What is the 'pattern' I need for stringregexp

[element72 0]

Say I'm looking at an html source that somewhere contains <td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>.

I want to capture fox (or cat) but in between <td> and </td>.

I just keep getting it wrong. My best effort is $array = StringRegExp($HTML, "(?i)<td>([(cat)(fox)]?)</td>")

[ViciousXUSMC 82]

There are better ways to parse html than regex, this is the first way I have it working.  The RegEx pro's are sure to come post some better ways.

 

$sString = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$sNewString = StringRegExpReplace($sString, "(?i)<td>(.*?)</td>", "$1")

MsgBox(0, "", $sNewString)

 

[element72 0]

Say I have many different symbols/characters in between <td> and </td> but all I want to identify is if fox or cat is in between them, while ignoring everything else.

[ViciousXUSMC 82]

Yeah I thought you wanted everything between <td> I need to fix it.

Almost not quite, seems it will not pick up the repeats present between <td> if its more than once.

#Include <Array.au3>

$sString = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$aNewString = StringRegExp($sString, "(?i)<td>.*?(cat|fox).*?</td>", 3)

_ArrayDisplay($aNewString)

 

Edited September 23, 2015 by ViciousXUSMC

[element72 0]

This should be possible with stringregexp() right? Or can this be done with a different function like IE UDF?

Edited September 23, 2015 by element72

[element72 0]

help me... please  Still can't figure it out... StringRegExp($HTML, "(?i)<td>([(fox)(cat)(.*)][(fox)(cat)(.*)][(fox)(cat)(.*)]?)</td>",3)

[jguinch 393]

This ?

#Include <Array.au3>
$HTML = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$array = StringRegExp($HTML, "(?i)<td>[^<]*(cat|fox)[^<]*</td>", 3)

_ArrayDisplay($array)

 

[ViciousXUSMC 82]

This ?

#Include <Array.au3>
$HTML = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$array = StringRegExp($HTML, "(?i)<td>[^<]*(cat|fox)[^<]*</td>", 3)

_ArrayDisplay($array)

 

Its missing the last set where both cat & fox are in a single string between <td></td>  I tried to look online how to "repeat" the capture but could not quite figure it out.  I assume some kind of back reference or something would be needed. Been eyeing this thread myself since I could not figure it out.

Closest I found was this, but none of it worked: http://www.regular-expressions.info/captureall.html

 

Edited September 24, 2015 by ViciousXUSMC

[jguinch 393]

Sorry, I didn't see the last elems contains two results.

Try this one :

#Include <Array.au3>
$HTML = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$aRes = StringRegExp($HTML, "(?i)(fox|cat)(?=[^<]*<\/td>)", 3)
_ArrayDisplay($aRes)

 

[ViciousXUSMC 82]

Sorry, I didn't see the last elems contains two results.

Try this one :

#Include <Array.au3>
$HTML = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

$aRes = StringRegExp($HTML, "(?i)(fox|cat)(?=[^<]*<\/td>)", 3)
_ArrayDisplay($aRes)

 

Works  not sure how lol.

(?i) Case Insensitive
(fox|cat) match & capture fox or cat
(?=[^<]*<\/td>) look ahead and match but do not capture Not < 0 or more times and </td> one time.

I just do not understand how this one picks up the duplicate cat/fox matches when the other regex was not able to.  I'll save this for notes though. 

[mikell 729]

ViciousXUSMC,

Hehe 

"(fox|cat)[^<]*</td>"  => matches a string containing fox|cat AND zero or more non-< chars up to </td>
When done, the search continues after the </td>

"(fox|cat)(?=[^<]*<\/td>)"  => matches fox|cat followed by 0 or more non-< chars and </td>
When done, the search continues after the match
It's the magic of the lookahead which is a zero-length assertion  

BTW [^<]* is used to check that there is no html tag between fox|cat and </td>

Edit :
Note that you can also do it like this "(?i)(?=(fox|cat)[^<]*</td>)"  

Edited September 24, 2015 by mikell

[ViciousXUSMC 82]

Thumbs Up!

Tested it on regex101 yes how you explained it makes sense, didn't think you could match like that only looking forward and ignoring what is behind, but it's a clever solution. 

[mikell 729]

Hmmmyes, but as usual in such topics about regex the OP doesn't provide enough details concerning the source... thus the regex fails if there are html font tags : <td><b>CATdss foxdf sdf</b></td> 

[jguinch 393]

Yes, for me this kind of regex should be done in two pass for more flexibility :

#Include <Array.au3>
$HTML = "<td>aaaa ©FOXaaa ttt</td> <td>aaacat bbbb aaa</td> <td>CATdss foxdf sdf</td>"

For $elem In StringRegExp($HTML, "<td>(.+?)</td>", 3)
    $aRes = StringRegExp($elem, "(?i)(cat|fox)", 3)
    If IsArray($aRes) Then _ArrayDisplay($aRes)
Next

 

[element72 0]

Hmmmyes, but as usual in such topics about regex the OP doesn't provide enough details concerning the source... thus the regex fails if there are html font tags : <td><b>CATdss foxdf sdf</b></td> 

good point. How would I solve this kind of problem? or include that into the regexp?

[mikell 729]

#Include <Array.au3>
$HTML = "<td>aaaa <b>©FOX</b>aaa ttt</td> <td>aaacat bbbb aaa</td> <td> <i><strong>CAT</i></strong>dssfox df sdf</td>"

$aRes = StringRegExp($HTML, "(?i)(fox|cat)(?=.*?</td>)", 3)
_ArrayDisplay($aRes)

Using 2 passes as said jguinch could be a more robust solution though, should be tested