RegEx - Trying to 'split string in string'

[supersonic]

Hi!

RegEx! ... once again. I'm sorry for that... But I don't get it solved on my own...

Here's the code I have so far:

#include <Array.au3>

Local $sTmp = ""

$sTmp &= "#4096750#" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=,FNGcJ7KMw5kCYNcZH24P9sdINiM=" & @LF
$sTmp &= "#4093600#,F47T1OveZnEgmjKE84Nfm+8bQQ4=;BundledBy,#4093601#" & @LF

Local $aTmp = StringRegExp($sTmp, ",(.*)\n|,|;", 3)

_ArrayDisplay($aTmp)

What I like to achieve:

- Read a '@LF'-delimited line, if it contains a ','.

- If a line contains a ',':

a] ... read it to the end of the line (= '\n')

b] ... or read the whole line and return all ','-delimited elements

c] ... or return the element between ',' and ';'.

The result should be an array containing this:

[0] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[1] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[2] EN28JlozIKrsBlTtM8ZKWw9auck=

[3] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[4] EN28JlozIKrsBlTtM8ZKWw9auck=

[5] FNGcJ7KMw5kCYNcZH24P9sdINiM=

[6] F47T1OveZnEgmjKE84Nfm+8bQQ4=

Please, could somebody give me a clue?

Greets,

-supersonic.

Edited March 10, 2011 by supersonic

[kylomas]

supersonic,

Don't know 'bout regexp but this is how I would do this (if I understand you correctly):

$arr1 = stringsplit($stmp,@lf)

for $i = 1 to $arr1[0]
    $arr2 = stringsplit($arr1[$i],',')          ; each element is a line
    
        for $j = 1 to $arr2[0]                  ; each element is a comma delimited value within the current line
            ;
            ;  do somethig with the segment...
            ;
        Next
next

kylomas

[Malkey]

Try this.

#include <Array.au3>

Local $sTmp = ""

$sTmp &= "#4096750#" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=,FNGcJ7KMw5kCYNcZH24P9sdINiM=" & @LF
$sTmp &= "#4093600#,F47T1OveZnEgmjKE84Nfm+8bQQ4=;BundledBy,#4093601#" & @LF

Local $aTmp = StringRegExp($sTmp, ",(.+?=)", 3)

_ArrayDisplay($aTmp)

[kylomas]

malkey,

what happens to the end of the string that is not delimited by a "=" sign?

kylomas

[supersonic]

Malkey,

thank you for that... It works fine!

But as kylomas stated, how to handle items without a '=' at the end?

[Melba23]

supersonic,

Try this one - it copes with elements without "=" at the end:

#include <Array.au3>

Local $sTmp = ""

$sTmp &= "#4096750#" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8,EN28JlozIKrsBlTtM8ZKWw9auck=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck,FNGcJ7KMw5kCYNcZH24P9sdINiM=" & @LF
$sTmp &= "#4093600#,F47T1OveZnEgmjKE84Nfm+8bQQ4=;BundledBy,#4093601#" & @LF

Local $aTmp = StringRegExp($sTmp, "(?U),(.*)(?=[\v,;^#])", 3)

_ArrayDisplay($aTmp)

The results match your requirements in the first post.

M23

[kylomas]

supersonic,

Let me borrow this thread for a minute...

@malkey and M23 - splitting a file at @CRLF and then splitting each record at some delimiter is a common deal. Do you guys always use regexp for this? I have a feeling that I've been doing this half-ass backward with the multiple splits and iterations.

kylomas

[Melba23]

kylomas,

It all depends on the form of the information you wish to extract. By definition you need a pattern of some sort if you want to use an SRE - if it is not "regular" then you are wasting your time! But SREs are not the only way to go - as GEOSoft often remarks, one of tricks of working with SREs is knowing when not to use them - although the little devils are really useful when you do!

So do not give up on iterative splits - they still have a place at times. But learning to use SREs (even to my amateurish level) is a valuable skill - and this is a good place to start your education. I do not want to put you off - but be warned, SREs are probably the most difficult things I have ever tried to master in coding.

M23

[supersonic]

Thank you Melba!

[kylomas]

M23,

Yes, of course, my question exactly. The pattern for all flat files is each liine is delimited somehow, normally by @CRLF, sometimes by "0x00".

I think I'm going to do some timings and see what I get.

Thanks for the advice!

kylomas

@supersonic - and thanks for the use of your thread, good luck!

[supersonic]

Hi!

How is it possible to return the second token only? (Please see 1st post for example.)

Like this:

[0] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[1] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[2] DvIFPwMcjKaVa04PlLQxDg1AKf8=

[3] F47T1OveZnEgmjKE84Nfm+8bQQ4=

It there an "escape sequence" like "if a string was found, go on with the next line"...???

Greets,

-supersonic.

Edited March 11, 2011 by supersonic

[Melba23]

supersonic,

I see you have not yet started the tutorial on that website I linked to.

Try this:

Local $aTmp = StringRegExp($sTmp, "(?U),(.*)(?=[\v,;]).*?\v", 3)

M23

[supersonic]

Melba,

shame on me - you are right!

But BIG thank you for helping me out again!

Edited March 11, 2011 by supersonic

[supersonic]

Melba,

I had to modify the SRE statement to "(?U),(.*)(?=[\v,;]).*?" in order to return all four lines.

So, I removed the last "\v".

Have I done it right?

[Melba23]

supersonic,

I get:

DvIFPwMcjKaVa04PlLQxDg1AKf8=
DvIFPwMcjKaVa04PlLQxDg1AKf8=
DvIFPwMcjKaVa04PlLQxDg1AKf8=
F47T1OveZnEgmjKE84Nfm+8bQQ4=

when I run the SRE I posted this morning with your initial data:

$sTmp &= "#4096750#" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=" & @LF
$sTmp &= "#4096756#,DvIFPwMcjKaVa04PlLQxDg1AKf8=,EN28JlozIKrsBlTtM8ZKWw9auck=,FNGcJ7KMw5kCYNcZH24P9sdINiM=" & @LF
$sTmp &= "#4093600#,F47T1OveZnEgmjKE84Nfm+8bQQ4=;BundledBy,#4093601#" & @LF

Are you saying that you do not?

M23

[supersonic]

Hello Melba,

I'd checked it several more times - I've got only three lines...

BUT - I've found the reason for that:

When the last line doesn't end with "@LF" then only three lines return...

With an trailing "@LF" four lines return...

Edited March 14, 2011 by supersonic

[Melba23]

supersonic,

Try this one:

Local $aTmp = StringRegExp($sTmp, "(?U),(.*)(?=[\v,;]).*?[\v=#]", 3)

Tricky little devils, these SREs!

M23

[GEOSoft]

M23

The secret here is <pattern>(?:\v|$)+

SmOke_N usually prefers <pattern>(?:\r\n|\r|\n|\z)

[Melba23]

George,

Thanks as always.

M23

[GEOSoft]

NP

At some point in time, if and when I get some time, I plan on adding another toolbox Insert Menu category to auto-insert commonly used portions of expressions such as what I gave you. I just need to get the time to add them to the database and to add the menu items.