UTF-16 and TextOutW()
in [MFC]
|
Prev: Queryinterface then Release fails
Next: https://accountservices.msn.com/EmailPage.srf?emailid=5d4c1ef525bfe00f&ed=BwZ12Kf8vfLCHttpServer missing in Dev. Studio 2008
From: Peter Olcott on 18 Nov 2007 11:28 Does TextOutW() always use UTF-16 as its basis, or are there some other levels of indirection? How can I tell exactly which UTF-16 characters are represented by a typeface?
From: David Webber on 18 Nov 2007 12:45 "Peter Olcott" <NoSpam(a)SeeScreen.com> wrote in message news:MKZ%i.534$ip1.466(a)newsfe21.lga... > Does TextOutW() always use UTF-16 as its basis, Yes, I believe so. > How can I tell exactly which UTF-16 characters are represented by a > typeface? Fonts contain lots of information about what they can do. Have a look at the structures FONTSIGNATURE, LOCALEFONTSIGNATURE, and CHARSETINFO in the help system. I looked into these a while ago but my memory is vague: I wanted to find out about support for musical flat, double sharp, etc symbols, but was disappointed. I came to the conclusion that fonts could implement symbols or not on an individual basis, and there was nothing to tell me if an individual one was implemented. [If anyone finds a way, I'd love to hear about it!] But I think these structures, (and the APIs which fill them in) should be able to tell you if the characters needed by a certain language are present. Dave -- David Webber Author of 'Mozart the Music Processor' http://www.mozart.co.uk For discussion/support see http://www.mozart.co.uk/mozartists/mailinglist.htm
From: Mihai N. on 19 Nov 2007 03:19 > Does TextOutW() always use UTF-16 as its basis, or are there > some other levels of indirection? UTF-16 always. > How can I tell exactly which UTF-16 characters are > represented by a typeface? The only way to tell is by checking if there is a glyph present for the character you want. For that you can use GetGlyphIndices (which does not work above BMP though), you can use ScriptGetCMap (Uniscribe) or you can do your own ttf/otf cmap parsing. But you should also remember that there are other things going: ligatures (so in the font you can have a glyph for the "xy" ligature but no "x" or "y" glyphs stand-alone) You also have font linking/mapping/fallback/substitution done by various Windows layers. So you can have a character on screen that is not in the currently selected font, but still shows ok. -- Mihai Nita [Microsoft MVP, Windows - SDK] http://www.mihai-nita.net ------------------------------------------ Replace _year_ with _ to get the real email
From: David Webber on 19 Nov 2007 04:46 "Mihai N." <nmihai_year_2000(a)yahoo.com> wrote in message news:Xns99ED35E4B4AMihaiN(a)207.46.248.16... > For that you can use GetGlyphIndices (which does not work > above BMP though), Hi Mihai, a couple of questions, if I may: Is the BMP *precisely* the set which can be represented by a single word in UTF-16 (ie without surrogates)? I am interested in Musical symbols - code points 1D100-1D1FF - which are clearly outside this range. But three useful ones are at 266D, 266E, 266F (flat, natural, sharp) - do you reckon it should work for those? [I'm suddenly thinking I could parse my strings at drawing time for those characters and replace them with 'b', ' ', '#' if not present in the font. That way titles like "Sonata in Eb" could be represented properly if the flat character is there, and with b (a poor substitute) if it isn't.] I don't know when unicode fonts with 1D100-1D1FF will become commonplace, but even if they did, one would have to know where the origin of the symbol is, in order to place it correctly with music. Do you know if there is a standard for this? Dave -- David Webber Author of 'Mozart the Music Processor' http://www.mozart.co.uk For discussion/support see http://www.mozart.co.uk/mozartists/mailinglist.htm you can use ScriptGetCMap (Uniscribe) > or you can do your own ttf/otf cmap parsing. > > But you should also remember that there are other things > going: ligatures (so in the font you can have a glyph for > the "xy" ligature but no "x" or "y" glyphs stand-alone) > You also have font linking/mapping/fallback/substitution > done by various Windows layers. So you can have a character > on screen that is not in the currently selected font, > but still shows ok. > > > > -- > Mihai Nita [Microsoft MVP, Windows - SDK] > http://www.mihai-nita.net > ------------------------------------------ > Replace _year_ with _ to get the real email
From: Peter Olcott on 19 Nov 2007 09:10
"Mihai N." <nmihai_year_2000(a)yahoo.com> wrote in message news:Xns99ED35E4B4AMihaiN(a)207.46.248.16... >> Does TextOutW() always use UTF-16 as its basis, or are >> there >> some other levels of indirection? > UTF-16 always. > > >> How can I tell exactly which UTF-16 characters are >> represented by a typeface? > The only way to tell is by checking if there is a glyph > present for the character you want. > For that you can use GetGlyphIndices (which does not work > above BMP though), That is plenty good enough for the time being. It will probably be a long time before I support any glyphs above the BMP. > you can use ScriptGetCMap (Uniscribe) > or you can do your own ttf/otf cmap parsing. > > But you should also remember that there are other things > going: ligatures (so in the font you can have a glyph for > the "xy" ligature but no "x" or "y" glyphs stand-alone) > You also have font linking/mapping/fallback/substitution > done by various Windows layers. So you can have a > character > on screen that is not in the currently selected font, > but still shows ok. > > I probably would map ligature to their individual Unicode values, and thus not retain the fact that they were ligatures. > > -- > Mihai Nita [Microsoft MVP, Windows - SDK] > http://www.mihai-nita.net > ------------------------------------------ > Replace _year_ with _ to get the real email |