks of text from strings that don't fit
The token keyword works with "time," and "time-bomb" (assuming you want to
find just "time"). I've found it to be really useful in a number of
circumstances in parsing out chunks of text from strings that don't fit the
standard word model - like SQL statements which can include commas,
parentheses
I think this underscores the need for the words keyword to be upgraded to
reflect real text. For instance, word delimiters could be a property containing
all the characters which might be word delimiters, all the punctuations for
example. I don't know how you would treat a hyphen.
Upon thinkin
Hmmm ... what will happen to "there is time,enough for it" - NB no space
before or after the comma.
I think you *want* to find "time" in that case - but I'm not sure if you
will by stripping out all non-letter characters from the word "time,enough".
Would it not be simpler (and faster) to fin
No worries Pete!
Phil
On 12/21/11 12:09 AM, Pete wrote:
Whoops, sorry Phil, replied before I saw your post.
On Tue, Dec 20, 2011 at 9:41 PM, Phil Davis wrote:
Hi Jim,
Try "token" instead of "word" as the chunk you're looking for. It will
work.
On 12/20/11 4:42 PM, Jim Hurley wrote:
If
Whoops, sorry Phil, replied before I saw your post.
On Tue, Dec 20, 2011 at 9:41 PM, Phil Davis wrote:
> Hi Jim,
>
> Try "token" instead of "word" as the chunk you're looking for. It will
> work.
>
>
> On 12/20/11 4:42 PM, Jim Hurley wrote:
>
>> If the sentence:
>>
>> "Now is the time, for all
I need is
> >
> > put wordOffset("time", "Now is the time, for all good men.")
> >
> > to show 4 when wholeMatches is true
> >
> > It appears that there is no way in LiveCode to find whole words in a
> variable reliably.
> >
Hi Jim,
Try "token" instead of "word" as the chunk you're looking for. It will work.
On 12/20/11 4:42 PM, Jim Hurley wrote:
If the sentence:
"Now is the time, for all good men."
is within the field "myField" , then
find word "time" in field "myField"
is successful. The word "time" is
ue
>
> It appears that there is no way in LiveCode to find whole words in a variable
> reliably.
>
> Wish there were.
>
> Jim Hurley
>
>
>
>> From: Bob Sneidar
>> To: How to use LiveCode
>> Subject: Re: is among the words AND find words
>&g
iveCode to find whole words in a variable
reliably.
Wish there were.
Jim Hurley
> From: Bob Sneidar
> To: How to use LiveCode
> Subject: Re: is among the words AND find words
> Message-ID:
> Content-Type: text/plain; charset=us-ascii
>
> Not sure if wholeMatches af
Not sure if wholeMatches affects the find command, but if you set wholeMatches
to true and then use wordOffset("time",field "myField") then I think you will
get the results you are looking for. wordOffset() returns 0 when it fails to
find a match.
Bob
On Dec 20, 2011, at 4:42 PM, Jim Hurley
If the sentence:
"Now is the time, for all good men."
is within the field "myField" , then
find word "time" in field "myField"
is successful. The word "time" is found despite the fact that it is not
delimited with spaces on either side.
But
"time" is among the words of "Now is the t
35 matches
Mail list logo
