Regular Expressions [solved]

[andybiochem]

Hi!

I'm trying to get my head around regular expressions, but it's doing my nut!

say I wanted to match the following list of strings:

12,25.5,ABC

1,97,Nam3

3,100,Z

but NOT:

12,25.5,ABC,zzz

(notice the ",zzz" bit at the end)

how do I do this with stringregexp?

So far I've got:

"([0-9]{1,}[,][0-9]{1,}[.]?([0-9]{1,})?[,][[:alnum:]]{1,})"

but this sees a match and returns it. I want the inputted string to match exactly the reg exp given....if there's more data in the string than specified in the reg exp I want it to fail.

Thanks in advance for any help!!

Edited July 30, 2008 by andybiochem

[Xenobiologist]

Hi,

you need to think of a regular method of your strings. Match all, but not if they end with zzz? Is that the rule?

Mega

[andybiochem]

Hi,

you need to think of a regular method of your strings. Match all, but not if they end with zzz? Is that the rule?

Mega

Not just zzz, anything at all after the 3rd item.

the way I'm thinking of my string to match is:

[any integer][comma][any number including decimal places][comma][any word/number/both]

if the whole string doesn't match this format then I want regexp to return = 0

but because it sees that format above within "12,25.5,ABC,zzz" it returns success (=1).

I want it to look at the whole string, not just pick and choose where it finds matches mid-string.

thanks for the reply tho!

Edited July 30, 2008 by andybiochem

[Xenobiologist]

Hi,

try this:

\d+,\d+[\.]*\d*,\w*(?=\s)

Mega

[SmOke_N]

"(?s)(\d+,\d+\.*\d*,\w+)(?:\z|\s)"

[andybiochem]

Thank you both!

SmOke_N yours seems to work for every example i've tried, so I'll go with that.

Cheers!

[andybiochem]

I eventually found some examples that wouldn't work with Smoke's reg exp

So this is how I managed to get it to work:

$string = "3,25.0,Text here" ;this should match
;$string = "zzz,123,32.33,Text Here" ;this shouldn't match
;$string = "123,32.33,Text Here...." ;this shouldn't match

;----- check string for reg exp -----
$aMatches = StringRegExp($string,"(?:\D*)(\d+,\d+.?\d?,[\w ]*)(?:\W*)",1)

;----- if no matches found... -----
If @error = 1 Then 
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- if match found not the same as original string... -----
If $aMatches[0] <> $string Then
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- other-wise good match -----
MsgBox(0,"","String format good!!!")

The important bit was the second If>EndIf section. If the 'matched' string wasn't identical to the original string, then the StringRegExp has cherry-picked the match out of a different, longer, string....thus for the purpose I needed this was a false-positive. I want to know if the WHOLE string matches the reg exp.

Maybe I've missed something obvious, but this is the only way I could find to get it to work.

Just thought I'd share.

[Gigglestick]

Yours:

(?:\D*)(\d+,\d+.?\d?,[\w ]*)(?:\W*)

Smoke_N's:

(?s)(\d+,\d+\.*\d*,\w+)(?:\z|\s)

These don't match. You said you were going to use Smoke_N's. Does it work with his?

You have a workaround. I'm just curious.

Edited July 30, 2008 by c0deWorm

[SmOke_N]

I eventually found some examples that wouldn't work with Smoke's reg exp

So this is how I managed to get it to work:

$string = "3,25.0,Text here" ;this should match
;$string = "zzz,123,32.33,Text Here" ;this shouldn't match
;$string = "123,32.33,Text Here...." ;this shouldn't match

;----- check string for reg exp -----
$aMatches = StringRegExp($string,"(?:\D*)(\d+,\d+.?\d?,[\w ]*)(?:\W*)",1)

;----- if no matches found... -----
If @error = 1 Then 
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- if match found not the same as original string... -----
If $aMatches[0] <> $string Then
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- other-wise good match -----
MsgBox(0,"","String format good!!!")

The important bit was the second If>EndIf section. If the 'matched' string wasn't identical to the original string, then the StringRegExp has cherry-picked the match out of a different, longer, string....thus for the purpose I needed this was a false-positive. I want to know if the WHOLE string matches the reg exp.

Maybe I've missed something obvious, but this is the only way I could find to get it to work.

Just thought I'd share.

You changed the situation:

Text<space>Here was not in your specs.

Edit:

You know, actually it would be MUCH better if you just posted the actual file you were trying to manipulate rather than trying to make scenarios. You leaving the slightest thing out changes an expression completely.

Edited July 30, 2008 by SmOke_N

[andybiochem]

You changed the situation:

Text<space>Here was not in your specs.

Yeah, my fault.

should have said:

[any integer][comma][any number including decimal places][comma][any words/numbers/Chr(32)s]

Really appreciated the help though, really made me think about reg exp in a different way.

[andybiochem]

You know, actually it would be MUCH better if you just posted the actual file you were trying to manipulate rather than trying to make scenarios. You leaving the slightest thing out changes an expression completely.

It wasn't actually a file I was working on. The scenario was in fact one I had made....this reg exp was a way of checking a user's input was in the correct format (a format I had devised) and that they hadn't just typed in junk.

[andybiochem]

Lol, found more faults. Now I'm working with:

"(?:\D*)(\d+,\d+(.\d+)*,[\w ]*)(?:\W*)"

@SmOke_N

Is there a less draw-out way of doing this?:

$string = "3,25.0,Text here" ;this should match
;$string = "zzz,123,32.33,Text Here" ;this shouldn't match
;$string = "123,32.33,Text Here...." ;this shouldn't match

;----- check string for reg exp -----
$aMatches = StringRegExp($string,"(?:\D*)(\d+,\d+(.\d+)*,[\w ]*)(?:\W*)",1)

;----- if no matches found... -----
If @error = 1 Then
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- if match found not the same as original string... -----
If $aMatches[0] <> $string Then
    MsgBox(0,"","String format not correct")
    Exit
EndIf

;----- other-wise good match -----
MsgBox(0,"","String format good!!!")

...ignore the msgboxes they're just there to check correct returns.

Edited July 30, 2008 by andybiochem

[therks]

I haven't analyzed everything.. but have you tried using ^ and $ to match the start and end of a string?

Just going off your first example of "[any integer][comma][any number including decimal places][comma][any word/number/both]" try this?

^\d+,\d+(\.\d+)?,[\w ]+$

[andybiochem]

I haven't analyzed everything.. but have you tried using ^ and $ to match the start and end of a string?

Just going off your first example of "[any integer][comma][any number including decimal places][comma][any word/number/both]" try this?

YES! That's it!! Thanks you!

^ and $ ...where on earth did they come from? lol!

Can this be added to the help file? Their use is mentioned here:

http://en.wikipedia.org/wiki/Regular_expression

Thanks for everyone's help! much appreciated.

[therks]

I could have sworn they were in the documentation as they're pretty standard, but when I checked sure enough they aren't there. They are mentioned indirectly with the (?m) flag. There is a lot of stuff that RegEx is capable of that's not mentioned in the AutoIt doc's though. There's a link near the top of that page which links here though, which is a more detailed description of the RegEx syntax.

[ptrex]

@Saunders / andybiochem,

Open the HELP file -> stringRegExp -> Enter

Look for (?m) down the list.

There you will find the ^ and $.

Sorry for the late reply though. Running a bit late have to catch up.

regards,

ptrex