Why Doesn't my Application Display Unicode?

Question

I have created an MFC application from scratch being careful from the start to use Unicode aware structures such as CStringW, LPCWSTR Etc. to store and process data. Unicode is also defined in the project.

Since I only one speak one language I tried the following test to ensure that a Unicode string was processed and stored correctly by the application.

In one of the Edit Boxes I entered ALT + 2061 and ALT + 2066 to display symbols not available on my keyboard link text The only thing displayed on the editbox is a square. Tryed the same think in Notepad and the symbols were correctly displayed. Is this just a font issue? If so what font should I be using?

UPDATE:

I copied several symbols from the unibook and pasted them into the Edit boxes. Apart from a small few symbols they were processed and saved correctly so I am happy with that.

Happy Christmas.

Answer 1

Yes. This is a font issue.

I would not worry about this, because the fonts installed on your customer's machine will be different than yours. If he needs kanji, he will have it. You can though force a specific font. Then you will see it correctly on your Windows.

I also recommend you against doing unicode with W (widechar) stuff. See http://stackoverflow.com/questions/1049947/should-utf-16-be-considered-harmful

Answer 2

U+2061 is "FUNCTION APPLICATION", which is a special nonprinting "operator" character, U+2066 is not defined yet (as of Unicode 5.2). Thus, what you see in your application is correct; probably you entered different codes in Notepad?

Answer 3

Make sure that the Alt-key input method you are describing works in other programs. It doesn't seem to work on my WinXP system.

You may just want to download Unibook from unicode.org and copy the chars from that just to be sure.

Answer 4

It is not a directly a font issue, it's an issue with your ALT+number method.

Using the alt key and the numberpad it's possible to enter any ASCII code, even those that don't have a key on the keyboard. It is not possible to enter unicode characters in this manner - everytime you enter a 4 digit code, it is just getting wrapped around to an 8 bit ASCII code:

2061 -> 100000001101 -> 00001101 -> ASCII character 13, a musical note

The reason it shows up as a square is because character 13 is a control character, so most modern fonts have no visual representation.

If you want to enter unicode characters using the numberpad then you'll need to use the extended mode Microsoft added - press and hold alt, then press and hold the plus (+) key on the numberpad. Now type the number of the unicode character you want and then release the alt and + key.

Answer 5

Go to Spy++ and look at the window class. Is it Edit, or is it some variant of RICHEDIT? A RichEdit window requires you to specify the font, which you can do with CRichEditCtrl::SetDefaultCharFormat.

Answer 6

I suggest you run the function IsWindowUnicode() somewhere in your app on the edit control. If it returns TRUE, then you know for sure that your app and your controls are unicode aware (and then it most likely is a font issue). However if it returns false, you have to look in your code and your project settings to find out why your app/window is not unicode aware.