I can obviously do it one character at a time using EM_GETCHARFORMAT, but it is extremely slow.
One idea is to somehow use the ITextDocument/ITextFont interfaces, the other is to use the EM_STREAMOUT message and manually parse the RTF. But I can't decide which approach is better and am very fuzzy on the implementation details. Wil开发者_如何学Pythonl appreciate any help, thanks!
I've found a solution that satisfies me and think will share it with you:
The ITextRange
interface contains a very useful method Expand
which can be used to find continuous runs of constant character (tomCharFormat
) and paragraph (tomParaFormat
) formatting.
Here is some sample code (warning: code is a proof-of-concept spaghetti without any error handling, apply refactoring as needed):
// Get necessary interfaces
IRichEditOle* ole;
SendMessage(hwndRichEdit, EM_GETOLEINTERFACE, 0, (LPARAM)&ole);
ITextDocument* doc;
ole->QueryInterface(__uuidof(ITextDocument), (void**)&doc);
long start = 0;
// Get total length:
ITextRange* range;
doc->Range(start, start, &range);
range->Expand(tomStory, NULL);
long eof;
range->GetEnd(&eof);
// Extract formatting:
struct TextCharFormat { long start, length; DWORD effects; }
std::vector<TextCharFormat> fs;
while(start < eof - 1)
{
doc->Range(start, start, &range);
long n;
range->Expand(tomCharFormat, &n); // <-- Magic happens here
ITextFont* font;
range->GetFont(&font);
DWORD effects = 0;
long flag;
font->GetBold(&flag);
if (flag == tomTrue) effects |= CFE_BOLD;
font->GetItalic(&flag);
if (flag == tomTrue) effects |= CFE_ITALIC;
font->GetUnderline(&flag);
if (flag == tomSingle) effects |= CFE_UNDERLINE;
font->GetStrikeThrough(&flag);
if (flag == tomTrue) effects |= CFE_STRIKEOUT;
if (effects)
{
TextCharFormat f = { start, n, effects };
fs.push_back(f);
}
start += n;
}
精彩评论