Inserting characters by their Unicode numbers in AI CS3 on WindowsXP...

[_Don_Glascock_]

Hi, folks --

We're running Illustrator CS3 on Windows XP, which is said by Adobe to support Unicode well.

When using the Text tool and while editing a line of text, one can insert a character by its numeric code using this technique: press & hold the Alt key, type in a decimal number *from the keyboard's keypad*, and release the Alt key.

Web pages like http://unicode.org/charts/PDF/U0370.pdf show the Unicode numberic values in hex for various characters, which I convert to decimal using the hex/decimal radio buttons of the Windows Calculator running in Scientific view.

Since inserting Unicode character 68(dec) (44(hex)) does indeed insert capital D using the Alt-UnicodeValueInDecimal-AltRelease trick, one might guess that the lower-case Greek letter Mu in Arial font could therefore be inserted by specifying its Unicode value of 956(dec) (3BC(hex)) after confirming its presence in the desired font through the Windows Character Map application. But in fact, Mu does not appear: instead, lower-case Mu appears if one supplies 1254(dec) using this trick. 1254(dec) should equal 4E6(hex), which is proving difficult to locate in Arial font in the Windows Character Map app.

If I'm working with a Unicode-friendly font, and if I can insert the desired character from that same font into MS-Word using Word's "type the number in hex, select the number, and press Alt-X" Unicode-insertion trick, then what am I doing wrong in AI-CS3? Is there some other numbering scheme at work, here?

Thanks for your help!

Don

[Harron_K_Appleman]

Use the Glyphs palette.

[_Don_Glascock_]

Thanks, Harron: I'm grateful for your prompt reply! I should have said previously that I know about the Glyphs palette, but for various reasons, I need to use the Unicode insertion trick I described previously.

What is the coding sequence that Abobe Illustrator CS3 uses in which 1254(dec) translates to lower-case Mu, but 956(dec) does not? I experiemented with Octal, but that doesn't work, and doesn't explain why 68(dec) does correctly insert a D.

Don

[Harron_K_Appleman]

The Alt+{code} method (or "trick," if you prefer) in Windows applications is generally limited to the extended ASCII set (decimal 0 through 255). Although there are clearly some oddball exceptions, such as your Alt+1254 example, Alt+{code} does not work for characters above decimal 255. You'll notice that Windows Character Map stops showing the 'Keystroke: Alt+{code}' shortcut in the bottom right corner once you are past 255.

I don't think what you are describing is limited to Illustrator. Have you tried it in Notepad?

In Notepad, with Arial Unicode, 1245 gives me mu (as does 0181, which is the mu within the ASCII set). 68 gives me D. 956 gives me ╝ (double-stroke lower right corner). 0956 gives me ¼ (the one-quarter fraction).

The same thing happens in Illustrator (12) for me.

[Harron_K_Appleman]

EDIT: First sentence of previous post

...generally limited to the standard ASCII set (decimal 0 through 127) and extended ASCII set (decimal 128 through 255).

[_Don_Glascock_]

Thanks for your observations, Harron.<br /><br />* In Notepad, with Arial Unicode, 1245 gives me mu (as does 0181, which is the mu within the ASCII set). 68 gives me D. 956 gives me &#9565; (double-stroke lower right corner). 0956 gives me ¼ (the one-quarter fraction). <br /><br />That's what confused me: in Illustrator, character codes greater than 255 seemed to be mapped, but not to the characters one would expect. I had written a Visual Basic program that would step through the character codes near 956 and then near 1254, but in neither case did I find character sequences that fit Unicode standard. Your explanation that only the ASCII and Extented ASCII codes are mapped makes sense (though it also makes me sad. :-)<br /><br />* Have you tried it in Notepad?<br /><br />As you report, Notepad and Illustrator seem to support only the ASCII and Extended ASCII codes. But WordPad seems to support Unicode: the Alt-<HexValue>-ReleaseAlt trick in WordPad produces Mu when one inserts 956, and 955 correctly inserts Lambda.<br /><br />It's probably true that apps are generally becoming more & more Unicode compliant, and perhaps Illustrator will one day offer Unicode keycode mappings (or perhaps there already is some other way that we don't know about...?). As you report, there's still the Glyphs palette for AI CS3, but I was hoping to find ways of getting Greek letters into (gulp) Illustrator 10 using this trick. It sounds like characters not in ASCII and Extended ASCII will be difficult indeed to insert into Illustrator 10 in Arial font: outlines don't really look as sharp, and Symbol font doesn't seem to have a "Bold" weight. <br /><br />Thanks for helping me to understand what's going on.<br /><br />Have a great day!<br /><br />Don

[_Don_Glascock_]

After working with this some more, it seemed wise to add a few more notes:

* In Notepad, with Arial Unicode, 1245 gives me mu (as does 0181, which is the mu within the ASCII set).

Just to show how confusing this subject is, 0181 is indeed mapped to lower-case Mu, but lower-case Mu in the Extended ASCII set is E6(hex) or 230(decimal). 181(dec) or B5(hex) is a line-drawing character, at least in this Extended ASCII table:

http://www.cdrummond.qc.ca/cegep/informat/Professeurs/Alain/files/ascii.htm

... and Notepad seems to agree with this table.

The only reason I mention this is to clarify that if you want to insert Extended ASCII characters using the Alt-Number-ReleaseAlt trick, you must apparently supply the code in decimal with no leading zero's.

And it would appear that the character code sequence holding 0181 and 1254 (and any of the other such ill-fitting characters mentioned in this thread) remains a mystery. I suspect that I could make good use of the sequence if I understood it, and I hope Adobe will one day either switch to Unicode or point my sometimes-seemingly-blind eyes toward the right spot in their documentation.

Don

[_Don_Glascock_]

Ah-hah!

In Illustrator, when using the Alt-Number-ReleaseAlt trick, if the number you provide is prefixed with a zero, then the *ANSI* character sequence is used:

http://cloford.com/resources/charcodes/symbols.htm

... and if you DO NOT prefix the number with a zero, then the *Extended ASCII* sequence is used.

Now, 1254(dec) is 4E6(hex). And 230(dec) is E6(hex). Both 1254 and 230 (using the Alt-Number-ReleaseAlt trick) produce lower-case Mu. So it would appear that only the least-significant byte is kept, and leading bytes are trashed, which is why the sequence appears to cycle through the Extended ASCII table this way.

Likewise for the ANSI table: 0956 (ie, 956(dec) for the ANSI table) is 3BC(hex), and is rendered as the "1/4" character, which, in the ANSI table, is BC(hex) or 188(dec). Again, it appears that all but the least significant byte is trashed, so that 0956 does what 0181 does: they both produce a lower-case Mu.

And Alt-U-Number-ReleaseAlt doesn't really work in Illustrator or Notepad, as suggested by the Alt-Zero-Number-ReleaseAlt trick for the ANSI character set. WordPad does Unicode (decimal) by default, and you can get ANSI characters via Alt-Zero-Number-ReleaseAlt.

The ANSI table doesn't really help me with Greek symbols for our math equations, unfortunately, but at least it's nice to know what to expect & what options remain on the table, so to speak.

Don

[_Abraham_Kifle_]

This site has a unicode character picker for all languages.
www.2keystrokes.com

[Harron_K_Appleman]

Interesting site, but I don't see how it helps the OP, who absolutely requires a keyboard shortcut method (to wit, is unwilling or not able to use the Glyphs palette nor the copy-and-paste feature of Windows CharMap).

[_Don_Glascock_]

* This site has a unicode character picker for all languages. www.2keystrokes.com

Thanks!

* Interesting site, but I don't see how it helps the OP, who absolutely requires a keyboard shortcut method (to wit, is unwilling or not able to use the Glyphs palette nor the copy-and-paste feature of Windows CharMap).

I don't wish to bore you with the details, so I'll just say that we're essentially locked into Illustrator-10 for a while longer, even though it doesn't support Unicode very well: no glyphs palette, and essentially no Unicode support. The Alt-trick stuff was a last hope for getting Unicode characters into AI-10, and it seemed best to ask about this with respect to AI-13, since 10 is so old. With a clearer understanding of ANSI vs Extended ASCII character sets, and how they relate to Unicode, I feel I've pursued this issue as far as possible.

Since we are already required to batch-convert our ~5-10k AI-8 and older drawings into AI-10 anyway, I added hooks the Visual Basic code to change characters from our purchased font sets into SymbolMT and Arial*MT, which seems to be going very well: I completed a 614-file run yesterday evening with few issues. Symbol won't look as nice, but we won't have to worry about missing fonts or require fonts on the audience PCs, or package fonts with our presentations.

Thanks again for all of the great help, everyone!

Don