ITextRange::SetChar method (tom.h)

Sets the character at the starting position of the range.

Syntax

HRESULT SetChar(
  long Char
);

Parameters

Char

Type: long

New value for character at the starting position.

Return value

Type: HRESULT

The method returns an HRESULT value. If the method succeeds, it returns S_OK. If the method fails, it returns one of the following error codes. For more information about COM error codes, see Error Handling in COM.

Return code Description
E_ACCESSDENIED
Text is write-protected.
E_OUTOFMEMORY
Out of memory.

Remarks

ITextRange::SetChar lets you specify the precise character code to use. However, string literals with similar looking glyphs can be misleading.

The characters set by this method are LONG instead of a BSTR. This hides the way that they are stored in the backing store, (as bytes, words, variable-length, and so forth).

Frequently on systems that do not have automatic word-wrapping, documents have hard carriage returns inserted just for line breaks. The following code shows a simple, but not perfect, way to convert such hard carriage returns back to blanks for the story that is associated with the range r.

    Sub EnableWrap(r As ITextRange)   // Convert false hard CRs to soft
        r.SetRange 0, 0               // r is set at start of story
        While r.Move(tomParagraph)    // Go to start of next paragraph
            If r.MoveWhile(C1_WHITE, 1) = 0 Then    // Next char isn't white space
                r.Move tomCharacter, -1
                r.SetChar = Asc(" ")    // Replace CR by blank
            End If
        Wend        // Loop till no more CRs in story
    End Sub

Alternatively, you could use the following inside the IF loop.

r.MoveStart tomCharacter, -1        // Select previous char (the CR)
r = " "        // Replace it with a blank

This approach enables you to wrap the text to other widths. However, the algorithm isn't perfect: it assumes that a hard carriage return that is followed by anything other than white space (blank, tab, line feed, carriage return, and so forth) should be replaced by a blank. The algorithm also assumes that the carriage return character is a single character like carriage return or the Unicode end-of-paragraph (EOP) 0x2029 character. And, the combination carriage return and line feed isn't matched and would require writing and executing more code (or use FindText(^p)). Another caution is that there are other cases, such as mixed code and documentation, where the algorithm does not work correctly.

However, ITextRange::SetChar is more efficient than a replace operation that is accomplished by a delete followed by an insertion. Thus, rewriting the code without using ITextRange::SetChar would probably be much slower.

The Char property, which can do most things that a characters collection can, has two big advantages: it can reference any character in the parent story instead of being limited to the parent range, and it's significantly faster, since LONGs rather than range objects are involved. Because of these advantages, the Text Object Model (TOM) does not support a characters collection.

Requirements

Requirement Value
Minimum supported client Windows Vista [desktop apps only]
Minimum supported server Windows Server 2003 [desktop apps only]
Target Platform Windows
Header tom.h
DLL Msftedit.dll

See also

Conceptual

ITextRange

Reference

Text Object Model