Simple MatchChunk question

[kaveh1000]

Can someone tell me why the following code does not work? I get empty for tStart and tEnd. Should I not get the offset of "e"?

Code: Select all

on mouseup
   put empty into tStart; put empty into tEnd
   get matchChunk("Hello", "e", tStart, tEnd)
   answer tStart && tEnd
end mouseup

[richmond62]

Well, I tried some code of my own:

Code: Select all

on mouseUp
   put "my socks are yellow" into rmSTRING
   get matchChunk (rmSTRING, "k", xStart, xEnd)
   answer xStart && xEnd
end mouseUp

and got NOTHING.

Also with this:

Code: Select all

on mouseUp
   --put "my socks are yellow" into rmSTRING
   get matchChunk ("my socks are yellow", "k", xStart, xEnd)
   answer xStart && xEnd
end mouseUp

Which surprised me given what the Dictionary says.
-

-
But, then I had a thought and tried this:

Code: Select all

on mouseUp
   put "my socks are yellow" into fld "fff"
   get matchChunk (fld "fff", "k", xStart, xEnd)
   answer xStart && xEnd
end mouseUp

and that didn't work either.

Nor did it work in LiveCode 8.2.0 dp-2.

[kaveh1000]

From what I remember, this should be looking at the offset of "e" inside "Hello", so tStart should return as 2.

[richmond62]

From what I remember

Where do you remember this from as it has obviously been a "dead duck" for quite some time?

I think you may be doing something wrong, because if I do this:

Code: Select all

on mouseUp
   --put "my socks are yellow" into rmSTRING
   put matchChunk ("my socks are yellow", "k", xStart, xEnd)
   --answer xStart && xEnd
end mouseUp

The messageBox returns TRUE.

And if I do this:

Code: Select all

on mouseUp
   --put "my socks are yellow" into rmSTRING
   get matchChunk ("my socks are yellow", "k", xStart, xEnd)
   put xStart && xEnd
   --answer xStart && xEnd
end mouseUp

I get nothing at all.

This is probably because the variable xStart should contain the number
of the char in the string where the matchChunk should start looking.

[bn]

Hi Kaveh,

matchChunk works also with regular expressions.

I think what you need is a regular expression if you want the offset of the char.

From the dictionary:

matchChunk(field "Title",myExpr,startMatch,endMatch)

The code below works in your example. But I am notoriously bad at regex.

Code: Select all

on mouseup
   put empty into tStart; put empty into tEnd
   get matchChunk("Hello", "([e])", tStart, tEnd)
   answer tStart && tEnd
end mouseup

I think only Thierry could really help here.

Kind regards
Bernd

[richmond62]

Interestingly enough:

Code: Select all

on mouseUp
   put empty into xStart;  put empty into xEnd
   get matchChunk ("my socks are yellow", "[(k)]", xStart, xEnd)
   answer xStart && xEnd
end mouseUp

Does NOT work.

[bn]

Interestingly enough this does work

Code: Select all

on mouseUp
   put empty into xStart;  put empty into xEnd
   get matchChunk ("my socks are yellow", "([k])", xStart, xEnd)
   answer xStart && xEnd
end mouseUp

Kind regards
Bernd

[kaveh1000]

OK, I got it. We need to grab the item we are interested in with brackets. so:

Code: Select all

on mouseup
   put empty into tStart; put empty into tEnd
   get matchChunk("Hello", "(e)", tStart, tEnd)
   answer tStart && tEnd
end mouseup

returns

Code: Select all

2,2

[kaveh1000]

Going further,

Code: Select all

on mouseup
   get matchChunk("Hello", "(el).*(o)", tOne, tTwo, tThree, tFour)
   answer tOne && tTwo && tThree && tFour
end mouseup

produces

Code: Select all

2 3 5 5

[bn]

Hi Kaveh,

you are cheating by using regex.

Kind regards
Bernd

[richmond62]

The Dictionary needs a spot of editing.

"([k])"

But for the right sort of brackets and their order a program was lost.

[kaveh1000]

@Bernd: But MatchChunk is specifically for regex I think.

[bn]

@Kaveh,

for me some of the examples did not look like regular expressions but the generic description of matchChunk in the dictionary

matchChunk(string, regularExpression [, positionVarsList])

suggests so. (I said I was guessing at regex) But I stay away from matchChunk but use charOffset instead.

Kind regards
Bernd

[kaveh1000]

Ah, but I have varying text, so e.g. in the following XML text:

Code: Select all

                 <Measure>
			<MeasureType>08</MeasureType>
			<Measurement>2137</Measurement>
			<MeasureUnitCode>gr</MeasureUnitCode>
		</Measure>
		<Measure>
			<MeasureType>01</MeasureType>
			<Measurement>234.0</Measurement>
			<MeasureUnitCode>mm</MeasureUnitCode>
		</Measure>

I need to find each chunk inside the <Measure> tags. Copy one, then the other. only regex will allow me to do that.

[dunbarx]

only regex will allow me to do that.

Not sure that is true. You could certainly use "ordinary" LC with its offset functions, then find the end tag, and put it all together, but these certainly are not as compact as regex.

Craig