RegExp StringPos

[elektron]

I'm writing a program to correct some formatting issues in print images of a HCFA-1500 form generated by an old SCO/UNIX system. Here's the deal.

The input: a text file.

The output: a text file.

I can't show the input due to HIPAA regulations, and I can't really show the program either, but I can let you guys know what I need and how the input is parsed into the program.

Essentially:

$hFile = FileOpen("PRINTIMAGE.txt", $READ_ONLY);

$content = FileRead($hFile, -1); read the file to end into a variable.

So, now the entire contents of this file is dumped into this one little variable. Now I can go through that string and apply regex matches to the string. Here's the problem though. In many cases, it's not as simple as a regex replace. I need to find the character position of the string that matches my regular expression and apply some formatting to it.

Can someone help me?

-- elektron

[jchd]

Without further specification, there's nothing anyone can do for you. Remember a regexp is a program which will apply rules that you create to match, extract or replace relevant data. If you can't make the rules explicit, no regex is going to automagically appear!

[elektron]

Without further specification, there's nothing anyone can do for you. Remember a regexp is a program which will apply rules that you create to match, extract or replace relevant data. If you can't make the rules explicit, no regex is going to automagically appear!

The problem is not writing the regex to create the matches, the problem is finding a way, in autoit, to find the location of the string that matches the given regular expression. Is that possible? Here's what I want:

; Continuation of code posted earlier

; ...

; $iPos = StringRegExMatch($content, $regexp) ; Returns the character position of the first regex match.

[Melba23]

elektron,

If you set the flag parameter to 1 or 2 the position of the end of the match is returned in @extended - here is a short example:

#include <Array.au3>

Global $aArray[3]

$sString = "Bla bla bla"

$iOffset = 1

For $i = 0 To 2
    $aTemp = StringRegExp($sString, "a", 1, $iOffset)
    $iOffset = @extended
    $aArray[$i] = $iOffset - 1
Next

_ArrayDisplay($aArray)

That should allow you to locate your match.

M23

[elektron]

elektron,

If you set the flag parameter to 1 or 2 the position of the end of the match is returned in @extended - here is a short example:

#include <Array.au3>

Global $aArray[3]

$sString = "Bla bla bla"

$iOffset = 1

For $i = 0 To 2
    $aTemp = StringRegExp($sString, "a", 1, $iOffset)
    $iOffset = @extended
    $aArray[$i] = $iOffset - 1
Next

_ArrayDisplay($aArray)

That should allow you to locate your match.

M23

That's great. It works! But I ended up changing my algorithm completely.

expandcollapse popup

Func parseFile($fileData)
;MsgBox(0, "", $fileData)

; If you're not me and you're reading this crap, then I just want to tell you, I'm sorry for you. Here DEFINITELY be dragons.
$sigRegExp = '([s]{6}[SIGNATURE]+[s][ON]{2}[s][FILE]{4}[s]+)([0-9]{2}[s][0-9]{2}[s][0-9]{2}[s]+)([s]{6}[SIGNATURE]+[s][ON]{2}[s][FILE]{4})'
$svcRegExp = '((((([0-9]{2})[s][s]?){3})+)[d]{2}sds{3}d{5}s{15}ds{5}(([dd]{2}s){2})sds{8}d{10})'
$fileByLine = StringSplit($fileData, @LF)
$returnResult = ""
For $i = 1 To $fileByLine[0] Step 1
  $sigResult = StringRegExp($fileByLine[$i], $sigRegExp, 0)
  $svcResult = StringRegExp($fileByLine[$i], $svcRegExp, 0)
  If( $sigResult = 1 ) Then ; We've found a match
   $buf = StringSplit($fileByLine[$i], "")
   $buf[46] = $buf[44]
   $buf[47] = $buf[45]
   $buf[44] = "2"
   $buf[45] = "0"
   $correctedString = ""
   For $j = 1 To $buf[0] Step 1
    $correctedString &= $buf[$j]
   Next
   $fileByLine[$i] = $correctedString
  EndIf
  If( $svcResult = 1 ) Then ; We've found a match
   $buf = StringSplit($fileByLine[$i], "")
   $buf[3] = $buf[4]
   $buf[4] = $buf[5]
   $buf[5] = "2"
   $buf[6] = "0"
   $buf[11] = " "
   $buf[12] = " "
   $buf[14] = " "
   $buf[15] = " "
   $buf[17] = " "
   $buf[18] = " "
   $buf[19] = ""
   $correctedString = ""
   For $k = 1 To $buf[0] Step 1
    If( $k = 45 ) Then
     $correctedString &= " "
    EndIf
    $correctedString &= $buf[$k]
   Next
   $fileByLine[$i] = $correctedString
   ; ConsoleWrite($i & ". " & $correctedString)
  EndIf
  $returnResult &= $fileByLine[$i] & @LF
Next
Return $returnResult
EndFunc

It's nasty, but not bad for a half-hours worth of coding. That regex matches the entire line(s) I need to extract from the data. Thanks for the help though!

[Robjong]

Hi,

Your pattern should work but if you care about it here is some advice to correct your pattern.

There is no need to put escape sequences in a character set.

[d]   -> d
[0-9]  -> d
[dd] -> d

If you want to match words do not put them in a character set, a character set will match any of characters in the set. For example "[ON]{2}" will match ON as well as NO.

[ON] -> (ON)   ; using a capturing group
[ON] -> (?:ON) ; using a non-capturing group
[ON] -> ON     ; just the word

It will be something like this:

$sigRegExp = 's{6}SIGNATUREsONsFILEs+d{2}sd{2}sd{2}s{6,}SIGNATUREsONsFILE'
$svcRegExp = '(((((d{2})ss?){3})+)d{2}sds{3}d{5}s{15}ds{5}((dds){2})sds{8}d{10})'

Edited May 30, 2012 by Robjong

[elektron]

Hi,

Your pattern should work but if you care about it here is some advice to correct your pattern.

There is no need to put escape sequences in a character set.

[d]   -> d
[0-9]  -> d
[dd] -> d

If you want to match words do not put them in a character set, a character set will match any of characters in the set. For example "[ON]{2}" will match ON as well as NO.

[ON] -> (ON)   ; using a capturing group
[ON] -> (?:ON) ; using a non-capturing group
[ON] -> ON     ; just the word

It will be something like this:

$sigRegExp = 's{6}SIGNATUREsONsFILEs+d{2}sd{2}sd{2}s{6,}SIGNATUREsONsFILE'
$svcRegExp = '(((((d{2})ss?){3})+)d{2}sds{3}d{5}s{15}ds{5}((dds){2})sds{8}d{10})'

Awesome, thanks for this.