Delete the 2nd until the last sentence of a set paragraph range on a Word

[Servant]

I tried a lot of techniques but still have no luck..

How can I delete the second sentence until the last sentence of a set paragraph range on a Microsoft Word document?

#include <Word.au3>

Global $oWord, $oDoc

$oWord = _Word_Create()
$oDoc = _Word_DocGet($oWord, 1)

Global Const $Count = $oDoc.Paragraphs.Count

For $i = 0 To $Count - 1
   $oRange = _Word_DocRangeSet($oDoc, -1, $wdParagraph, $i, $wdParagraph, 1)

   ; Here will be placed the missing code
Next

Sample of the beginning of a Word document:

This is a sentence 1. This is a sentence 2. This is a sentence 3.

This is a sentence 4. This is a sentence 5. This is a sentence 6.

This is a sentence 7. This is a sentence 8. This is a sentence 9.

Sample of the final result:

This is a sentence 1.

This is a sentence 4.

This is a sentence 7.

Edited April 15, 2014 by Servant

[FireFox]

Hi,

I don't use the Word UDF, so this may not be the best method :

;$s is the text of your paragraph
;$s2 is the replacement text
$s2 = StringRegExpReplace($s, "(?m)(.*?\.)(?:.*?)$", "$1")

_

Br, FireFox.

Edited April 15, 2014 by FireFox

[water]

This deletes from the first "." in each paragraph to the end of the paragraph.

#include <Word.au3>

Global $oWord = _Word_Create()
If @error <> 0 Then Exit MsgBox(16, "Word UDF: _Word_DocFind Example", "Error creating a new Word application object." & @CRLF & "@error = " & @error & ", @extended = " & @extended)
Global $oDoc = _Word_DocOpen($oWord, "Test Leerzeichen.docx", Default, Default, True)
If @error <> 0 Then Exit MsgBox(16, "Word UDF: _Word_DocFind Example", "Error opening 'Test Leerzeichen.docx'." & @CRLF & "@error = " & @error & ", @extended = " & @extended)

Local $oRangeFound, $oRangeText
$oRangeFound = _Word_DocFind($oDoc, ".", 0) ; Search the whole document
If @error Then Exit MsgBox(16, "Word UDF: _Word_DocFind Example 3", "Error locating the specified text in the document." & @CRLF & "@error = " & @error & ", @extended = " & @extended)
; Create a new range (duplicate to not alter the result of the find operating)
$oRangeText = $oRangeFound.Duplicate
$oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, $WdCharacter, 1, $wdParagraph, 1) ; Move the start of the range past the "." and the end of range to the end of the paragraph
$oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, Default, Default, $wdCharacter, -1) ; Move the end of the range one character to the left to not delete the new line character
$oRangeText.Text = ""
While 1
    $oRangeFound = _Word_DocFind($oDoc, ".", 0, $oRangeFound) ; Search the next "."
        If @error Then ExitLoop
    $oRangeText = $oRangeFound.Duplicate
    $oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, $WdCharacter, 1, $wdParagraph, 1)
    $oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, Default, Default, $wdCharacter, -1)
    $oRangeText.Text = ""
WEnd

[Servant]

This is my better solution:

expandcollapse popup

#include <Word.au3>

Global $oWord = _Word_Create()
If @error <> 0 Then Exit MsgBox(16, "Word UDF: _Word_DocFind Example", "Error creating a new Word application object." & @CRLF & "@error = " & @error & ", @extended = " & @extended)

Global $oDoc = _Word_DocGet($oWord, 1)
If @error <> 0 Then Exit MsgBox($MB_SYSTEMMODAL, "Word UDF: _Word_DocGet Example", _
"Error accessing collection of documents." & @CRLF & "@error = " & @error & ", @extended = " & @extended)

Global $pCount = $oDoc.Paragraphs.Count
Local $oRange, $oRange2, $sCount, $1st, $p
Local $sFindText, $oFind, $oRangeFound, $oRangeText

For $i = 0 To $pCount - 1
  $oRange = _Word_DocRangeSet($oDoc, -1, $wdParagraph, $i, $wdParagraph, 1)

  $sCount = $oRange.Sentences.Count

  While 1
    If $sCount > 1 Then
      $oRange2 = _Word_DocRangeSet($oDoc, -1, $wdParagraph, $i, $wdSentence, 1)
      $1st = $oRange2.Text
      $oRange2 = _Word_DocRangeSet($oDoc, -1, $wdParagraph, $i, $wdParagraph, 1)
      $p = $oRange2.Text

      $sFindText = StringReplace($p, $1st, "")
      $sFindText = StringReplace($sFindText, @CR, "")
      $oFind = _Word_DocFind($oDoc, $sFindText, $oRange2, Default, Default, False, False, False)

      If @error <> 0 Then
        $oRangeFound = _Word_DocFind($oDoc, ".", $oRange2) ; Search the $oRange2

        If @error Then Exit MsgBox(16, "Word UDF: _Word_DocFind Example 3", "Error locating the specified text in the range." & @CRLF & "@error = " & @error & ", @extended = " & @extended)

        ; Create a new range (duplicate to not alter the result of the find operating)
        $oRangeText = $oRangeFound.Duplicate
        $oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, $WdCharacter, 1, $wdParagraph, 1) ; Move the start of the range past the "." and the end of range to the end of the paragraph
        $oRangeText = _Word_DocRangeSet($oDoc, $oRangeText, Default, Default, $wdCharacter, -1) ; Move the end of the range one character to the left to not delete the new line character
        $oRangeText.Text = ""
      Else
        $oFind.Delete
      EndIf

    EndIf

    $oRange2 = _Word_DocRangeSet($oDoc, -1, $wdParagraph, $i, $wdParagraph, 1)
    $sCount = $oRange2.Sentences.Count
  WEnd
Next

But please review the sentence below:

The FYE 2012 Transfer Pricing Report shows that the set of comparable companies chosen to benchmark the O&M services has a three year period weighted average (“PWAVG”) interquartile range (“IQR”) of 3.3 percent to 19.3 percent and a one year IQR of 4.0 percent to 21.8 percent.

After that code was run, the sentence in the new document was:

The FYE 2012 Transfer Pricing Report shows that the set of comparable companies chosen to benchmark the O&M services has a three year period weighted average (“PWAVG”) interquartile range (“IQR”) of 3.

And this does not seem to happen consistently I think it's because of my new code.

For example, the following sentence before and after did not have this problem.

Before:

For transaction 4, the mark-up on total cost (OI/TC) KPMG calculated is -23.8 percent, while the mark-up of total cost PwC presented in the report is -31.3 percent, which happened to be mark-up on total revenue (OI/Revenue);

And After:

For transaction 4, the mark-up on total cost (OI/TC) KPMG calculated is -23.8 percent, while the mark-up of total cost PwC presented in the report is -31.3 percent, which happened to be mark-up on total revenue (OI/Revenue);

Is it possible to fix this issue?

I think when this code execute:

$oFind = _Word_DocFind($oDoc, $sFindText, $oRange2, Default, Default, False, False, False)

  If @error <> 0 Then

...and produce the error "4 - $sFindText could not be found" it will then execute your code but it will treat the decimal point in a number as the end of the sentence..

Edited April 22, 2014 by Servant

[water]

 

How can I delete the second sentence until the last sentence of a set paragraph range on a Microsoft Word document?

The solution depends on how you define a "sentence". You need to search for ". " for "sentecnes" within a paragraph or for ".P" where "P" is the control character for a new paragraph.

[jchd]

Would that be ".¶" aka '.' & ChrW(0xB6) ?

[water]

Yes.



 

[jchd]

Beware that a correctly written paragraph may in the general case also end in various other characters like ! ? ) … » ” „ ’ ‼ ⁇ ⁈ ⁉ ❩ ❫ ¿ ¡ and most probably another set of quotes and exotic punctuation marks when, for instance the paragraph ends with a citation from some non-english language. Incorrect punctuation only adds more difficulty.

So why not rely on paragraph marks only? Sounds more reliable.

[water]

Because in post 2 the OP showed an example with only period as an ending character.

But you are correct, sentences can end with a lot of characters.

Let's see what the OP needs

[jchd]

I was just saying, both for the OP and/or for future reference.

[water]

I see. So I think to properly identify "sentences" within a paragraph SRE would be needed?

[jchd]

I don't have Office installed here and I don't remember how powerful/painful regexpes are in Word.