Page 2 of 2
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 3:00 pm
by MachineCode
Still not much of a big deal, because you can just expand the borders a bit. In your examples, what is next after the last ">"? Maybe a #CRLF$? So include that as the right border, ie. ">"+#CRLF$. Easy. But the point of the tip is, to simply pluck something out of two borders. It's up to the coder to decide what the borders are.
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 3:08 pm
by TomS
@MachineCode: You can't just expand the borders. Some tags may includes spaces, some not. Some attributes may be encapsulated by doublequotes, some by singlequotes, some are not encapsulated at all.
You can't check for all those variations. That's bad style and very slow.
@Trond: HTML doesn't allow single quotes. It's the browsers that allow it. In HTML every value that is not numeric (positive integer), must stand between doublequotes.
<table border=1 width="100%">
<table border=true width=100%>
<table border="true">
Just because it works doesn't mean it's correct. That's why there are so many sites that only look good in one browser.
Now one has to decide, whether to stick to the rules and ignore bad codes, or to try to compensate for all imaginable errors.
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 3:33 pm
by MachineCode
TomS wrote:You can't check for all those variations
I know. But for my use, I would be testing for the same specific borders, so it works for me. And the request for such a command is a good one anyway, for general-purpose non-HTML plucking.
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 4:03 pm
by Trond
TomS wrote:@Trond: HTML doesn't allow single quotes. It's the browsers that allow it. In HTML every value that is not numeric (positive integer), must stand between doublequotes.
Where did you get that information? The standard does not agree with you.
http://www.w3.org/TR/html401/intro/sgmltut.html#h-3.2.2
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 4:27 pm
by TomS
Well, seems I'm not totally up2date.
Here it says, that single quotes are allowed, too.
http://www.w3schools.com/html/html_attributes.asp
But I think that's just because all browsers accept it.
It's a trend to use single quotest in php, asp etc, because programmers don't want to escape the quotes.
echo "this is a <a href='target' >hyperlink</a>";
I'm pretty sure that this was not valid in html before 4...
Anyways, I don't like the style. Just as I hate double quotes in php, but that's just me.
It also says: "always quote". Only numerical values can stand unquoted.
Althought the majority of browser doesn't seem to have a problem with that either.
I remember IE accepting
height, heigth, and heigt as an element attribute.
Again: Just because it works doesn't mean it's correct.
EDIT: As for the feature request: Have a look at this post and the follow ups
http://purebasic.fr/english/viewtopic.p ... 94#p353094
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 5:03 pm
by kenmo
Just to add one more similar solution:
Code: Select all
Procedure.s Middle(String.s, Open.s = "", Close.s = "", Start.i = 1, NoCase.i = #True)
Protected iOpen.i, iClose.i, opLen.i
Protected qString.s, qOpen.s, qClose.s
Protected Result.s
If (NoCase)
qString = LCase(String)
qOpen = LCase(Open )
qClose = LCase(Close )
Else
qString = String
qOpen = Open
qClose = Close
EndIf
If (Open)
opLen = Len(Open)
iOpen = FindString(qString, qOpen, Start)
If (iOpen)
iClose = FindString(qString, qClose, iOpen + opLen)
If (iClose)
Result = Mid(String, iOpen + opLen, iClose - iOpen - opLen)
Else
Result = Mid(String, iOpen + opLen)
EndIf
Else
Result = ""
EndIf
Else
If (Close)
iOpen = Start
iClose = FindString(qString, qClose, Start)
If (iClose)
Result = Left(String, iClose - 1)
Else
Result = ""
EndIf
Else
Result = String
EndIf
EndIf
ProcedureReturn (Result)
EndProcedure
[/size]
This one is a little different. If the closing (or opening) delimiter isn't found, it absorbs everything right up to the string boundary. Same if either delimiter is left blank. So I guess the name "middle" is sort of a misnomer...
But for anything more complex, like HTML parsing, you can't go wrong with Regular Expressions.
Re: PluckString command to extract string from another
Posted: Fri May 20, 2011 8:02 pm
by Vitor_Boss®
Hi, if I understood you need split an string, I wrote a code to do that returning result as ARRAY. Enjoy
Code: Select all
Procedure Split(Array Result.s(1), Expression.s, Delimiter.s, Limit.l=-1)
Protected.i i, ii, C
Protected.l Length, Size, Position
ii = CountString(Expression, Delimiter)
If ii=0
ReDim Result(0)
Result(0) = Expression
Else
If Limit > 0 And ii > Limit-1
ii = Limit-1
EndIf
ReDim Result(ii)
Size = Len(Delimiter)
Position = 1
For C = 0 To ii-1
Length = FindString(Expression, Delimiter, Position) - Position
Result(C) = Mid(Expression, Position, Length)
Position + Length + Size
Next
Result(C) = Mid(Expression, Position)
EndIf
EndProcedure