Help with StringRegExReplace and subpattern replacement

[chadskinner]

I have a bunch of log files I need to change about %30 of the line where a : was used instead of a coma. I can't just replace all the colons or the time/date stamp would get messed up. The kicker is there is roughly 15 different syntax's before and after the colon.

Example

Result1:Sub1:55.09

Result1:Sub2:63.10

NewResult1:105.10

NewResult2:110.20

I need to replace all those colons in about a 100+ files. So I've been trying stringregexreplace

I find my files via filefindnext file turn each one into a string, replace the parameters, write it back to the file.

So far I've Tried things like

stringregexreplace($fileString,'(?!{6,9}d{1})((?!{3}d{1})',',')

This one just replaces the entire string with a coma. I guess I don't understand what 'non-capturing' group means.

So I tried it this way

stringregexreplace($fileString,'?!{6,9}d{1}(?!{3}d{1}',',')

again replaces the entire string.

Does anyone know if it's possible to just replace a substring inside a pattern match and not destroy the rest of this. I think I could write wayyyy too much code and do this with StringRegEx and a lot of loops but it seems like StringRegExReplace should be able to do this.

simple example

$fileOpen = FileOpen($dir & 'my.log',0)

$fileRead = FileRead($fileOpen)

fileclose($fileOpen)

$fileString = String($fileRead)

MsgBox(1,'FileString', $fileString)

$newString = StringRegExpReplace($fileString,',[A-Za-z]{3}[0-9]{1}(',',')

MsgBox(1,'NewString', @error & $newString )

Any thoughtsreferences would be greatly appreciated.

Thanks,

Chad

Edited July 18, 2012 by chadskinner

[chadskinner]

BTW - each of the @$@##$$ smiley faces is actually comma close parens but I'm too stupid to figure out how to get this editor to stop doing that.

[UEZ]

Can you give another example how your log file looks like incl. the time and how the result should be?

Thanks,

UEZ

[AZJIO]

; $fileString = FileRead($dir & 'my.log')



$fileString = _

        'Result1:Sub1:55.09' & @CRLF & _

        'Result1:Sub2:63.10'



$newString = StringRegExpReplace($fileString, ':[A-Za-z]{3}d', '')

MsgBox(0, 'Yes?', @error & @CRLF & $newString)

[chadskinner]

Entry has colons separating values instead of comas

INFO,2012-05-23 15:03:58,934,Result1:Reading1:94.75

INFO,2012-05-23 15:03:58,934,Result1:Reading2:94.75

INFO,2012-05-23 15:03:59,276,Result2:Reading1:81.21

INFO,2012-05-23 15:03:59,276,Result2:Reading2:81.21

INFO,2012-05-23 15:03:59,464,NewResult1:61.58

INFO,2012-05-23 15:03:59,558,NewResult2:55.67

INFO,2012-05-23 15:03:59,651,NewResult3:59.13

INFO,2012-05-23 15:03:59,744,NewResult4:69.21

INFO,2012-05-23 15:03:59,869,NewResult5:62.32

INFO,2012-05-23 15:03:59,947,NewResult6:60.78

INFO,2012-05-23 15:04:00,088,NewResult7:59.16

INFO,2012-05-23 15:04:00,259,NewResult8:57.67

INFO,2012-05-23 15:04:00,322,NewResult9:59.28

INFO,2012-05-23 15:04:00,384,NewResult10:55.65

Part of the log file that is correct

INFO,2012-05-23 15:05:42,658,Result[0]V,1.11

INFO,2012-05-23 15:05:44,592,Result[1]V,1.12

INFO,2012-05-23 15:05:46,559,V1,1.11

INFO,2012-05-23 15:05:48,509,V2,0.00

INFO,2012-05-23 15:05:50,459,sysV,1.11

INFO,2012-05-23 15:05:52,378,sys2V,1.34

INFO,2012-05-23 15:05:54,312,sys3V,1.32

INFO,2012-05-23 15:05:56,246,sys4V,1.32

INFO,2012-05-23 15:05:58,243,sys5V,1.51

INFO,2012-05-23 15:05:58,539,main,77.50

How the bad section should look (note time stamp is preserved but the rest of the colons are gone)

INFO,2012-05-23 15:03:58,934,Result1,Reading1,94.75

INFO,2012-05-23 15:03:58,934,Result1,Reading2,94.75

INFO,2012-05-23 15:03:59,276,Result2,Reading1,81.21

INFO,2012-05-23 15:03:59,276,Result2,Reading2,81.21

INFO,2012-05-23 15:03:59,464,NewResult1,61.58

INFO,2012-05-23 15:03:59,558,NewResult2,55.67

INFO,2012-05-23 15:03:59,651,NewResult3,59.13

INFO,2012-05-23 15:03:59,744,NewResult4,69.21

INFO,2012-05-23 15:03:59,869,NewResult5,62.32

INFO,2012-05-23 15:03:59,947,NewResult6,60.78

INFO,2012-05-23 15:04:00,088,NewResult7,59.16

INFO,2012-05-23 15:04:00,259,NewResult8,57.67

INFO,2012-05-23 15:04:00,322,NewResult9,59.28

INFO,2012-05-23 15:04:00,384,NewResult10,55.65

My plan was to read the file into a string, parse the string with regexreplace and write it back to the file. I have 100's of these improperly formatted files.

[BrewManNH]

Try this on one of your text files and see if it does what you're looking for.

#include <file.au3>
#include <_array.au3>
Global $aFile
_FileReadToArray("test.txt", $aFile)
For $I = 1 To $aFile[0]
    $aFile[$I] = StringRegExpReplace($aFile[$I], "Result(d+):", "Result$1,")
    $aFile[$I] = StringRegExpReplace($aFile[$I], "Reading(d+):", "Reading$1,")
Next
_ArrayDisplay($aFile)

I'm just learning RegEx so this could probably be done in one line instead of 2, but I have no idea how to do it that way yet.

Also, change the text file name in the _FileReadToArray line to one of your filenames, with path if not in the same folder as the script.

[BrewManNH]

I gave it some more thought and came up with a little bit better RegEx, probably still not as good as it could be, but I'm learning. muttley

#include <file.au3>
#include <_array.au3>
Global $aFile
_FileReadToArray("test.txt", $aFile)
For $I = 1 To $aFile[0]
    $aFile[$I] = StringRegExpReplace($aFile[$I], "([a-zA-Z]{6})(d+):", "$1$2,")
Next
_ArrayDisplay($aFile)

This pattern will match any string of 6 letters followed by any number of digits and then a colon. So, it matches the "Result##:" and "Reading##:" strings and replaces the colons with commas accordingly.

[Malkey]

This example shows the uselessness of arrays.

; If any of the three characters before the ":" are a space or a colon (:), then do not replace that colon with a coma.
$sSring = StringRegExpReplace(FileRead("test.txt"), "([^: ]{3}):", "$1,")

ConsoleWrite($sSring & @LF)

[chadskinner]

Holy crap, I owe you a beer sir. You taught me more in that one line than I thought possible. Elegant job, Thanks!!

[BrewManNH]

This example shows the uselessness of arrays.

Not sure how it demonstrates anything about the utility of using arrays, but a very nice approach to the situation nonetheless.

[UEZ]

Here another way:

$sText = _
"INFO,2012-05-23 15:03:58,934,Result1:Reading1:94.75" & @CRLF & _
"INFO,2012-05-23 15:03:58,934,Result1:Reading2:94.75" & @CRLF & _
"INFO,2012-05-23 15:03:59,276,Result2:Reading1:81.21" & @CRLF & _
"INFO,2012-05-23 15:03:59,276,Result2:Reading2:81.21" & @CRLF & _
"INFO,2012-05-23 15:03:59,464,NewResult1:61.58" & @CRLF & _
"INFO,2012-05-23 15:03:59,558,NewResult2:55.67" & @CRLF & _
"INFO,2012-05-23 15:03:59,651,NewResult3:59.13" & @CRLF & _
"INFO,2012-05-23 15:03:59,744,NewResult4:69.21" & @CRLF & _
"INFO,2012-05-23 15:03:59,869,NewResult5:62.32" & @CRLF & _
"INFO,2012-05-23 15:03:59,947,NewResult6:60.78" & @CRLF & _
"INFO,2012-05-23 15:04:00,088,NewResult7:59.16" & @CRLF & _
"INFO,2012-05-23 15:04:00,259,NewResult8:57.67" & @CRLF & _
"INFO,2012-05-23 15:04:00,322,NewResult9:59.28" & @CRLF & _
"INFO,2012-05-23 15:04:00,384,NewResult10:55.65" & @CRLF & _
"INFO,2012-05-23 15:05:42,658,Result[0]V,1.11" & @CRLF & _
"INFO,2012-05-23 15:05:44,592,Result[1]V,1.12" & @CRLF & _
"INFO,2012-05-23 15:05:46,559,V1,1.11" & @CRLF & _
"INFO,2012-05-23 15:05:48,509,V2,0.00" & @CRLF & _
"INFO,2012-05-23 15:05:50,459,sysV,1.11" & @CRLF & _
"INFO,2012-05-23 15:05:52,378,sys2V,1.34" & @CRLF & _
"INFO,2012-05-23 15:05:54,312,sys3V,1.32" & @CRLF & _
"INFO,2012-05-23 15:05:56,246,sys4V,1.32" & @CRLF & _
"INFO,2012-05-23 15:05:58,243,sys5V,1.51" & @CRLF & _
"INFO,2012-05-23 15:05:58,539,main,77.50"

$sNew = StringRegExpReplace($sText & @CRLF, "(.*)(:)(d*.d*rn)", "$1,$3")
MsgBox(0, "Test", $sNew)

Br,

UEZ